According to atmospheric modeling and satellite observations, cold fronts can cause trans-regional transport (TRT) of haze particles from the North China Plain (NCP) to Yangtze River Delta (YRD) in winter. However, compositions and aging of haze aerosols during the TRT have not been studied. We showed the TRT PM_2.5 dominated by organic matter (OM) (30%) and secondary inorganic ions (36%) in the NCP and 29% and 60% in the YRD. Following the TRT, abundant spherical primary OM particles (i.e., tarballs) (71% by number) mainly from residential coal burning in rural areas of the NCP unexpectedly occurred in the YRD. The inert tarballs display similar sizes (~300 nm) and O/C ratios (~0.15), but the mixture of nitrate, sulfate, and secondary OM as the coatings completely convert the hydrophobic tarballs into hydrophilic ones in the TRT. The aging and transport of tarballs from the NCP to YRD further indicate that the TRT not only brought various trace gases (e.g., CO, SO₂, NO_x, and VOCs) but also carried large numbers of nanosized primary particles (e.g., tarball, metal, fly ash, and soot) with secondary coatings over 1000 km. The findings suggest that these many nanosized tarballs containing brown carbon and highly toxic species in the NCP influence regional climate and human health in northern and eastern China, which needs more attention. Although the NCP and YRD have different energy consumption structures in winter and are two isolated administrative regions, we emphasize the need for a coordinated cross-regional emission reduction strategy for TRT haze control.

Sand and dust storms(SDS) severely hinder global sustainable development. Currently, most research focuses on assessing the health impacts of SDS. This study is the first to integrate SDS hazards with the vulnerability and exposure characteristics of affected populations, energy, and food systems to develop a comprehensive risk assessment model for analyzing the compounded risks of SDS to public health, energy security, and food security under different Shared Socioeconomic Pathways (SSPs) by employing the Intergovernmental Panel on Climate Change (IPCC) risk framework. The results show that sub-Saharan Africa and regions in the Middle East are key hotspots for the triple risks of SDS-related health, energy, and agricultural threats. Most countries face SDS risks to vulnerable populations, energy systems, and agricultural production, with dominant risks varying across different regions. For instance, in China, the western regions are at higher SDS-energy risks, while the eastern regions are more vulnerable to SDS-agricultural risks. In contrast, India is predominantly exposed to SDS-agricultural risks, and North Africa is mainly threatened by SDS-energy risks. Under high-emission scenarios, SDS-health risks in densely populated areas of eastern China, the Middle East, and West Africa are expected to intensify due to population growth. Even under low-emission scenarios, India, the Middle East, and North Africa may still face SDS-risks related to imbalances in energy and food supply. This study highlights the critical importance of promoting clean energy transitions, building robust health protection systems, and advancing climate-resilient agricultural practices in vulnerable regions to support global sustainable development.

Addressing the issue of polycyclic aromatic hydrocarbons (PAHs) and their nitro-oxidized derivatives (N+DNPAHs) pollution in China requires a comprehensive grasp of their refined spatio-temporal distribution. Although the existing simulation studies have filled the gap of refined data in the observational studies, they usually neglect the oxidation products of PAHs. In this study, a numerical model capable of simulating atmospheric PAHs and N+DNPAHs was developed using the CMAQ framework, and the observational data were collected to verify the applicability of the model in China. The model performed better in simulating PAHs compared to N+DNPAHs. The simulated PAHs and N+DNPAHs were 9.16% and 66.94% lower than the observed values, with MFBs of -0.01 and -0.07, and MFEs of 0.76 and 0.89, respectively. Furthermore, the spatio-temporal distribution of PAHs and N+DNPAHs in China and their influencing factors were analyzed and quantified. PAHs pollution was mostly concentrated in the Central, North, and East China, while N+DNPAH pollution was primarily concentrated in the Central and East China. PAHs and N+DNPAHs exhibited similar seasonal variation patterns: winter (53.42 ng·m^-3, 0.49 ng·m^-3) > autumn (27.18 ng·m^-3, 0.40 ng·m^-3) > spring (17.56 ng·m^-3, 0.19 ng·m^-3) > summer (12.18 ng·m^-3, 0.17 ng·m^-3), with emissions, diffusion conditions, and chemical reactions being important influencing factors. In addition, this study assessed the health risks associated with atmospheric PAHs and N+DNPAHs in China. The average lifetime cancer risk (ILCR) was 1.93 × 10^-6, with approximately 93.83% of residents living in areas where the ILCR exceeds the acceptable risk threshold of 1.0 × 10^-6. The further analysis shows that PAHs and N+DNPAHs contribute 96.82% and 3.18% to ILCR, 94.91% and 5.09% to cancer incidence, respectively.

The Beijing-Tianjin-Hebei region, one of the most polluted and densely populated areas in China, has experienced significant improvements in air quality as a result of the Clean Air Action Plan. However, the problem of ozone (O₃) pollution has become increasingly severe, and the compound pollution of O₃ and PM_2.5 has emerged as a major challenge. Hydroxyl radicals (OH), as a key factor in determining atmospheric oxidative capacity (AOC), play an important role in the removal of primary pollutants and the formation of secondary pollutants. This study utilizes both observation-based models (OBM) and parameterization methods to assess the long-term trends of OH concentrations in the Beijing-Tianjin-Hebei region from 2015 to 2023. The results show that OH concentrations exhibit significant diurnal and seasonal variations, peaking in the afternoon and during the summer. Both methods reveal a notable increasing trend in OH concentrations, with rates of 0.113 × 10⁶ cm^-3 year⁻¹ (OBM) and 0.112 × 10⁶ cm^-3 year^-1 (parameterization). The interannual increase in OH is most pronounced in Handan and Xingtai, while Tianjin shows a smaller rate of increase, a trend consistent in both methods. Analysis of parameterization results indicates that the increase in OH concentrations is closely related to a significant decrease in NO₂ concentrations (-38.48%), with the largest reduction occurring in summer (-45.92%), corresponding to the highest increase in OH (+53.42%). Additionally, photolysis rates (J_NO₂) play a significant role in the increase of OH during the summer. Correlation analysis reveals a positive correlation between OH and O₃, and a significant negative correlation between OH and NO₂, CO, SO₂, and PM_2.5. 

During the winter of 2023–2024, PM_2.5 concentration in Shanghai, China, exhibited a unexpected significant resurgence, reaching the highest levels since 2018–2019. However, the underlying causes remain unclear, necessitating a comprehensive investigation into key pollution drivers and effective mitigation strategies.This study analyzed observed compositions of PM_2.5 data from winter 2023–2024 in Shanghai, employing the Community Multiscale Air Quality (CMAQ) model for simulation and characterization, supplemented by XGBoost and SHAP machine learning methods for further interpretation. The findings identified four major haze pollution episodes during this period, with nitrate emerging as the most significantly increased component in three events. Source apportionment revealed that 56.2% of the nitrate originated from local emissions in Shanghai and its adjacent Jiangsu and Zhejiang provinces, while 26.6% was transported from the North China Plain. Further analysis indicated that volatile organic compounds (VOCs), relative humidity (RH), and ammonia (NH₃) were the primary factors influencing nitrate concentration, underscoring the critical role of precursor availability under unfavorable meteorological conditions.Building on this analysis, we conducted sensitivity experiments on the three key precursors of nitrate—NO_x, VOCs, and NH₃—to evaluate the effectiveness of different emission reduction strategies. The results demonstrated that the response of particulate nitrate and total PM_2.5 concentrations varied with reduction percentage and study area. In Shanghai's winter conditions, ammonia reduction was the most effective strategy for mitigating nitrate pollution. However, for overall PM_2.5 mitigation, VOCs reduction was most effective at lower reduction levels, while NH₃ reduction became more impactful at higher reduction percentages. Notably, these effects varied across different regions.This research provides critical insights into the factors contributing to this year’s PM_2.5 pollution episodes in Shanghai and offers evidence-based guidance for targeted emission control strategies, particularly in years with adverse meteorological conditions.

Since 1996, the Flexible Array of Radars and Mesonets (FARM) facility instruments, including the Doppler on Wheels (DOW) mobile radars and in situ surface instrumentation, have deployed in the eyewalls/eyes of 20 hurricanes. This dataset comprises near-surface fine temporal and spatial resolution data inside a diversity of hurricane eyewalls, allowing for resolution of small-scale HBL structures, near the points of landfalls, over a range of hurricane intensities (Cat 1-4).  A unique dataset was obtained during the 2024 hurricane season that allows, for the first time, assessment of the three-dimensional winds below radar level and how these winds relate to previously mapped near-surface DOW analyses detailing the kinematics of coherent boundary layer structures.  During the landfall of Hurricane Helene (Cat 4), two co-located DOWs, with 10-m masts measuring wind speed/direction, were deployed within a “nanonet” ring of ~150 m diameter 1-m AGL Pods, measuring wind speed/direction.  These data are being used to directly relate specific DOW-observed HBLR structures, including vertical winds directly determined from vertical-pointing radar data and DOW-observed proximate (within 500 m - 3 km range) very low-level horizontal wind fields, and Pod and mast array surface wind gusts, divergence and vorticity.  Are analysis of the two-DOW and anemometer nanonet data collected during Hurricane Helene will examine how specific HBLR affect near surface (1-10 m AGL) winds and kinematic fields at the smallest spatial and temporal scales.  Our results will be compared to previous DOW analyses with a focus on the following questions.  How do directly-measured vertical wind speeds depend on location within specific HBLR; are predictions dating back to Wurman and Winslow (1998) verifiable using directly measured, nearly co-located radar and in situ measurements?  Do directly measured vertical winds validate the multiple-Doppler derived results of Kosiba and Wurman 2014 and Kosiba et al. 2025? 

How do tornadoes form?  How do tornadoes strengthen, maintain, and dissipate? What are the winds like in tornadoes very near the ground and how do they vary with height?  The BEST (Boundary-layer Evolution and Structure of Tornadoes) project seeks to answer these questions using integrated data from the DOW mobile radars, soundings, and surface weather stations.  The 2024 tornado season started early for the BEST project, intercepting a tornado near Harlan, IA on April 26^th.  Several weeks later, on 21 May, the DOW collected data from less than a quarter mile away as a tornado traveled through the town of Greenfield, IA.  In town, the DOW-calculated wind speeds were as high as 309-318 mph, some of the fastest wind speeds ever determined.  Work is underway to integrate DOW-calculated wind metrics with detailed damage surveys.  Very high resolution dual-Doppler DOW data and thermodynamic data were collected in the Duke, OK tornado on 23 May, allowing for the exploration of processes that contribute to tornado evolution and intensity changes.  Changes in buoyancy very near the tornado and rapid generation of vertical vorticity along convergence boundaries helped maintain the Duke, OK tornado.  While case studies are useful, we are working to integrate these very valuable data sets collected during the 2024 tornado season with previously collected data sets in order to develop a more comprehensive picture of low-level tornado structure and tornado genesis and maintenance mechanisms.  This presentation will discuss preliminary analyses of the 2024 tornado data and how these fit into our evolving theories of tornado formation and evolution and low-level wind profiles.      

A microburst is a severe small‐scale meteorological event that develops rapidly, producing intense downdrafts that result in catastrophic divergent winds near the ground. The fine‐scale structure and evolution of real‐case microbursts are rarely analyzed due to the limitation of regular observational platforms and the computational capacity for numerical simulations. On 12 September 2020, a microburst was effectively captured by China's S‐band operational weather radar network and an X‐band polarimetric phased‐array radar (XPAR). XPAR observations can identify precursor signatures of the rapidly evolving microburst before the occurrence of surface wind disasters, demonstrating superior spatiotemporal resolution in monitoring compared to the S‐band radars. To disclose the evolution of small‐scale structure and the underlying physical processes, this study utilizes the Weather Research and Forecasting (WRF) model and the ensemble Kalman filter (EnKF) system to assimilate XPAR data. The simulation successfully reproduces the fine‐scale structure and evolution of the microburst with a misocyclone. The analysis indicates that the microburst's downdraft is primarily triggered by the middle‐level hydrometeor loading. The misocyclone generates a downward‐directed perturbation pressure gradient then accelerates the microburst's downdraft toward the surface. This study is the first observation and simulation of a real‐case microburst using the WRF‐EnKF system assimilating XPAR data. The investigation of the impact of misocyclone on the microburst enhances our understanding of microbursts' forcing mechanism.

Accurate and timely estimation of surface precipitation is essential for decision-making during extreme weather events and effective water resources management. Polarimetric weather radar is the primary tool used for quantitative precipitation estimation (QPE). However, significant differences in topographical environments and meteorological characteristics in different precipitation regimes can degrade the generalization capability of a deep learning model trained in a specific region when applied to other regions. In addition, many areas have a lack of rain gauges and/or radar observations that are sufficient to train a machine learning model. This paper presents a Domain-Adaptive Regulation Transfer Model (DARTM) for polarimetric weather radar quantitative precipitation estimation, which incorporates a long-distance regulation module and a short-distance adaptation module. To adaptively adjust the number of shared features in the neighborhood and alleviate the negative knowledge transfer problem, the DARTM model adopts a dynamic adaptive layer and a novel loss function. The feasibility and performance of the DARTM model are demonstrated and quantified using U.S. Weather Surveillance Radar-1988 Doppler (WSR-88D) observations and surface gauge measurements from three different precipitation regimes. Experimental results suggest that DARTM can not only improve the accuracy of precipitation estimation in data-scarce regions but also enhance the model’s generalization capability across diverse geophysical regions.

Despite consistent advancements in tropical cyclone (TC) track forecasting, accurately predicting unusual TC tracks remains a significant challenge. This study applies the orthogonal conditional nonlinear optimal perturbations (O-CNOPs) method to the Weather Research and Forecast (WRF) model to improve ensemble forecast reliability of unusual TC tracks. Ensemble forecast experiments were conducted for eighteen forecast periods involving five TCs, all of which exhibited sharp turns, to evaluate the effectiveness of O-CNOPs. Results demonstrate that O-CNOPs outperform singular vectors (SVs) and bred vectors (BVs) by providing more stable and reliable improvements in TC track forecasting skills, both from deterministic and probabilistic perspectives. Notably, O-CNOPs show a superior ability to generate ensemble members that accurately predict the sharp turns of TCs at lead times from one to five days. These results highlight the superiority of the O-CNOPs method over the SVs and BVs methods in improving the forecasting accuracy of TC tracks, particularly for forecasting unusual TC tracks. This study underscores the potential of O-CNOPs to be extended to real-time TC forecasting and to play an important role in operational TC forecasts.

Weather forecasting faces dual challenges from random errors and systematic biases. To systematically address both types of errors, this study proposes a unified generative super ensemble prediction framework. For large-scale weather systems, the framework leverages probabilistic diffusion models to learn climatological distributions from reanalysis data, effectively suppressing the growth of random errors. When applied to three major deterministic models (Pangu, Fengwu, and Fuxi), the anomaly correlation coefficient for 500 hPa geopotential at day 10 reaches 0.676, further improving to 0.683 when incorporating ECMWF ensemble forecasts. For mesoscale systems like tropical cyclones, this study develops a super ensemble prediction method that targets systematic biases by analyzing historical performance characteristics of different models through machine learning approaches. While maintaining the physical consistency of continuous fields, this method significantly improves the accuracy of key forecast indicators such as tropical cyclone track and intensity. This unified framework demonstrates the complementary advantages of generative models and machine learning in weather prediction across different scales, providing a new technical pathway toward theoretical predictability limits while establishing a foundation for the automated integration of meteorological forecast products into industry applications.

The multiscale interactions of initial condition (IC) perturbations for 12-h convection-permitting ensemble forecasting were explored in this study. The heavy-rainfall events occurring in South China in the rainy season during 2013–2020 were focused on and classified into three types: weak-forcing, strong-forcing, and tropical cyclone (TC) cases. The impacts of both large- and small-scale IC perturbations on multiscale characteristics of forecast perturbations and the forecast performance were investigated. Both the upscale growth of small-scale IC perturbations and downscale propagation of large-scale IC perturbations favored perturbation growth, with damping impacts of both processes beyond the first few hours. Compared with upscale growth, downscale propagation showed greater impacts on forecast perturbations in both magnitude and location and more evident variable-dependent impacts. The small-scale IC perturbations spread the high-probability rainfall downstream, while the large-scale IC perturbations spread the high-probability rainfall in all directions to the low-probability areas. Both of these two types of rainfall spreading led to increasing locational perturbations. Probabilistic forecasts of heavy short-duration rainfall benefited from both the small- and large-scale IC perturbations, leading to performance improvements of multiscale over single-scale IC perturbations. In particular, the small- and large-scale IC perturbations more benefited the forecasts of weak-forcing heavy rainfall with larger spatial forecast errors and at earlier lead times, respectively. The flow regimes with large forecast uncertainties, which were determined by the complicated interactions between synoptic-scale forcing and moist convection, directly boosted both upscale growth and downscale propagation.

The China Meteorological Administration (CMA) Global Ensemble Prediction System (GEPS) is a weather prediction model based on an atmosphere-land coupled system developed and operated by the CMA. It provides probabilistic weather forecasts for the 1-15 day range. Ensemble forecasting generates multiple forecast scenarios (ensemble members) using slightly different initial conditions or model configurations to capture the uncertainty inherent in weather prediction. The CMA GEPS is one of China's key tools for operational weather forecasting and research. Currently, the GEPS primarily focuses on atmospheric and land interactions without incorporating feedback from the ocean or sea ice, which enhances computational efficiency. However, interactions between the atmosphere and ocean play a critical role in many weather phenomena, particularly in the tropics, influencing events such as the Madden-Julian Oscillation (MJO), El Niño, and La Niña. Oceans significantly affect atmospheric circulation patterns, and neglecting these interactions can lead to less accurate forecasts, especially for medium- to long-range predictions.To address both the need for computational efficiency and the impact of ocean-atmosphere interactions, a new approach has been introduced to extend the CMA GEPS forecasts to 35 days, including subseasonal predictions. This method incorporates bias-corrected sea surface temperature (SST) forecasts from a fully coupled atmosphere-land-ocean model (2-tiered SST). Preliminary results indicate that using this SST forcing method provides significant improvements for weather predictions, as well as for weeks 2, 3, and 4 forecasts, when compared to fixed SST and relaxed SST (e-folding) methods. System performance is measured using key metrics such as 500hPa height, surface temperature, precipitation, and MJO skill scores (RMM1 & RMM2), along with 850hPa and 200hPa zonal winds, Outgoing Longwave Radiation (OLR), and the prediction of both strong and weak MJO cases.

Sea surface temperature anomalies (SSTAs) over the North Atlantic (NA) have a significant impact on the weather and climate in both local and remote regions. This study first evaluated the seasonal prediction skill of NA SSTA using the North American multi-model ensemble and found that its performance is limited across various regions and seasons. Therefore, this study constructs models based on the long short-term memory (LSTM) network machine learning method to improve the seasonal prediction of NA SSTA. Results show that the seasonal prediction skill can be significantly improved by LSTM models since they show higher capability to capture nonlinear processes such as the impact of El Nin ̃o-Southern Oscillation on NA SSTA. This study shows the great potential of the LSTM model on the seasonal prediction of NA SSTA and provides new clues to improve the seasonal predictions of SSTA in other regions.

It is an urgent need to understand the ability of current artificial intelligence (AI) models in simulating atmospheric mesoscale aspects. This paper compares mesoscale kinetic energy spectra from an 11-day experiment simulated by a novel AI-based model (Pangu) and a physics-based model (MPAS), using ERA5 reanalysis as a reference. Based on the commonly used evaluation metrics of latitude weighted root mean square error (RMSE) and anomaly correlation coefficient (ACC), the AI-based model has better short to medium-range weather forecasting skill compared to the physics-based model. However, the AI-based model cannot replicate the mesoscale -5/3 spectral slope and underestimates the mesoscale energy at wavelength smaller than 1000 km. As altitude increases and scale decreases, the deviation of the AI-based model from the reanalysis significantly increases. These features prove that the AI-based model has the lower effective resolution compared to the physics-based model with the close nominal resolution. Compared to the physics-based simulations, AI-based model has stronger downscale energy flux at larger mesoscales, which is dominated by divergent kinetic energy flux. But it rapidly becomes the weakest at smaller mesoscales. The diagnosed vertical velocity of AI-based model and its related budget terms are closest to those of the reanalysis at large scales. Overall, the AI-based model Pangu shows closer agreement with ERA5 at large scales, likely due to its use of the latter as training data, but significantly underestimates mesoscale kinetic energy compared to the physics-based model MPAS. Note that these findings are specific to the models and configurations used and should be interpreted with caution.

Rapid urbanization demands making cities and human settlements inclusive, safe, resilient, and sustainable. The present study uses multi-temporal Local Climate Zones (LCZ) to assess the spatio-temporal growth of the city over the past decade and its impact on its thermal environment. Kolkata is the 3rd largest Class-I city in India having tropical savanna climate that witnessed vertical growth near its peripheral areas and satellite cities between 2010 and 2020. Around 21% of the city has experienced change in its LCZ pattern. Horizontally the built-up area increased only 2% but LCZ 2 (compact mid-rise) has increased 10% at the expense of low-rise (LCZ 3 and 6) areas. LSTday between 2004-2024 increased significantly at a rate 0.12℃/year. In areas where LSTday has increased significantly, LCZ 2 has increased more than 50% while the low-rise (LCZ 3 and 6) and open mid-rise (LCZ 5) areas have decreased by around 4%, 11% and 34% respectively along with vegetation cover (LCZ B) decreasing around 5%. Near the CBD of the city, which is more than 200 years old, LSTday increased at a rate 0.15-0.24℃/year while, in the peripheral regions, the rate of the same is around 0.35-0.85℃/year. However, the human-made East Kolkata Wetlands (EKW), which is a designated Ramsar site, has shown a decreasing LSTday, along with some areas near Victoria and Fort William where greening has taken place. The fringe areas of the city are mostly going under unplanned construction of residential buildings on agricultural lands, which has the potential to change the thermal environment of that area. The study will help in identifying such potent areas to help the city planner in managing them in a more sustainable way.

Energy simulations are essential for designing energy-efficient and thermally indoor comfort buildings. These simulations require high-resolution local climate data, such as the Typical Meteorological Year (TMY). TMY is an hourly dataset of meteorological variables describing typical conditions for 12 representative typical months at a specific location. Traditionally, TMY datasets are derived from ground-based observations. However, with the accelerating impacts of global warming, future TMY datasets are also needed. Currently, future climate projections are primarily obtained from General Circulation Models (GCMs), but these datasets have coarse temporal and spatial resolutions, limiting their direct applicability to local studies. To bridge this gap, this study employs dynamical downscaling (DD) using the Model for Prediction Across Scales–Atmosphere (MPAS-A) to generate future TMY datasets for major cities in Indonesia. The DD was performed using a variable-resolution mesh with refinements up to 10 km, configured for the Maritime Continent region. The MPI-HR model from CMIP6 was selected as the optimal GCM for providing initial and boundary conditions (IBC). A climatological 6-hourly median from a 10-year period of the raw MPI-HR was calculated to create a 1-year dataset with 6-hour resolution, representing median conditions, as IBC. Two simulations were conducted to represent historical (2005-2014) and future (2025-2034) climates, with the future projection based on the RCP 8.5 scenario. The accuracy of the downscaling approach was evaluated by calculating correlation coefficients and root mean square differences between the historical downscaled results and observational datasets, including ground-based measurements and ERA5-land reanalysis. The final dataset provides seven key meteorological variables—global horizontal irradiance, direct normal irradiance, diffuse horizontal irradiance, air temperature, precipitation, wind speed, and relative humidity—with an hourly time resolution and 10 km spatial resolution.

Recently, rising temperature caused by global warming and the urban heat island phenomenon is becoming more serious, posing significant health risks and straining electricity demand. Green spaces and rivers are expected to mitigate urban thermal environments by influencing local microclimate. This study focuses on the cold-air drainage phenomenon in urban green spaces – a process in which cold air generated within green spaces on a calm nights spread into the surrounding city area through gravity flow. This phenomenon typically happens when average wind speeds range between 0.1~0.3 m/s with a cooling effect expanding 200~250 m into adjacent urban area (Narita et al., 2004). Additionally, this study explores the influence of small urban rivers on surrounding microclimates, where river-induced temperature drops have been reported to extend about twice the width of the river (Moriwaki et al., 2012). To investigate these phenomena, dense observations are required. However, commercially available meteorological equipment is bulky, costly and requiring an eternal power source, making it impractical for multi-point or mobile observations traversing the area. This research developed an affordable, compact, and energy-efficient observation system and conducted microclimate measurements at a large urban forest park and a small canal near our campus located in Chiba prefecture adjacent to Tokyo. Observations on the canal during summer revealed that there was no significant cooling effect of the canal on the surrounding thermal environment. In contrast, during winter, cold air accumulated within the river channel after sunset, indicating the potential cooling influence on nearby urban areas. In the urban forest park, observations in winter showed that while daytime temperatures within the forest were higher than those in the surrounding area, the nighttime temperature were lower. It suggests the urban forest park may amplify the winter cold in adjacent urban area.

Urban greening, particularly through increased tree cover, is widely recognized as one strategy for mitigating the impacts of urban heat island in cities.  However, an accurate quantification of the influence of urban vegetation on building energy consumption is difficult due to two ways feedbacks between building energy consumption and urban canyon microclimate. This quantification is however essential for developing effective climate adaptation and mitigation strategies. This study investigates the impact of urban trees on building air-conditioning (AC) energy demand by considering both temperature and humidity effects in seven cities (Riyadh, Phoenix, Dubai, New Delhi, Singapore, Lagos, and Tokyo) during the hot season. For this purpose, we develop a coupled urban ecohydrological and building energy model (Urban Tethys-Chloris - BEM), which can simulate the effects of urban densities and tree cover scenarios as well plant physiological and biophysical properties. Using this model, we evaluated the relative contributions of tree shade, temperature reduction, and increased humidity on AC energy consumption. Results showed that well-watered trees provided the largest average summer AC energy reduction (-17%) in hot-dry climates (Riyadh, Phoenix), primarily due to shading. While humid climates also saw AC energy reductions (averaging -6% to -9%) thanks to shade. However, the increased humidity induced by vegetation impacted dehumidification loads, particularly when high ventilation rates are applied. In these humid cities, optimal AC energy reduction was achieved with up to 40% tree cover. These findings offer valuable insights for urban planning strategies aimed at minimizing AC energy demand through urban greening.

In recent years, the worsening heat environment in urban areas has become a major social issue, with a significant increase in heatstroke cases. To address this, the Japan Meteorological Agency (JMA) and the Ministry of the Environment issue heatstroke alerts based on the Wet-Bulb Globe Temperature (WBGT). However, urban meteorological conditions vary greatly due to differences in land cover and building structures, causing significant microclimatic variations at a fine scale. As a result, large-scale forecast information alone is insufficient for accurately assessing heat conditions in living spaces. To implement effective heat countermeasures, detailed meteorological observations and thermal comfort evaluations tailored to specific environments are necessary.JMA provides weather forecasts based on numerical weather prediction models, with the Meso-Scale Model (MSM) being a key example. MSM predicts meteorological conditions around Japan at a 5 km grid resolution and is widely used for disaster prevention and daily weather information. However, MSM data represent average meteorological conditions within each grid and do not necessarily reflect actual conditions at specific observation sites. Since microclimates at the human scale are strongly influenced by land cover and buildings, MSM’s resolution is insufficient for detailed analysis, necessitating corrections based on observed data.This study conducts independent meteorological observations and applies statistical downscaling to MSM data using these observations. Outdoor observations consider land cover differences, while indoor observations account for variations in structures and floors. Between November and December 2024, temperature and humidity corrections using a naturally ventilated radiation shield showed acceptable accuracy outdoors and significant improvements indoors, despite not fully meeting the required range. However, outdoor observations exhibited large daytime temperature variations, highlighting the need for a system to reduce direct solar radiation effects. To enhance accuracy, forced ventilation was introduced for further measurements.At AOGS2025, we will present summer observation results and thermal comfort evaluations for different environments.

Japan’s active promotion of tourism in recent years has led to a significant increase in foreign tourists. However, alongside this trend, urbanization and the heat island phenomenon have worsened outdoor thermal environments, presenting a critical challenge. As demand for comfortable and sustainable outdoor spaces grows, such spaces require a comprehensive understanding of the relationship between thermal conditions and human behavioral patterns. Numerous studies have explored these relationships. For instance, Akagawa et al. (2007) investigated the thermal environment and visitors’ stay duration in a rooftop garden of a large commercial facility, revealing a negative correlation between the Standard Effective Temperature (SET*) and the likelihood of individuals remaining in an unshaded square on sunny days. Similarly, Ando et al. (2012) examined an outdoor amenity space in a major business district in central Tokyo, suggesting that people in summer exhibit higher tolerance for solar radiation than in autumn, indicating seasonal adaptation to heat.However, these studies were conducted over a decade ago, and with the progression of heat intensification, both human behavioral patterns and thermal environmental characteristics may have evolved. Furthermore, traditional research methods primarily relied on visual observation for behavioral analysis and multi-point measurements for microclimatic assessment, both of which labor-intensive and time-consuming, often restricting data collection to short periods. In this study, we developed an analytical workflow for efficiently processing and analyzing long-term observational data. Using high-resolution video captured by a custom-built camera (Sony Spresense) optimized for YOLOv8 analysis, we applied AI-based automatic detection (YOLOv8) to extract the number of users and their movement trajectories in outdoor spaces. Furthermore, we visualized detailed thermal environment distributions through 3D simulations, enabling extended analysis. At AOGS2025, we will present seasonal thermal environment evaluations for spring, early summer, and the post-rainy season, along with an investigation of user behavioral patterns based on these evaluations. 

With the occurrence of super-typhoon Haiyan (2013, 170kts) of intensity far exceeded the regular category-5 Tropical Cyclones (TCs) and its catastrophic damage, more than 10 years ago, Lin et al. 2014 proposed the need of adding Category ‘6’ to the Saffir-Simpson TC scale. In the Saffir-Simpson scale, each sequential jump of category from 1 to 5 is about 13-22kts, and the threshold of category-5 is 137kts. If adding 23kts and use 160kts threshold to form a new category ‘6’, these extra-ordinary TCs like Haiyan can be more accurately categorized. In addition, not only TC intensity (wind speed) matters, TC’s kinetic energy is a function of the square of wind speed while its destructive potential the cube of wind speed (Emanuel 2005). Therefore, the kinetic energy and destructive potential of Category ‘6’ TCs can be of 150% or double than that of ‘regular’ Category-5 TCs (Lin 2014). Since then, more cases have been observed, e.g., Meranti (165kts), Hagibis (160kts), Patricia (185kts), and more (Lin  2017, Huang  2017; Rogers  2017, Lin  2021, Wehner and Kossin 2023, Lin MS in preparation 2025). In this presentation, we explore the characteristics of these important category ‘6’ TCs. We ask scientific questions including their rapid intensification characteristics (Lin  2021). Also, because ocean is the energy source for TCs, what type of ocean condition (sea surface temperature, subsurface temperature structure and heat content, salinity) can support such extra-ordinary TCs. Finally, we also discuss whether these Category ‘6’ TCs are purely originated from global warming or there is possible influence from natural climate variability.  Reference:Lin, I., R,F. Rogers et al., A Tale of Two Rapidly-Intensifying Supertyphoons: Hagibis (2019) and Haiyan (2013), BAMS, 2021. Lin, I., I.Pun et al., ‘Category-6’ Supertyphoon Haiyan in Global Warming Hiatus: Contribution from Subsurface Ocean Warming, GRL, 2014 

Tropical cyclones (TCs) have a huge impact on land when its landfall. TC was generated when environmental conditions have horizontal wind shear and/or convergence of winds and so on. In this study, I focus on low latitude TC and review the strong low latitude TC observed during the 20^th century. The average of annual TC genesis number in the south of 10 N is around 5, on the other hand, total TC genesis number over the western north Pacific is around 27. The strongest low latitude TC at south of 7N since 1951 is TC Bopha of 935hPa in December 2012. Satellite measurements for TC started over the western north Pacific in 1977. On the other hand, aircraft reconnaissance measurements were conducted since 1945. The coverage of TC measurements was not sufficient by aircraft reconnaissance, but direct observation is able to obtain the TC intensity. 940hPa was observed at south of 7N by TC Ophelia in 1958 and TC Kate in 1970. Surface observation at Sonsorol Island in Palau at 5N 132E near TC Kate measured 1005hPa and at the same time aircraft reconnaissance measured 959hPa. Data rescue of TCs were conducted before 1945, and low latitude strong TCs were found in the early 20^th century. The quality of measurements was not sufficient, but the evidence of the TC records is precious. In March 1907, schooner Ponape anchored at Woleai Atoll in Micronesia at 7N 143E measured 922hPa. Widespread damages over the Micronesia Islands were recorded associated with this TC. In April 1905, Seeadler anchored at Pohnpei in Micronesia at 6N 158E measured 951hPa. The frequency of low latitude strong TCs are rare, however once the TC center hit, huge damage occurred due to very steep pressure gradient.

In recent decades, the intensity and intensification rate of tropical cyclones (TCs) in the North Indian Ocean, particularly in the Arabian Sea, have significantly increased. Using the high-resolution (HighResMIP) CMIP6 multi-model ensemble we found that in the near future (2018–2050) under the global warming scenario, the tropical cyclone intensity will continue to increase in the Arabian Sea during the pre-monsoon season (April–June). Also, the TC rapid intensification rate is projected to increase by 25% in the Arabian Sea during the pre-monsoon season. In contrast to the Arabian Sea, the multi-model ensemble is projecting a decrease in TC intensity in the Bay of Bengal during the pre-monsoon season. On the other hand, no major change is projected during the post-monsoon season (October–December). However, in the Bay of Bengal too the TC rapid intensification rate is projected to increase by 5% during the post-monsoon season (October–December). The projected increase in the intensification rate in the Arabian Sea during the pre-monsoon season and in the Bay of Bengal during the post-monsoon season is linked with the changes in the TC duration and ocean-atmosphere dynamic-thermodynamic conditions. Globally, the CMIP models are projecting an increase in the intensity of TCs, however, our results indicate that with global warming the intensity and rapid intensification rate of TCs in the north Indian Ocean display complex features strongly suggesting careful monitoring of the future changes in the TC activity over this basin.

Heavy rainfall is a defining characteristic of tropical cyclones (TCs) and a significant contributor to the disasters they cause. Understanding how TC rainfall responds to climate change is critical, yet studies based on observational data remain limited compared to those relying on climate model simulations. Here, using high-resolution satellite observational rainfall data and numerical model results, we find that between 1999 and 2018, TC rain rates have exhibited contrasting trends in different regions. Globally, the TC rain rate increased by 8 ± 4%, primarily driven by enhanced rainfall in the outer regions due to increased atmospheric water vapor associated with rising surface temperatures. In contrast, the rain rate in the inner-core region of TCs decreased by 24 ± 3%, likely attributable to an increase in atmospheric stability. These findings provide valuable insights into the evolving climate characteristics of TC rainfall and their underlying mechanisms.

Extreme tropical cyclones (TCs) can lead to significant losses in terms of both lives and properties. In August–October of 1959, a total of five Category 5 (C5) TCs generated over the western North Pacific. The C5 TC records account for 14.6% of all TC records, ranked first in history. Further analysis indicates that both enhanced C5 TC number and increased mean duration in C5 stages are main contributors. However, it remained to be investigated whether high-resolution numerical simulations can reproduce extreme TCs in 1959, as well as the physical mechanisms. Utilizing the Nonhydrostatic Icosahedral Atmospheric Model (NICAM) with 28-km horizontal resolution and 78 vertical levels, two 3-month experiments, each with 5 ensemble members, are designed: (1) the HIST experiment with 1951-1980 climatology sea surface temperature (SST); (2) the REAL experiment with 1959 SST. The results demonstrate that NICAM successfully reproduces extreme TCs in 1959. TC records with a minimum central pressure < 950 hPa account for 12.6% of all TC records, significantly exceeding 7.1% in the HIST experiment. On the other hand, accumulated cyclone energy, extreme TC number, and mean duration per extreme TC in the REAL experiment are all greater than those in the HIST experiment. The intensified TCs might result from genesis positions and environmental factors. In the REAL experiment, TCs tend to generate over the southeastern part of the basin, showing a southeastward shift compared to the HIST experiment. The shift can be explained by the genesis potential index and dynamic genesis potential index. The enhanced TC intensity may be partly due to the eastward TC genesis positions, which may lead to increased intensification time and lifetime maximum intensity. Meanwhile, the increase in low-level vorticity and the decrease in vertical wind shear in the main developing regions of TCs may also play a role.

Nitrogen-containing organics (NCO) in aerosols play a crucial role due to their light absorption and biotoxic properties, thereby impacting regional haze, health and climate. This study investigated the diurnal variation, chemical characteristics, and potential formation pathways of  (NCO)  in ambient aerosol, frost and cloud water in Megacities and rural regions in China by UHPLC-Q-TOF-MS., UHPLC-Orbitrap MS.  The results showed that NCO accounted for over 60% of urban organic aerosols, with O/N < 3 compounds being the major contributors (>70%). The predominance of ESI+ suggested the prevalence of reduced NCO. These NCO included alkyl, cyclic, and aromatic amides in CHON compounds, which were associated with anthropogenic activities and may be capable of forming light-absorbing chromophores or posing harm to human health. Nitroaromatic compounds were more prevalent in winter than in summer. Further analysis of atmospheric formation and precursor-product pairs suggested that CHON are derived from the oxidization or hydrolyzation processes, revealing potential transformations of these aerosols.Both solution and particles analyses in frost collected in rural Northeast China under non-hazy and hazy days were conducted. The average NCO molecular formulas were1934, 3114, 2224, and 3042 in ESI-, and 5954, 6921, 27 5547, and 6629 in ESI+, for the non-hazy solution, hazy solution, non-hazy particle, and hazy-particle samples, respectively. Nitrophenol (C6H5NO3) and methyl nitrophenol (C7H7NO4) were the most abundant NCO. Cloud water samples collected at the submit of Mt. Shanghuang in 2023, 1320 soluble organics in ESI+, and 870 in ESI- were identified. CHON dominated the molecular composition, accounting for 52.1% in ESI+ and 47.2% in ESI-. Nitrobenzene compounds predominated in the ESI-, while amine chain amides were dominant in the ESI+. Precursor matching analysis revealed that most of the chain amides detected were identified through nucleophilic addition reactions among carbonyl groups and ammonia.

Nitrogen-containing organic aerosol, consisting of nitroaromatics, N-heterocyclic compounds, and organic nitrates, are a group of key species in organic aerosol affecting aerosol optical properties and nitrogen cycle. The formation and aging of nitrogen-containing organic aerosol, however, are still not well understood hindering our understanding on their atmospheric evolution and impacts. In this study, nitrate-mediated photooxidation of some important nitroaromatics in atmospheric aqueous phase under different pH and temperature conditions were investigated. The dynamic changes in light absorption of nitroaromatics were measured, the photolysis rates and oxidation products as well as the aging processes were characterized. Besides nitroaromatics, the nighttime formation processes of secondary organic nitrates were investigated based on size-resolved measurements with a soot particle long-time-of-flight aerosol mass spectrometer. It was found that aqueous processing played an important role in the nighttime formation of particulate secondary organic nitrates in large size particles, especially in fog-rain days. In addition, N-heterocyclic compounds from aqueous-phase reaction of dicarbonyls with amines and ammonium under different pH were also studied. We identified for the first time 155 new N-heterocyclic compounds and their formation pathways were characterized.

Nitro-phenolic compounds (NPs) are key constituents of brown carbon, contributing to visibility degradation, climate forcing, and atmospheric oxidation processes. Despite increasing research, the abundances, sources, and atmospheric fate of NPs remain poorly understood. In this study, we conducted continuous measurements of eighteen gaseous NPs using high-resolution Time-of-Flight Chemical Ionization Mass Spectrometry (ToF-CIMS) at a background site in South China. Our observations revealed substantial NP from secondary formation in continental outflows, with mono-NPs exhibiting daytime photochemical peaks and di-NPs enriched at night. Budget analysis with a photochemical box model highlighted that, in addition to OH oxidation of aromatics, NO₃ oxidation significantly contributed to daytime mono-NP formation and nighttime di-NP production.  Concurrent gas and particle-phase measurements revealed a wide range of NPs in both phases. The partitioning of NPs exhibited significant diurnal variations, with particulate fractions ranging from 8.6% to 53%, deviating substantially from theoretical estimates, emphasizing the need for improved parameterization in partitioning models. A field-based apparment parameter was derived from the observation data, and can significantly enhanced the accuracy of NP gas-particle partitioning estimates. This study provides quantitative insights into NP formation, gas-particle partitioning dynamics, and atmospheric impacts, highlighting the need for ruther investigation.

Nitro-aromatic compounds (NACs) are important atmospheric pollutants that impact air quality, atmospheric chemistry, and human health. Understanding the relationship between NACs formation and key environmental driving factors are crucial for mitigating their environmental and health impacts. In this work, we combined an ensemble machine learning (EML) model with the SHapley Additive exPlanation (SHAP) and positive matrix factorization (PMF) model to identify the key driving factors for ambient particulate NACs covering primary emissions, secondary formation, and meteorological conditions based on field observations at urban, rural, and mountain sites in eastern China. The EML model effectively reproduced ambient NACs and recognized that anthropogenic emissions (i.e., coal combustion, traffic emission, and biomass burning) were the most important driving factors, with the total contribution of 49.3%, while significant influences from meteorology (27.4%), and secondary formation (23.3%) were also confirmed. Seasonal variations analysis showed that direct emissions presented positive responses to NACs concentrations in spring, summer, and autumn, while temperature had the largest impact in winter. By evaluating NACs formation and loss under various locations in winter, we found that anthropogenic sources played a dominant role in increasing NACs levels in urban and rural sites, while reduced ambient temperature along with secondary formation from gas-phase oxidation was the main reason for relatively high particulate NACs levels at the mountain site. This work provides a reliable modelling method for understanding the dominant sources and influencing factors for atmospheric NACs and highlights the necessity of strengthening emission sources controls to mitigate organic aerosol pollution.

The study of mono- and polycyclic- compounds containing nitro and hydroxyl groups in atmospheric particulate matter holds significant importance in environmental health and atmospheric chemistry. In this research, a self-constructed VACES-OE-LC-MS integrated system was utilized to online detect various mono- and polycyclic compounds containing nitro and hydroxyl groups in the Shanghai area. The concentrations of these compounds in atmospheric particulate matter ranged from several picograms to several nanograms per cubic meter, with the total concentrations of mono- and polycyclic- compounds being on the same order of magnitude. On clear days, the concentrations of the aforementioned compounds exhibited a dual-peak pattern at noon and in the evening. However, on cloudy/rainy days, the overall concentration of polycyclic compounds decreased by approximately half compared to mono-cyclic compounds, with the noon peak delayed and the evening peak disappearing. The high noon concentrations were driven by intense solar radiation and OH radical reactions, while the evening peak was primarily driven by NO₃ radical reactions and NO_x accumulation. The delayed noon peak and the disappearance of the evening peak on cloudy days probably related to the weakening of daytime photochemical reactions and the suppression of nighttime NO₃ radical reactions due to factors such as wet deposition of NO_x and gaseous precursors.

Heterogeneous NO₂ hydrolysis forms nitrate and nitrous acid but is believed to proceed slowly in the atmosphere. Accelerated reactions in microdroplets have gained significant attention, but whether NO₂ hydrolysis is accelerated in deliquesced particles remains unclear. We address the gap by measuring size-dependent NO₂ hydrolysis rates of sulfate or halide-containing droplets. Results show that the reaction rates in Na₂SO₄ droplets increased by a factor of ~25 as the particle radius decreased from ~40 to ~4 µm. An even higher enhancement of ~100 times was observed in NaCl and NaI particles, likely due to the NO₂-halide interactions. The enhancement observed in NaBr particles, however, was comparable to that observed in Na₂SO₄ particles. Kinetic modeling results illustrate that the accelerated reactions are due to ~6 orders of enhancement factor (EF) of surface reaction rates over the bulk-phase reaction rates. Compared to Na₂SO₄ particles, the surface reaction rates increase by factors of 2.3, 1.5, and 4.4 in NaCl, NaBr, and NaI particles, respectively. Under ambient conditions, EFs can increase up to 10⁸, corresponding to the ambient nitrate production rates of >1 µg m^–3 h^–1. The rates are comparable to the N₂O₅ hydrolysis and OH+NO₂ reaction pathways, making NO₂ hydrolysis a crucial source of reactive nitrogen species.

Air quality in Northeast Asia has significantly improved due to various policies. However, long-range transport and secondary formation continue to contribute to persistent air pollution issues. In this study, the Community Multiscale Air Quality (CMAQ) model was used to simulate air quality in Northeast Asia for the years 2020 and 2050. The Tagged Species Method was applied to analyze regional contributions. To project future air quality, outputs from the CMIP6 were used as the initial and boundary conditions for the WRF meteorological model, providing a foundation for future meteorological simulations. Additionally, future emissions were estimated using the Air Quality for Northeast Asia (AQNEA) scenario, which integrates three energy scenarios and two air quality policy scenarios. An analysis of PM₂.₅ concentrations in 2020 revealed that major cities in China exhibited the highest contributions from both local emissions and emissions from nearby cities. In contrast, Seoul, located downwind, showed a higher contribution from transboundary sources. Although Tokyo had the highest local contribution, its overall pollution levels remained lower than those of other major cities. Future emission scenarios based on AQNEA indicate that incorporating the NetZero energy scenario would lead to substantial emission reductions and improved air quality. Conversely, in developing regions such as North Korea and Mongolia, emissions are projected to increase under the baseline scenario. This study underscores the necessity for region-specific air quality management policies that consider both local characteristics and long-range transport effects.

Particulate matter (PM) toxicity is governed by both particle size and chemical composition. For comprehensive environmental and health impact assessments, concurrent analysis of source-specific particle size distributions (PSDs) and chemical compositions is essential. Such analysis would ideally be conducted in real time. While real-time analysis facilitates prompt and effective air quality management strategies, the main challenge lies in the rapid and precise utilization of high-time-resolution data for accurate assessment of temporal variations in source contributions.To address these challenges, this study developed an automated real-time source apportionment system through an integrated offline-online framework. The offline phase uses PMF to analyze historical PSD and chemical composition data, establishing stable source profiles that serve as reliable reference patterns. These well-characterized source profiles are crucial for accurate source identification and are then used as constraints in the online phase, enabling more stable and computationally efficient source apportionment. Our framework integrates offline-derived constraints with continuous measurements and HYSPLIT trajectory modeling. This integrated approach enables robust real-time analysis while maintaining computational efficiency, thereby overcoming the instability and delays typically associated with unconstrained and long-term source apportionment methods.Field validation at our monitoring station demonstrated the system's capabilities through several key performance indicators. The framework demonstrated consistent performance by identifying traffic-related sources contributing to 80% of total particle number concentration and 19% of PM_1.0 mass concentration, aligning with the established long-term averages. The successful integration of offline constraints with online analysis verified the framework's capability for real-time source contribution estimation. Spatial distribution further showed that the short-term source spatial patterns were comparable to long-term average distributions. Validation tests confirmed the system's reliability in distinguishing temporal variations and spatial distribution patterns, demonstrating its potential for comprehensive source analysis. These validation results establish the framework's effectiveness as a scientific foundation for real-time air monitoring and regulation protocols.

The aerosol optical depth (AOD) is a critical parameter in atmospheric science, providing essential insights into the concentration and distribution of aerosols in the Earth's atmosphere. However, accurate assessment of AOD remains challenging due to its inherent spatial variability, even over relatively small geographic areas. This study aims to explore the spatial variability of AOD within the Kao-Ping Experiment (KPEx) as part of the 7-SEAS (Seven SouthEast Asian Studies) 2024 international field campaign in southern Taiwan. The campaign was divided into four campaigns and lasted from February to March 2024. During KPEx, the dense network of 5 AERONET sites, 2 MPLNETS Lidar stations, and more than 20 air quality monitoring stations plays a crucial role.  This result reveals that the AOD retrieved from the AERONET network shows a consistent relationship with the satellite MODIS AOD with a correlation coefficient (r) reaching 0.71. There are significant spatial and temporal variations in AOD across different AERONET stations, with urban areas typically showing higher AOD and island or higher altitude locations displaying distinct patterns. A significant variation in the relationship between AOD and PM_2.5 among the sites was also observed with the highest value being 0.68. Urban areas typically show stronger correlations between AOD and particulates due to higher pollution sources, whereas meteorological impacts vary across different locales, affecting aerosol dispersion and concentration. Besides, taking the advance of dense MPLNET and AERONET networks, the machine learning method is also applied using the KPEx data to estimate the vertical profile of PM_2.5 within the Southern Taiwan area. In conclusion, this study wants to reveal the comprehensive properties of aerosol and air pollution patterns in Southern Taiwan. The relationship between aerosol properties and PM_2.5 is also studied to fill up the vertical profile of PM_2.5 in the area.

With the continuous advancements in artificial intelligence and computational technology, AI-based models introduced new possibilities for weather forecasting. Leading tech companies such as Google, Huawei, and Microsoft, along with research institutions, have launched atmospheric models based on large-scale AI, initiating a new “data-driven” paradigm in meteorological research. Despite the capabilities of some existing models in atmospheric environment forecasting, these models still require high-resolution gridded data and underutilize ground-level observational data. Additionally, the relatively low resolution of available gridded data limits the ability to forecast at the urban scale and prevents true “end-to-end” forecasting. In response to these limitations, we present a novel AI-based atmospheric forecasting model, Fuyao. This model innovatively incorporates a “non-uniform grid” design, maintaining the strengths of large atmospheric models while enhancing the utilization of ground-level observational data for air pollutants. Fuyao effectively forecasts various atmospheric elements, including ground-level particulate matter and gases. In specific experiments, Fuyao has demonstrated superior performance over the CAMS model in forecasting PM2.5, PM10, O3, CO, SO2 and NO2 across the Beijing-Tianjin-Hebei region, with higher correlation coefficients (R) and lower root mean square errors (RMSE). Furthermore, during heavy ozone and PM2.5 pollution episodes, Fuyao outperforms traditional models like WRF-Chem. The Fuyao model can accurately forecast air quality in different cities and regions, solving the problems of accuracy, efficiency and operational cost in air quality forecasting. 

Coastal cities with various emission sources, such as Busan in Korea, a coastal metropolis where a major port sits, experiences distinct local air mass recirculation due to land-sea breezes. These breezes cause the air mass to move out to the ocean during the land breeze and return inland during the sea breeze, cycling between ocean and land areas rich in water vapor and precursor gases, respectively. This local recirculation can alter fine particles' physical and chemical properties. To investigate these physicochemical changes in fine particles during land-sea breeze periods, we conducted measurements of PM₂.₅ composition and size distributions, and meteorological factors in Busan across multiple seasons from 2021 to 2022. Initially, we developed an algorithm to extract land-sea breeze periods during measurement periods, based on meteorological parameters. During the extracted land-sea breeze periods, we observed several common features, including an influx of water vapor during sea breezes that led to increased specific humidity. As the circulation transitioned to the land breezes, relative humidity increased due to lowered temperature, followed by a rise in aerosol liquid water content (ALWC). Concurrently, the concentrations of inorganic ions and PM₂.₅ increased gradually. These findings highlight the significant role of ALWC in PM₂.₅ formation under land-sea breeze conditions, providing insights into the impact of local meteorology on air quality in coastal cities.

Under China's energy policies, anthropogenic emissions have gradually decreased, stabilizing and slowly reducing nitrogen wet deposition. To assess the effects of these policies and the COVID-19 pandemic, rainwater samples were collected from Matsu (a coastal area with high anthropogenic influence, n = 99) and Pengjiayu (a marine area with low anthropogenic influence, n = 97) between April 2020 and April 2024. In terms of ion composition, total dissolved nitrogen (TDN) and dissolved inorganic nitrogen (DIN) showed a summer-low, winter-high trend, while dissolved organic nitrogen (DON) exhibited no clear seasonal variation. By comparing the NH₄⁺ to non-sea-salt sulfate (nss-SO₄^2-) ratio, the NH₄⁺/nss-SO₄^2- ratio was higher in the coastal region (1.07) than in the marine region (0.85), indicating a mixture of anthropogenic ((NH₄)₂SO₄) and marine (NH₄HSO₄) sources in the coastal area, while marine sources (NH₄HSO₄) dominated in the marine area. In terms of flux, DIN was nearly twice as high in the coastal region (30.9 mmol m⁻² yr⁻¹) compared to the marine region (17.8 mmol m⁻² yr⁻¹), while DON flux was similar in both regions (0.47 vs. 0.34 mmol m⁻² yr⁻¹). Regarding the contribution of DON to TDN flux, the coastal region contributed 1.69%–34.0% of TDN flux, while the marine region contributed 0.12%–60.1%, highlighting the significant role DON plays in areas with lower anthropogenic influence. Moreover, the reduction in DIN flux showed a decreasing trend of -7.1 mmol m⁻² yr⁻¹ in the coastal region and -4.2 mmol m⁻² yr⁻¹ in the marine region, reflecting significant reductions in anthropogenic emissions due to China's energy policies and the COVID-19 pandemic.

As one of the most densely populated cities worldwide, air pollution issues over megacity Delhi, the capital of India, have caused great concern. Diwali Festival is one of the most important Indian festivals celebrated nationwide by displaying extensive fireworks. Intense anthropogenic emissions from short-term Diwali firework displays coinciding with open biomass burning during the post-monsoon seasons further exacerbated the air quality in Delhi. Many studies have well documented the impacts of Diwali firework displays on the PM mass and chemical composition, trace metals, particle number and size distribution, and real-time evolution of firework-derived chemical species. In the literature, there is no study to understand the morphology, size, composition, and mixing state of individual particles specifically related to fireworks in India. This gap is pivotal for identifying particle sources, understanding particles evolution process, and comprehending potential health effects during the Diwali Festival. The highly polluted scenario during the Diwali Festival provides a unique opportunity to study fine particles originating from firework displays and biomass burning, and possibly identifying the distinct tracers for firework emissions during the Diwali celebrations in Delhi. In this study, we investigate the variations in chemical composition, morphology, mixing state, size, and number fraction of individual aerosol particles before, during, and after the Diwali Festival in Delhi using transmission electron microscopy (TEM) coupled with energy-dispersive X-ray spectrometry (EDS). To our knowledge, this is the first study aimed at identifying the individual particle types related to firework displays and exploring the sources and evolution process of aerosol particles at the microscopic scale during the Diwali Festival.

The complex structures and diverse emission sources in urban coastal environments add complexity to understanding the spatiotemporal distributions of air pollutants, while mesoscale atmospheric circulation can influence air quality by affecting the chemical processes of air pollutants. In this study, we conducted highly spatially resolved multi-point measurements of air pollutant concentrations (CO, NO, NO₂, O₃, PM₁₀, and PM_2.5) using cost-effective air quality sensors in Ulsan, a developed urban coastal city in South Korea. Ulsan features extensive industrial areas along the coast, with residential areas located inland and surrounded by forests. Sensors were deployed in various emission environments, including industrial, residential, suburban, and urban background sites, and field measurements were conducted for two weeks in both summer and winter from 2023 to 2025. We found that air pollutant concentrations and their diurnal patterns varied depending on emission environments. Notably, O₃ concentrations decreased as NOx level increased, with daytime NOx concentration peaks driven by combustion activities in traffic and industrial areas. Furthermore, we examined the effects of sea breeze on surface O₃ concentrations and the associated chemical processes. As the sea breeze transported the O₃-laden air masses inland, O₃ levels gradually decreased in industrial and residential areas due to NO titration but increased again in forested regions further inland. Based on these O₃ concentration gradients, we evaluated the impact of sea breeze on O₃ concentrations by quantifying O₃ advection rates and further explored approaches for assessing the chemical budget of O₃ concentrations. We expect this study to enhance the understanding of coastal urban air pollution, as well as the effects of sea-land breezes on chemical processes of surface O₃.

Dhanbad, also known as the “Coal Capital of India,” has been subjected to air pollution for years. Dhanbad has no public bus facilities for intra-city commuting, and the city heavily relies on 3-wheeler (3W) autos. These autos mainly run on diesel fuel, producing a lot of exhaust emissions, which adds to the already existing air pollution. To manage the worsening on-road air quality issues, it is necessary to increase the share of electric vehicles (EVs), especially the 3W autos. EVs have no exhaust emissions and thus can help combat local air pollution. This research work focuses on identifying the key factors influencing the purchase and adoption of EVs. Despite the prevalence of subsidy schemes, the market share of 3W EVs is significantly less. To learn more about the reason behind this, a word-of-mouth (WOM) roadside survey was conducted in Dhanbad from 3W auto drivers’ point of view. Several factors for lesser market penetration were found. Income from running the vehicles was the most important factor, followed by vehicle operating speed, charging time, etc, of the 3W EVs. This study provides a detailed analysis of these barriers using various statistical techniques. Additionally, an Analytical Hierarchical Process (AHP) model predicted that the services offered by 3W EVs make them the best option available in the city to buy. With the projected market penetration of 3W EVs, the CO2 emission in the duration of 2022-2030 is expected to be reduced by 36,564 tons, this includes scope-2 grid emissions. It was observed that under the optimistic scenario, the city can avoid 26 tons of PM emissions from the exhaust during this period. A detailed study focusing on the life cycle assessment of 3W EVs is mandatory to understand its overall environmental benefits.

The ambient PM_2.5 pollution in subtropical Asia (SA) has surpassed the World Health Organization's recommended value, threatening human health, and ecosystems. Meteorological condition is a key factor in driving variations of PM_2.5-related air pollution, and the links between air pollution and synoptic situation have been widely studied. However, it is not fully clear how climate affects PM_2.5 variability on an interdecadal timescale, which undermines our ability to predict interdecadal variations of PM_2.5 in the SA. Here, we show that the thermal forcing of the Tibetan Plateau (TP) plays a significant role in exacerbating PM_2.5 pollution in the SA during winter. The TP thermal forcing strengthens the East Asian subtropical jet and descending airflow, and thus exacerbates air pollution by reducing precipitation, lowering planetary boundary layer height, and enhancing atmospheric stability. The interactions among the TP thermal forcing, atmospheric circulation, and PM_2.5 concentrations form a feedback loop that substantially increases air pollution in SA. The pollution exacerbated by the TP will further impact human health in China and India, and will have serious effects over the next 30 years. Our findings corroborate the dominant role of thermal forcing by the TP on poor air quality in SA.

Secondary organic aerosol (SOA) is an important component of atmospheric fine particles and adversely affects human health, air quality and climate change. Isoprene is the most significant biogenic non-methane volatile organic compound (VOC) emitted into the earth's atmosphere. Ammonia (NH3) is the most critical alkaline gas in the atmosphere, affecting ecological cycles and human health, and mainly comes from agriculture, industrial processes and vehicular emissions. Isoprene and NH3 are essential precursors for organic and inorganic aerosol in the atmosphere, but their molecular interactions are unclear. Stabilized Criegee intermediates (SCIs) are essential radicals produced by ozonolysis of alkenes, and have been confirmed to play a vital role in SOA formation. The fundamental role of NH3 in SOA formation from ozonolysis of isoprene was investigated based on smog chamber experiments. The molecular information was determined based on high-resolution orbitrap mass spectrometry. Results show that NH3 can react with Criegee intermediates to generate amines, which significantly reduces the yield of SOA from isoprene ozonolysis. This study discovered a new pathway for NH3 to change SOA composition through a novel reaction with Criegee intermediates. The fundamental mechanism of NH3 in SCI-derived SOA formation was revealed at the molecular level. The roles of NH3 in biogenic SOA are more complex than their representation in atmospheric chemistry models and need to be reevaluated as the chemistry between NH3 and isoprene was revealed.

The National Science Foundation (NSF) of the United States approved the Airborne Phased Array Radar (APAR) Mid-scale Research Infrastructure-2 proposal in 2023 to develop the next generation airborne polarimetric, Doppler weather radar mounted on the NSF/National Center for Atmospheric Research (NCAR) C-130 aircraft. Polarimetric measurements are not available from current airborne tail Doppler radars. The APAR system will consist of four removable C-band active electronically scanned arrays (AESA) strategically placed on the fuselage of the aircraft. Each AESA measures approximately 1.5 x 1.5 m and is composed of 2368 active radiating elements arranged in a total of 37 line replaceable units (LRU). Each LRU is composed of 64 radiating elements that are the building block of the APAR system. The development will be completed in summer 2028. APAR adopts a phased approach as an active risk assessment and mitigation strategy. The APAR Team includes partners in industry and in the university community. One of the outstanding challenges in observational meteorology is to observe simultaneous kinematic and microphysical information that will lead to a better understanding and description of the heating profiles within convective storms via phase changes of water. APAR will fill this critical gap especially in those weather systems originate over the ocean, e.g., tropical cyclones, maritime MCSs, explosive cyclones, etc., that have impacts from mesoscale to climate. The authors will update current progress and accomplishments of APAR during the first year of the development, including the activities of the university and private industry partners and the engagement of collaboration with research and operational communities.

The National Science Foundation (NSF) of the United States approved the Airborne Phased Array Radar (APAR) Mid-scale Research Infrastructure-2 proposal in 2023 to develop the next generation airborne polarimetric, Doppler weather radar mounted on the NSF/National Center for Atmospheric Research (NCAR) C-130 aircraft.  The APAR Observing Simulation, Processing, and Research Environment (AOSPRE) was developed to simulate APAR's measurement capabilities for heavy precipitation and high-impact weather events. Using Cloud Model 1 (CM1) and Weather Research and Forecasting (WRF) model output to provide various storms of interest and their surrounding environments, simulated NCAR C-130 flights are operated within the model space.  AOSPRE is linked to a NSF NCAR wide INtegrating Field Observations and Research Models (INFORM) to (1) establish and support best practices and methods for comparisons between models and observations, (2) exploit, assess and quantify the impacts of integrating observations and models to improve understanding of the prediction and predictability of the Earth system, and (3) improve the design, planning, deployment strategy of field programs and instrument development. The AOSPRE will be expanded into a field program planning tools as wells as a post campaign re-analysis tool with DA capability. AOSPRE is developed as an open-source software. The first version of AOSPRE software has been released to the research and operational community in the last quarter of 2024. This paper will provide an overview of the AOSPRE and report the recent development of the AOS to better simulate the characteristics of a phased array radar on a moving platform. In addition, the authors will outline how AOSPRE will be used as a component in the future APAR data analysis software system.

The indirect radar reflectivity assimilation method, which assimilates the retrieved hydrometeor from radar reflectivity data, is simple and efficient in severe weather forecasting applications. However, it suffers from retrieval errors due to the uncertainties in discerning multiple hydrometeors based solely on reflectivity observations. To mitigate these inaccuracies, the dual-polarization radar data is incorporated into the background-dependent indirect reflectivity assimilation method in this study. First, the contribution of multiple hydrometeor species to the whole reflectivity is first estimated using the observed reflectivity and background microphysical information, then the hydrometeor classification algorithm (HCA) product from dual-polarization radar observations is introduced to correct the dominant hydrometeor type if in error, and last the contribution factors are adjusted and used to retrieve multiple hydrometeor species from reflectivity data. Through a single squall-line case, it is demonstrated that the incorporation of the HCA product from dual-polarization radar data leads to more reasonable hydrometeor identification, with more supercooled rainwater above melting layer and more graupel at low levels, thereby refining the hydrometeor analysis. With the 15-minute rapid update cycling configuration, the changes in analysis field enable more cold rain processes, resulting in more intense latent heat release at higher levels and stronger cooling near the surface in the forecast. This in turn strengthens updraft motions and cold pools in the convective regions, thereby improving the reflectivity and precipitation forecasts. Four cases’ quantitative evaluations of the 0-3-h reflectivity and precipitation forecasts further validate the effectiveness of incorporating dual-polarization radar data in the assimilation process.

As the main results of collision in liquid precipitation, coalescence, breakup and their balance can be well captured and presented from polarimetric radar perspectives. Based on S-band polarimetric radar measurements during the record-breaking rainfall in Zhengzhou, China, on 20 July 2021, the time-series, vertical profile, spectra, and spatial variation of polarimetric radar measurements are thoroughly investigated to show different perspectives of microphysical processes in the rainstorm. The results show that: (i) The strongly connected views from different angles agree with each other and enrich the understanding of microphysical processes. (ii) Excluding quality issues of radar measurements, the transitions between different microphysical processes, are the sources of uncertainties of radar-measured reflectivity (Z_H) and differential reflectivity (Z_DR), and they can be clearly identified and differentiated, which also accounts for a theory-realty contradiction between measured and retrieved Z_H and Z_DR. (iii) Integrating microphysical information, i.e, the averaged relationships between radar variables, into radar quantitative precipitation estimation (QPE) algorithms could effectively mitigate the QPE uncertainties caused by microphysical transitions.

For quantitative precipitation estimation (QPE) based on polarimetric radar (PR) and rain gauges (RGs), the quality of the radar data is crucial for estimation accuracy. A combined radar quality index (CRQI) is proposed to represent the quality of the radar data used for QPE and an algorithm that uses CRQI to improve the QPE performance. Nine heavy rainfall events that occurred in Guangdong Province, China, were used to evaluate the QPE performance in five contrast tests. The QPE performance was evaluated in terms of the overall statistics, spatial distribution, near real-time statistics, and microphysics. CRQI was used to identify good-quality data pairs (i.e., PR-based QPE and RG observation) for correcting estimators (i.e., relationships between the rainfall rate and the PR parameters) in real-time. The PR-based QPE performance was improved because estimators were corrected according to variations in the drop size distribution, especially for data corresponding to 1.1 mm < average D_m < 1.4 mm, and 4 < average log₁₀ N_w < 4.5. Some underestimations caused by the beam broadening effect, excessive beam height, and partial beam blockages, which could not be mitigated by traditional algorithms, were significantly mitigated by the proposed algorithm using CRQI. The proposed algorithm reduced the root mean square error by 17.5% for all heavy rainfall events, which included three precipitation types: convective precipitation (very heavy rainfall), squall line (huge raindrops), and stratocumulus precipitation (small but dense raindrops). Although the best QPE performance was observed for stratocumulus precipitation, the biggest improvement in performance with the proposed algorithm was observed for the squall line.

Dust aerosols can transport iron from land to ocean through atmospheric deposition. Insoluble ferric iron within dust particles may also be reduced to ferrous iron through atmospheric chemical reactions. This study examines the trends and driving factors of soluble dust iron deposition over East Asia using an enhanced version of the Community Earth System Model (CESM). We improved the model to incorporate desert dust mineralogy and atmospheric chemical processes that promote iron dissolution, enabling a detailed analysis of dust iron evolution. Our findings emphasize the critical roles of both dust emission and atmospheric processing in enhancing iron dissolution, which in turn affects soluble iron deposition and marine ecology.

Sand and dust storms (SDS) imposed imminent risks to human health, including increased risks on asthma, heart attack, and premature death. Regardless of recent advances in dust modeling and observation, the long-term trend of global dust remains highly uncertain, hindering efforts to assess and mitigate dust impacts. This study presents an international collaboration, coordinated under the World Meteorological Organization (WMO) Sand and Dust Storms Warning Advisory and Assessment System (SDS-WAS), to develop long-term climatology of global dust life cycle. In the first phase of the project, long-term trends of dust surface concentration and column dust loading are reconstructed using four global dust reanalysis datasets developed by the National Aeronautics and Space Agency (NASA), Naval Research Laboratory (NRL), European Centre for Medium-Range Weather Forecasts (ECMWF), and Finnish Meteorological Institute (FMI). Through intercomparisons of these four global reanalysis datasets, with ground and satellite observations, the strengths and limitations of each dataset are thoroughly assessed. Results from the ensemble reanalysis showed that the number of days with daily dust PM₁₀ exceeding 45 ug/m³ (the WHO guidance) ranged from zero to a few days in areas not affected by dust to over 1,600 days in the dustiest regions during the five-year periods. The number of days of population exposure to high dust levels increased in 48% of countries while decreasing in 20% of countries from 2003-2007 to 2018-2022. Among the countries with increased dust exposure, two-thirds of them fall into high- or very high-income categories. Despite of the decrease in dust exposure in North Africa, the dust exposure levels of these countries remain the highest in the world, imposing a substantial burden to human health.  

Saharan dust constitutes approximately 50-60% of the total global dust and can impact regional climate, environment, and ecosystems through both direct and indirect effects. However, the long-range transport of Saharan dust to East Asia and its specific effects on the region's weather and climate remain poorly understood. By combining satellite observations with model simulations, multiple reanalysis data, and HYSPLIT trajectory analysis, we systematically investigate the long-range transport of Saharan dust to East Asia and further examine its impact on direct radiative forcing, clouds, and precipitation in the region. A quarter of dust cases in East Asia originate from the Sahara Desert. The long-range transported Saharan dust is typically located in the upper troposphere of East Asia. The total annual average amount of Saharan dust transported over East Asia is 33.05 ± 9.78 Tg/year. Saharan dust can be transported eastward throughout the year and contributes approximately 35.8% of the dust to the upper troposphere in northern China in spring, which is nearly equivalent to the amount of dust lifted from the East Asian dust source. Furthermore, regarding regional impact, Saharan dust exerts a cooling effect on both the surface and the upper atmosphere, while simultaneously causing a warming effect within the atmosphere. Dust from the Sahara and Middle East, carried by the jet stream, contributes to 47.7% of high-altitude dust in the Taklimakan Desert before rain. It increases ice water content and particle radius by 59.6% and 72.4%, boosting precipitation by 57.6%. This study provides new insights into Saharan dust's role over East Asia, improving understanding of its transport and atmospheric sources, and clarifying its impact on radiative forcing, clouds, and precipitation.

In May 2022, Iran experienced extreme dust activities, with the highest frequency of dust episodes and dusty days in the past seven decades. This study explores the drivers behind this event, focusing on land surface conditions and meteorological factors, with meteorological station data, satellite observations, and reanalysis datasets. The results show that May 2022 was a record-breaking month for dust activity in Iran over 77 years, with a mean of 14.8 dusty days, far exceeding the climatological mean of 5.2 days. Dust events were more frequent and lasted longer, with a 4.7-fold increase in events lasting 24-72 hours. Low vegetation cover and soil moisture in Iran and upstream regions (Iraq, Syria, and Saudi Arabia) enhanced dust emissions. Additionally, abnormally strong surface winds and a high frequency of strong wind events intensified dust mobilization. Atmospheric circulation patterns, including a strengthened Persian low-pressure system and a high-pressure gradient over the Arabian Peninsula, facilitated dust transport into Iran. This study highlights the critical roles of land degradation and atmospheric dynamics in extreme dust events and underscores the need for further research on the impacts of climate variability and change on dust activity.

Taklimakan Desert (TD) serves as a significant source of high-altitude airborne dust over the Tibetan Plateau (TP). However, systematic understanding of its transport mechanism requires further exploration beyond isolated case studies. This study effectively identified the atmospheric circulation patterns and transport mechanisms that facilitate airborne dust movement from TD to TP in the spring, utilizing an obliquely rotated principal component analysis alongside various reanalysis and satellite datasets. The findings indicate that out of the five identified circulation patterns, three — specifically the northwest high-pressure (NWH), the northern high-pressure with warm anomaly (NH-W), and the northern high-pressure with cold anomaly (NH-C) — favor the occurrence of dust storms in the TD. In the NWH and NH-W patterns, dust is transported to the northeastern TP and the northwestern and central TP as a result of the interaction between dynamic, thermal, and terrain factors. This process features an elevated boundary layer, increased temperature, a steeper temperature lapse rate, and a more significant surface sensible heat flux in the TD. In contrast, the NH-C pattern restricts dust transport due to the presence of downdrafts along the north slope of the TP, and a reduced boundary layer and stable temperature stratification in the TD. This research provides valuable insights into the crucial function of atmospheric circulation in transporting dust from TD to TP, which is beneficial for assessing the condition of the cryosphere and developing environmental protection strategies.

East Asian desert is one of the major dust sources in the world, contributing approximately 40% of global annual dust emissions and exerting a substantial impact on regional environment and climate. Recent reports indicate that the frequency of dust storm outbreaks in East Asia has declined significantly, primarily due to improved natural conditions (i.e., reduced surface wind speeds, increased precipitation, and enhanced soil moisture) and the promotion of vegetation cover resulting from afforestation efforts in north China. However, there is limited research on how the properties of dust, following long-range transport to downstream regions, respond to this decline. From 2010 to 2024, we conducted continuous observations of height-resolved dust aerosols using a ground-based polarization lidar in Wuhan (30.5°N, 114.4°E), located in central China. The dust optical depth (DOD) showed a consistent decline at a rate of -0.011 yr^-1 until August 2020, accounting for ~22% of the total deduction in aerosol optical depth (AOD). The dust mass concentration and columnar mass density decreased by 2.03 μg·m^-3 and 1.97 mg·m^-2 per year, respectively. However, since 2021, long-range transported Asian dust has become more frequent and intense during spring, with significant higher springtime DOD values of 0.16-0.20 from 2021 to 2024, compared to 0.12 in 2020. This study provides a detailed climatology of dust properties over central China, offering valuable insights into the response of downstream dust aerosols to the reduction of dust emissions from the East Asia desert as well as to regional climate change.

The unique terrain and complex atmospheric boundary layer (ABL) processes result in a distinctive spatiotemporal distribution of dust in the Tarim Basin; however, this distribution remains unclear under continuous clear-sky conditions. In this study, 382 cases were selected to investigate the spatiotemporal evolution of dust and its potential mechanisms based on MERRA-2 and ERA5 reanalysis datasets combined with MODIS satellite observations during the warm seasons from 2000 to 2023. Taking the typical case of a completely cloudless on July 24–27, 2016, the dust aerosol optical depth (DAOD) at the margin of the Tarim Basin increased with time. The climatological characteristics showed a high DAOD in the northern, western, and southwestern regions and a relatively low DAOD in the central area. Nocturnal low-level jets dominated by northeasterly winds enhance the low-level westward airflow in weak anticyclonic systems, causing dust accumulation in the west and north of the basin. Vertical mixing within the ABL during the daytime increases dust loading in the residual layer, and these dust particles can ascend to high altitudes after breaking through the ABL by the vertical circulation. The dust loading at the lower level during the daytime was higher than that at night, whereas the opposite was true for the upper level. The downward airflow in the northwest slope of the Tibetan plateau weakens at night, leading to dust being uplifted to higher altitudes and transported outside the Tarim Basin by the westerlies. These results enhance our understanding of dust distribution and related mechanisms in Tarim Basin and support the development and utilization of climatic resources in this region.

Northeast Asia is one of the world's primary dust sources, and the frequency of dust events in this region has risen sharply in recent years, posing significant threats to the ecological environment and socio-economic systems. However, the influence of local and upwind underlying surface factors on dust intensity in Northeast Asia remains poorly understood. Investigating the variation characteristics and driving factors of dust intensity is crucial for understanding the mechanisms of dust occurrence and for developing effective prevention and control strategies. This study analyzes long-term trends in dust intensity and underlying surface factors across Northeast Asia using ground-based meteorological station data and raster datasets, including the Normalized Difference Vegetation Index (NDVI), soil moisture, and snow cover, from 2000 to 2024. The key findings are as follows: (1)Spatially, dust intensity is highest in the arid and semi-arid regions of Mongolia and northern China, with three distinct high-intensity centers identified. (2)Temporally, dust intensity exhibits a weakening trend over time, with the highest intensity occurring during spring.(3) Dust intensity in Northeast Asia is significantly influenced by both local and upwind underlying surface conditions. Specifically, NDVI in the preceding summer and following spring has the most pronounced impact on dust intensity in northern China, the Korean Peninsula, and Japan, while soil moisture is the dominant factor affecting dust intensity in Mongolia.These findings provide valuable insights into the spatial and temporal dynamics of dust intensity and highlight the critical role of underlying surface factors in dust generation and transport.

Leading economic nations need a green energy transition to avoid disastrous consequences caused by climate change. Barren regions are ideal for wind-power and solar-energy industries, providing year-long sunlight and ideal wind and sun power resources. The problem is that such regions are prone to dust storms with their adverse impacts. Dust storm forecasting necessitates including additional physical processes, which increase computational costs, a challenge for traditional weather forecasting systems. Current dust storm predictions are conducted separately using lower-resolution models, resulting in delayed and less accurate forecasts and warnings for severe events, hindering the effective operation and risk management of renewable energy facilities. To address this problem, this work proposes iDust, an innovative paradigm that deeply integrates dust storm forecasting into a global high-resolution weather forecasting model, utilizing only one-eighth of additional computational resources. iDust enables timely forecasting of severe dust storms with concentrations exceeding 1000 μg/m³ and dust transport over long distances. The new paradigm of iDust integrated weather forecasting model construction expands the coverage of operational forecasting systems to include the prediction of damaging dust storms. It provides in-depth customized support for emerging solar and wind power industries and a much-needed high-fidelity global severe dust storm dataset for scientific studies.

The Madden-Julian Oscillation (MJO) is the most prominent mode of intraseasonal oscillation in the tropical atmosphere and it has large impacts on a variety of weather and climate phenomena across different spatial and temporal scales. How well the MJO itself can be predicted could affect the subseasonal predictability of various meteorological variables associated with the MJO, and is therefore a matter of concern to the scientific and operational communities.By using the latest hindcast results from the subseasonal-to-seasonal (S2S) phase II models, this study attempts to understand the diverse prediction skill for the MJO events at the forecast start date, by investigating the preference of the prediction skill to the observational conditions.We classify the MJO events into high-skill and low-skill groups based on the SACC of individual MJO events. Compared to the low-skill MJO events, the high-skill events are characterized by a stronger intraseasonal convection–circulation couplet over the IO before the forecast start date, which could result in a longer zonal propagation range during the forecast period, thereby leading to a higher score for assessing the prediction skill. The difference in intraseasonal fields can further be attributed to the LFBS of IO sea surface temperature (SST) and quasi-biannual oscillation (QBO), with the high-skill (low skill) events corresponding to a warmer (colder) IO and easterly (westerly) QBO phase. The physical link is that a warm IO could increase the low-level convective instability and thus amplify MJO convection over the IO, whereas an easterly QBO phase could weaken the Maritime Continent barrier effect by weakening the static stability near the tropopause, thus favoring eastward propagation of the MJO. It is also found that the combined effects of IO SST and QBO phases are more effective in influencing MJO prediction skill than individual LFBS.

Recentstudies have shown that individual Madden–Julian oscillation (MJO) events can be categorized into four types based on their propagation characteristics: standing, jumping, slow-propagating, and fast propagating MJOs. While their structures and impacts are well documented in observations, their representation in state-of-the-art climate models has not been investigated. This study evaluates the four types of MJOs in phase 6 of Coupled Model Intercomparison Project (CMIP6) models. While most models show reasonable MJO propagation characteristics, their performance varies among MJO types, with better performance for the propagating MJO types. To identify the factors that drive each MJOtype, background conditions at the surface, troposphere, and stratosphere are examined. We found that the standing and fast MJOs are closely associated with sea surface temperature (SST) and precipitable water (PW) distributions. Specifically, negative and positive SST and PW anomalies are observed over the equatorial Pacific during the standing and fast propagating MJOs, respectively. In contrast, the jumping and slow-propagating MJOs do not show noticeable SST and PW distributions, suggesting that they may be more strongly influenced by the internal dynamics than by the background conditions. It is further found that the models with more frequent standing MJO than fast MJO generally have stronger negative biases in both SST and PW over the equatorial western Pacific. Although the standing MJO is also preferred when the stratospheric wind at 50 hPa is westerly, such a relationship is absent in the models. These results suggest that MJO diversity in climate models could be improved by better simulation of the mean state, especially for the standing and fast-propagating MJOs.

This study investigates the influence of equatorial Rossby (ER) waves on modulating the eastward propagation of the Madden-Julian Oscillation (MJO). Using classification methods, the standing and jumping MJO clusters are identified as those impeded by ER waves. In these clusters, MJO convection splits and shifts poleward, in contrast to the continuously propagating MJO, which moves through the MC without disruption. Column-integrated moist static energy (MSE) budget analysis and timescale separation was conducted to understand the dynamic mechanisms underlying the different behaviors among the MJO clusters. The results revealed that in the positive MSE tendency anomalies east of the MJO convection, facilitating the MJO eastward propagation, are mainly contributed by the horizontal and vertical advection. However, these positive anomalies straddle to the equator in the standing and jumping MJO clusters, leading convection to propagate poleward. The differences in MSE tendency anomalies are primarily contributed by two nonlinearly rectified terms: the meridional advection of MJO-scale winds to the low-frequency background MSE and the vertical transport of low-frequency background MSE by the MJO-filtered vertical velocity. These differences arise from the destruction of the subsidence anomaly and the anomalous poleward flows to the east of MJO convection by the ER waves. The variations in ER wave intensity among the MJO clusters are primarily influenced by the background vertical shear of zonal wind and moisture anomalies. This study provides new insights into the mechanisms driving MJO propagation and the sophisticated interactions between ER waves and the MJO. 

This work presents a prediction method of the Madden-Julian Oscillation (MJO) using reservoir computing, which is a brain inspired machine learning technique. The reservoir computing model of this study was trained to skillfully forecast the time evolution of the real-time multivariate MJO index (RMM), a macroscopic variable that represents the state of the MJO. The prediction skill of our model was extended by refinement of the training data. A novel filter (realtime bandpass filter) was developed to extract the recurrency of MJO signals from the raw atmospheric data and restrict the degrees of freedom of the training. The efficacy of the reservoir computing was further enhanced by selecting a suitably correlated time-delay coordinate of the RMM for the training. The constructed model demonstrated the skill to predict the RMM sequence for a month from the pre-developmental stages of the MJO, comparable with that of the dynamical models.

The Madden-Julian Oscillation (MJO) is the key tropical variability having widespread impact on worldwide extreme events via its global teleconnections. Yet, the dynamics of the MJO teleconnection are not fully understood, especially on how MJO teleconnection interacts with extratropical climate variability and how it responds to various MJO propagation speeds. In this study, we present an alternative dynamical view of MJO teleconnection using the method of dynamical mode decomposition (DMD). The DMD can identify spatiotemporal coherent structures from high-dimensional nonlinear flow. When applying this method to tropospheric upper-level atmosphere, the dynamics of unforced motion can be decomposed into DMD modes, and the MJO teleconnection can be interpreted as spatio-temporal resonance between the DMD modes and the forcing from MJO. As some DMD modes have regional spatial features resembling some climate variabilities, such as the Pacific North American (PNA) and North Atlantic Oscillation (NAO), the interaction between the MJO forcing and these DMD modes explains how MJO-forced circulation modulates those climate variabilities on intraseasonal time scales. The analysis also explains how MJO teleconnection responds to fast and slow MJOs, in terms of temporal resonance between the DMD modes and the MJO’s propagation speed. The study further compares this alternative view of MJO teleconnection in terms of DMD with the classical view in terms of dynamics associated with vorticity equation. The results here have implications in model simulation and prediction of MJO teleconnections.

The East Asian monsoon system shows strong annual cycle with the wet season in summer and dry in winter. Apart from the annual cycle, the monsoon system exhibits distinct climatological intraseasonal oscillations (CISOs) (Wang and Xu 1997), also known as the fast annual cycle (LinHo and Wang 2002). This study focuses on investigating the relationship between CISOs, South China Sea (SCS) westerly summer monsoon onset, and the commencement of Philippine rainy season. The SCS-CISO mode during the spring to summer transition period (March-June) is determined as the intraseasonal (20-73 days) variations of the area (10°-20°N, 110°E-120°E) mean OLR data during the 44 years from 1979-2022. A statistically significant dry (wet) singularity over the SCS is identified with the positive (dry) peak in early May and negative (wet) valley in late May. When an individual-year SCS-CISO mode shows distinct shift from dry to wet during the time window from mid-April to early-June (pentads 22-31) and the wet phase coincides with the wet singularity, the year is identified as a year of normal SCS-CISO. It turns out that 72% of the 22 normal SCS-CISO years show concurrent wet SCS-CISO and westerly monsoon onset, and about one half of the concurrency was influenced by MJO through an intensified southwest-northeast oriented moisture transport path from Indian Ocean extended to northern SCS and extratropical western North Pacific. The extended moisture transport path is associated with a MJO-Rossby wave like anomalous cyclonic circulation. It can enhance the Taiwan-Okinawa Mei-yu front and the extreme rainfall over northern Philippines. However, the fact that only in one half of the normal SCS-CISO years showing MJO influence suggests the spring to summer fast annual cycle and CISO over the SCS cannot be fully explained by MJO.

The Indo-Asian summer monsoon (IASM) is an essential component of climate system on Earth, driven by the seasonal cycle of solar radiation. It comprises a smoothed annual cycle and abrupt variations nearly fixed to the calendar day, such as onset, withdrawal, active/break cycles, and Baiu/Meiyu. They are linked to the climatological intraseasonal oscillation (CISO), which is interpreted as the phase-locking of the intraseasonal oscillation (ISO) to the annual cycle. Understanding of the CISO is vital for sub-seasonal-to-seasonal prediction in the IASM region. Yet its mechanisms still remain elusive. Here, we study possible phase-locking mechanisms of the ISO using 42-year of the outgoing longwave radiation (OLR), atmospheric, and oceanic reanalysis data. We first calculated the phase angle of the OLR time series for each year over the Arabian Sea (AS), in which the CISO signal is most prominent. Individual years are classified according to the phase information, where we target years in phase with the CISO. Composite analysis on the in-phase years suggests that the phase-locking of the ISO initiates in the AS in late March. Subsequently, dry-wet cycles repeat over the boreal summer. Lead-lag correlations between OLR, sea surface temperature (SST) and column-integrated moist static energy indicate that cloud-radiation-SST feedback is highly active in maintaining the CISO. The AS receives the maximum amount of solar radiation in late March under clear sky and weak wind conditions. Furthermore, we found the significance of a shallow ocean mixed layer over the AS for the CISO. Increased solar radiation and cloud-radiation-SST feedback in late March, and the shallow mixed layer all jointly contribute to the phase-locking of the ISO over the Indian Ocean. Our research points to the important role of the upper ocean to the IASM, which also extends to the western Pacific in linking to the Baiu/Meiyu.

The predictability of the propagation of the Madden-Julian Oscillation (MJO) across the Maritime Continent (MC) remains a challenge. Recent research has shown that constructive interference (synchronization) between oceanic Rossby waves (RWs) and the MJO is possible in the IO basin related to basin resonance timescales. Using upper ocean heat content (OHC) to diagnose oceanic RWs and an MJO track dataset (Kerns and Chen 2020), we examine MJO events that propagated across the MC and those events that did not. We find that about 80 days before the onset of propagating MJO events, anomalously positive westward propagating off-equatorial OHC, indicative of oceanic RWs is evident in the IO basin. The positive OHC anomalies are enhanced about 10 days before the MJO as the oceanic RWs synchronize with the MJO winds. The anomalously positive westward propagating OHC are not as prevalent prior to non-propagating MJO events. These results suggest that synchronization between oceanic RWs with the MJO, through upper OHC modulation, might be important in contributing to MJO propagation across the MC. Consequently, upper OHC anomalies and oceanic RWs in the IO basin could be a potential source of predictability for the propagation of the MJO across the MC.

In urban areas with high-rise and densely distributed buildings, highly turbulent flows are generated, which affects heat transport from urban surfaces to the atmosphere. Such heat transport will influence the development of isolated thunderstorms in urban areas. Sometimes, strong precipitation induced by such thunderstorms may appear as a natural disaster. Sensible heat fluxes from urban surfaces are considered to play a role in causing such thunderstorms. This study numerically investigates the impacts of the geometrical features of urban surfaces and the heat transfer from urban areas on the generation of isolated thunderstorms and the resulting local-scale precipitation by combining a mesoscale meteorological model and a building-resolving large-eddy simulation (LES) model. The Weather Research and Forecasting (WRF) model and the PALM model are used in this study. In the LES, we examine the impacts of changing surface sensible heat fluxes on the turbulent flows and heat transfer in urban areas having various geometrical features in Osaka City and quantitatively assess the changes in turbulent momentum and heat transfers with the changes in urban surface fluxes in the urban areas. The district with high-rise buildings with a higher packing density effectively transfers heat, compared with the district with a lower packing density of buildings. The impact of the changes in surface sensible heat fluxes on precipitation intensity over a mesoscale area is examined with the WRF model through changing the surface fluxes. The change in the precipitation intensity and amount through changes in heat fluxes from the urban surfaces is quantitatively assessed from the hybrid analysis using the WRF and PALM models.

Using the WRF-ARW mesoscale numerical model coupled with a single-layer urban canopy model and a realistic anthropogenic heat (AH) profile, we investigate how urban land cover and AH influence on an extreme rainfall event on 22 May 2020 in the Guangdong-Hong Kong-Macao Greater Bay Area (GBA). To assess their relative impacts, we conducted two sensitivity experiments: one replacing urban region with cropland cover and another maintaining urban land cover but eliminating AH by setting the sensible heat flux to zero. Our results reveal that urbanization amplifies both the spatial extent and intensity of extreme rainfall, driven by the combined effect of urban land cover change and AH, with AH playing a dominant role. Replacing urban regions with cropland enhances near-surface thermal mixing and convergence uplift through decreased latent heat flux. Additionally, low-level wind speeds in the mega-city region decrease due to strong surface friction, leading to enhanced rainfall in the inner urban areas. Furthermore, AH significantly increases sensible heat flux, raising low-level temperatures. Combined with intensified convergence zones, higher convective available potential energy, and lower convective inhibition, this creates a more favorable environment for convection triggering along the eastern boundary of urban areas.

Forecasting the location and timing of convective initiation (CI) and the development of convective clouds (CC) remains a challenging issue due to the impact of small-scale mechanisms. The limitations of operational observational networks, which have coarse resolution and insufficient data during clear skies, hinder our understanding of the dynamic and microphysical processes involved in CC. To address these challenges, extensive field campaigns utilizing new observational technologies and targeted strategies have been launched worldwide to investigate the characteristics of convective clouds in relation to various environmental factors and the mechanisms that trigger convection. In South Korea, a Korea Precipitation Observation Program (KPOP), International Collaborative Experiments for Mesoscale Convective Systems in Seoul Metropolitan Area (KPOP-MS) were initiated during the summer of 2023. This study examines the small-scale convection related to the initiation and development of the CC that occurred on July 30, 2023, during the KPOP-MS field campaign.             The convection initiated under a thermal low during the heat wave warning on 30 July 2023. The CC was identified and tracked to characterize the developing stages and reveal the microphysical processes involved. Initially, moisture supplied by the sea breeze (SB) played a crucial role in the development of the CC. However, as the upward motion became disorganized, the CC weakened. The CC then redeveloped over the Seoul metropolitan area due to the interaction between the sea breeze and the cold pools, which tilted the updraft toward the downdraft. This created favorable conditions for upright convection, facilitating the development of the CC. Ultimately, the CC weakened again due to the outflow disrupting the updraft. In conclusion, we examined the small-scale convection mechanisms by analyzing the thermodynamic and dynamic processes associated with the CC case during the KPOP-MS field campaign.

Boundary layer jets (BLJs) over the South China Sea (SCS) play a critical role in influencing coastal heavy rainfall in South China through moisture transport and convergence processes. Using 24 years of ERA5 reanalysis datasets (1998-2021), we employ a multivariate self-organizing map (SOM) approach to objectively classify BLJs over the SCS in the early-summer rainy season based on zonal (u) and meridional (v) wind components at 950 hPa. Our SOM clustering identifies six distinct BLJ types (from SOM1 to SOM6), each characterized by unique jet intensities, dominant wind directions, and spatial distribution of jet cores. SOM1-SOM5 predominantly feature southwesterly winds, while SOM6 is characterized by easterlies. SOM1-SOM3 show jet cores located to the south but with different wind directions, whereas SOM4-SOM4 have northern jet cores and different jet intensities. Our results indicate that the formation of these BLJ types is primarily driven by synoptic-scale background conditions. Significant differences between these types are observed in their seasonal occurrence patterns and diurnal phase characteristics. Each BLJ type shows distinct diurnal cycles with clockwise wind rotation driven by inertial oscillations. Moreover, the sequential occurrences of distinct BLJ types reveales dynamic inter-type transition mechanisms that drive the spatiotemporal organization of jet diversity.Distinct rainfall patterns are associated with each BLJ types: SOM1, characterized by southerly winds, leads to coastal torrential rainfall in werstern Guangdong (Yangjiang), while the southwesterly SOM2 is linked to weak coastal rainfall. The nearly westerly SOM3 drives offshore heavy rainfall at the jet exits. Weaker SOM3 mainly induces coastal rainfall, while stronger SOM4 primarily results in inland raifnall. Easterly SOM6 is associated with negligible precipitation.

Typhoon Gaemi (2024) exhibited a looping track before making landfall in
northeastern Taiwan and produced extreme rainfall over south Taiwan. In this study,
both global MPAS and regional TWRF models are utilized to explore the dynamic
processes in Gaemi’s looping track and the associated rainfall using 1-km horizontal
resolution. The approaching northwestward Gaemi is significantly stagnated near
northeastern Taiwan and significantly deflected southward by the induced intense
northerly flow at the west quadrant of the inner vortex and then northward by the
recirculating flow of the outer typhoon circulation around south Taiwan when flow
channeling considerably weakens. Wavenumber-1 potential vorticity (PV) tendency
budget analysis indicates that the earlier southward and later northward translations
are in response to horizontal PV advection from the northerly channeling flow and the
recirculating flow, respectively. Meanwhile, differential diabatic heating and vertical
PV advection can contribute moderate inland and offshore translation, respectively.
Both model rainfall forecasts agree well with the observation, highlighting that severe
rainfall over south Taiwan mainly results from orographic lifting of the outer typhoon
circulation near the earlier southward looping and the associated strong convergence
with ambient southwesterly near the later northward looping that helps to produce
heavy rainfall over the southwestern plains.

Comparing the cloud microphysics parameterization schemes is challenging due to the differences in the complexity of their microphysical process representations. Meanwhile, weather radars can provide significant opportunities for validating model results, especially in atmospheric microphysics processes. However, the comparison between models and observations is not straightforward because radar observables cannot be directly correlated with the model simulation variables. Cloud-resolving model Radar SIMulator (CR-SIM) (Oue et al. 2020), an advanced radar simulator, can generate various radar or lidar variables by utilizing model simulation outputs, therefore it makes the qualitive verification of microphysics processes between model and radars possible. We recently developed Weather Research and Forecasting (WRF) Double-Moment 6-class (WDM6) schemes by adding the prognostic graupel volume mixing ratio (Park et al., 2024) (WDM6_BG), which has been developed based on the WDM6 scheme with the prognostic cloud ice number concentration (WDM6_NI) (Park and Lim, 2023). Furthermore, the WRF Double-moment 7-class (WDM7) microphysics scheme, which includes the prognostic hail mixing ratio based on the WDM6 scheme, has been modified by adding the prognostic variable of cloud ice number concentration (WDM7_NI). These new versions of the WDM6 scheme have been incorporated into CR-SIM. By comparing the radar-observable variables from simulation results using recently developed cloud microphysics schemes (WDM6_NI, WDM7_NI, and WDM6_BG) with radar observations through CR-SIM, we can determine the relative importance of prognostic microphysical variables in simulating realistic microphysical processes. The detailed comparison results from the various versions of WDM6, focusing on the simulated microphysical processes, will be discussed at the conference.     *This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (grant no. RS-2023-00208394)

Using radiosonde data from April to September between 2008 and 2017, we conducted a statistical analysis of low-level jets (LLJs) characteristics across China. Our analysis focused on comparing the spatiotemporal variations and vertical structures of two types of LLJs: synoptic-system-related low-level jets (SLLJs) and boundary layer jets (BLJs), while also investigating their associated weather patterns and precipitation distributions.  LLJs occur most frequently in Northeast and South China and least in the Tibet Plateau. LLJs typically feature wind speeds ranging from 10 to 20 m/s, with winds predominantly coming from the southwest, except in Northwest China where northwesterly winds are more common. In terms of altitude, the jet nose is generally located around 700 m above ground level. A bimodal distribution of the jet core is found in North China, Central China, Northwest China, and Tibet, while a unimodal distribution is found in Northeast China, South China, and Southwest China. While BLJs are more prevalent in Northwest China and parts of the eastern coastal areas, SLLJs generally accounted for a higher proportion of occurrences across China.  Diurnal variations revealed that the frequency of LLJs was generally higher at 08 LST compared to 20 LST in most regions, except for the Tibetan Plateau and some eastern coastal areas. LLJs are often associated with strong low-level vortices. SLLJs are commonly observed in Southwest and South China, influenced by the southwest vortex and the Meiyu front, respectively. In contrast, BLJ are primarily observed in regions where terrain-induced baroclinic effects or enhanced inertial oscillations occurred.  The two types of LLJs exhibit distinct precipitation patterns. For instance, in the North China Plain, anomalous precipitation associated with BLJ occurs near the terrain, regardless of the position of the jet core. In contrast, SLLJ are linked to anomalous precipitation near the exit of the jet core.

The parameters defining hydrometeor characteristics in cloud microphysics schemes are a source of uncertainty, influencing microphysical processes both directly and indirectly, thereby altering surface precipitation. Our study obtained the range of hydrometeor characteristic parameters from observations during the International Collaborative Experiment conducted in the PyeongChang 2018 Olympics and Paralympic Winter Games (ICE-POP 2018) field campaign over the Korean Peninsula. 256 experimental sets were generated by altering the hydrometeors characteristic parameters using Latin hypercube sampling (LHS) method, and the perturbed parameter ensemble (PPE) experiments were conducted to assess precipitation performance. Based on the PPE results, a generalized linear model (GLM) was constructed using the RMSE of the simulated precipitation and the corresponding parameter values from the experiments. Additionally, Bayesian optimization was applied to identify the optimal parameter set within the observed range of parameter values. For the simulation, the Weather Research and Forecasting (WRF) Double-Moment 6-class (WDM6) cloud microphysics scheme was employed for the winter-time three representative precipitation cases observed during ICE-POP 2018, which includes cold low (CL), warm low (WL), and air-sea interaction (AS) cases. The optimized parameter-set (OPT) experiment presented the decreasing snow mixing ratio due to the reduced deposition and accretion processes in the CL case. In the WL case, the mixing ratio of rain and snow decreased due to the reduced snow melting and deposition/accretion processes. In the AS case, the snow mixing ratio decreased due to the reduced accretion and increased sublimation. Graupel mixing ratio also decreased due to the reduced accretions. In all cases, OPT experiment showed an improved RMSE for precipitation compared to the control experiments. These findings highlight that the optimizing parameters, defining hydrometeor characteristics, can alter microphysical processes and improve precipitation simulation performance.This work was funded by the Korea Meteorological Administration Research and Development Program under Grant (RS-2023-00240346)

Twelve precipitation events over the Pearl River Delta (PRD) in South China, occurring with low-level southeastward-moving shear lines prior to the monsoon onset over the South China Sea, were identified and analyzed using multiple observations. The impacts of the PRD urban agglomeration on precipitation were investigated using  convection-permitting ensemble simulations with the WRF model. These simulations incorporated the synoptic background averaged over the twelve events and compared scenarios with and without the cities in the PRD. Results reveal that these events are characterized by prevailing westerly flows in the mid-troposphere and low-level southwesterly flows south of the shear lines, which transport air with high equivalent potential temperature to the PRD.  Accumulated precipitation from 0600 local solar time (LST) to 0000 LST (+1d) is predominantly located to the north of the city cluster and near its northeastern border. Simulations indicate that the existence of urban heat islands (UHI) in megacities significantly enhances downstream convection initiation and precipitation along the northeastern urban boundaries. This UHI-induced local convection, combined with urban dynamic effects, hinders further inland transport of warm, moist air by impeding low-level southwesterly flows and depleting moisture resources. Consequently, the shear line-associated precipitation north of the urban cluster is reduced.

A tropical island city-state located near the equator, Singapore experiences a warm and humid climate throughout the year. Its geographical location and rapid urbanization make it highly susceptible to extreme weather events, such as intense rainfall associated with convective thunderstorms or extreme air temperatures elevated by additional urban heat. Focusing on rainfall, the weather systems contributing to extreme events include localized thunderstorms, Sumatra squalls, and monsoon surges—each shaped by complex interactions between land, air, and ocean dynamics in the Southeast Asia region. Sumatra squalls, originating over the mountains of Sumatra Island, are driven by southwesterly winds and cold pools that propagate eastward, often leading to heavy rainfall in Singapore, typically during the early morning hours. Studies suggest that urbanization significantly influences rainfall patterns, especially in coastal cities like Singapore, by altering land-atmosphere interactions. Urban environments can generate localized wind convergence zones, enhancing convection and increasing the likelihood of heavy rainfall. Additionally, surface heating variations across different land covers can trigger localized convergence and thunderstorms, often resulting in intense rainfall. This study presents results from several case studies of extreme rainfall events in Singapore representing afternoon convection and Sumatra squall conditions. The analysis is conducted using the uSINGV model, a customized urban version of the operational Numerical Weather Prediction system SINGV for Singapore and the region developed at CCRS/MSS based on the UK Met Office Unified Model (UM) Regional Atmosphere and Land configuration. The results also highlight the performance of high-resolution (100-meter) model simulations in capturing the processes responsible for the initiation, organization, and intensification of deep convection in the above-mentioned weather systems. The findings aim to enhance the understanding of complex weather patterns in urbanized tropical environments, contributing to the development of more accurate numerical models for weather and climate forecasting.

Shading has been widely recognized as an effective strategy for ameliorating human heat stress. The shading effect varies across different urban microclimates, influenced by urban morphology. However, existing research on evaluating shading effects mainly focuses on point-based observations or block-scale simulations, resulting in limited regional representation. To address the gaps in regional-scale evaluations, we evaluate the improvement of human heat stress (measured by Universal Thermal Climate Index, UTCI) in shaded areas of the Yangtze River Delta (YRD) by considering/not considering solar radiation when calculating the UTCI. The method primarily relies on accurate descriptions of the urban microclimate. Therefore, the Weather Research and Forecasting (WRF) model, incorporating LCZ land use data, is used to simulate the thermal environment during the summers (June–August) of 2020 and 2022 in YRD. The results indicate that the daytime average outdoor UTCI in compact high-rise building (LCZ1) is 38.19℃, with 53% of the time in the two summer years experiencing “very strong heat stress”, which is higher than sparsely built (LCZ9). After shading, the urban demonstrates a more effective shading effect than suburban. Additionally, the UTCI decreases by an average of 6.36°C across the LCZs, with large low-rise (LCZ8) demonstrating the­ best shading effect, reaching a maximum UTCI decrease of 11.01°C at 1000 LST. Furthermore, the shading effect is more significant under high-temperature and moderate-humidity conditions, whereas excessively high or low humidity diminish shading effects. Our research bridge the scale gaps in shading effect evaluations, and can provide more comprehensive guidance in improving the urban environment.

In 2023, recorded as the hottest year on Earth, extreme weather events, such as flash floods, intense rainfall, record-breaking heatwaves, and wildfires, devastated many regions. In this study, by applying a dynamical downscaling method on a bias-corrected multi-model ensemble from CMIP6, we present model projections of heatwaves under an intermediate future scenario, SSP2-4.5, in the Pearl River Delta (PRD), one of the most densely populated and urbanized areas. Our study diverges from traditional heatwave research by employing the excess heat factor (EHF), a novel index that accounts for both short- and long-term temperature anomalies. The EHF in this study is based on a three-day mean wet bulb globe temperature (WBGT), rather than the conventional dry bulb temperature, offering a more comprehensive assessment of heatwave characteristics in the PRD's high-humidity conditions. Our preliminary results show a significant increase in areas facing at least a moderate daytime heat threat (≥29.5°C), rising from 39% of the PRD’s land area in the 2010s to 95% by the 2090s. Additionally, the 90th percentile of daily minimum WBGT from the 2010s will align with the 53rd by the 2090s, suggesting that current extreme values during the night will become a new norm by the century’s end. Our results further indicate that, under SSP2-4.5, heatwaves in the PRD are projected to become more frequent, intense, extensive, prolonged, and earlier onset. The average number of days in heatwave conditions is expected to increase from 11 days currently to 40 days by the 2040s and as many as 98 days by the 2090s. The longest heatwave duration could extend to nearly two months by the 2090s. The onset of the first heatwave will shift earlier by one month—from 20th June in the 2010s to 15th May in the 2090s.

Singapore is a highly urbanized coastal city in the tropics, facing increased weather and climate risks with rising temperatures and more wet and dry extremes. The latest Singapore’s Third National Climate Change Study (V3) provides 2-km high-resolution climate change projections for Singapore and the broader Southeast Asian region to the end of 2100. In addition to rising temperatures, Singapore is also facing Urban Heat Island (UHI) effects, a phenomenon where urban areas experience significantly warmer temperatures than their surrounding rural areas, which is not addressed in V3. To bridge this gap, we employed uSINGV, a high-resolution urban-scale modelling version of numerical modelling system (SINGV) developed at CCRS, to further downscale the future climate projections to 100 m. uSINGV offers both increased model resolution and better representation of urban surfaces through an urban canopy model. In this study, we selected both historical and future hottest days from V3 to drive 100 m uSINGV simulations. The results reveal that for future hottest days, Singapore’s island wide UHI intensity is not expected to exceed that of the historical period, although near-surface temperatures will increase. The future island wide UHI intensity peaks during the night, while daytime shows a negative UHI intensity (or cool island effects). This island-wide negative UHI intensity during the day is primarily attributed to sea breezes cooling the southern coast and parts of inland Singapore, resulting in a lower island-wide average temperature compared to rural sites. This suggests that in addition to UHI intensity, we may need other measures to fully capture the complexity of future heat scenarios that Singapore will encounter.

Compound Daytime-Nighttime Heatwave (CDNHW) has caused greater socio-economic damage than independent heatwaves because the daytime heat persists into the nighttime. This study investigates the urbanization effect on CDNHW in the Seoul metropolitan area over 25 years (2000–2024). CDNHW was defined as a nighttime heatwave (TMIN>25 ℃) preceded by daytime heatwave (TMAX >33 ℃) based on ASOS (AutomatedSynoptic Observing System) Seoul station data. The urban heat island (UHI), defined as the temperature difference between urban and rural areas, was stronger during nighttime (1.47 K) than prior daytime (0.27 K) on CDNHW days. This pattern was also observed during the main period for CDNHW (from 21-Jul to 20-Aug). These suggest that UHI played a critical role in nighttime warming on CDNHW days. To understand the impact of UHI on CDNHW, numerical experiments were conducted using WRF-SLUCM for the 2018 CDNHW case (28-Jul to 12-Aug, 2018), a representative case of CDNHW. Three experiments were performed, prescribing each different peak of anthropogenic heat (AH) values (70, 140, and 210 W/m2). In most case, higher temperatures in Seoul were simulated in higher AH experiments, with temperature differences between experiments being more pronounced during nighttime than daytime.In this study, these phenomena were examined with a focus on local circulation. In Seoul, three local circulations were identified; land-sea breeze, mountain-valley breeze, and urban-rural breeze. During daytime, land-sea contrast (LSC) was larger than mountain-valley contrast (MVC), leading to dominant cooled westerly sea breezes over Seoul. In contrast, at nighttime, MVC was stronger than LSC, inducing dominant easterly winds that were heated by the Foehn-like effect. These features were intensified as AH increased. In other words, AH resulted in a cooling effect induced by the sea breezes in the daytime and a warm advection by easterly winds at night.

Integrated models of criteria air pollutants (CAPs) (e.g., O₃ and PM) and greenhouse gases (GHGs) (e.g., CO₂) provide a powerful tool for the development of synergetic strategies to mitigate air pollution and climate change. Such models, however, are lacking, due mainly to a historic separation of CAPs and GHGs modeling.  An integrated model system has been developed by our group to bridge the gaps based on Weather Research and Forecasting model coupled with Chemistry (WRF-Chem) and the WRF Greenhouse Gas Model (WRF-GHG) (referred to as WRF-Chem-GHG). WRF-Chem-GHG simultaneously simulates the evolution of CAPs and GHGs and their interactions.  In this work, regional-to-street scale simulations are performed using WRF and WRF-Chem-GHG over eastern U.S. (12-km), northeastern U.S. (3-km), and Greater Boston (GB) (1-km) and the Model of Urban Network of Intersecting Canyons and Highways (MUNICH) over GB on street-scale.  WRF and WRF-Chem-GHG results are evaluated using observations from the U.S. NOAA’s Meteorological Aerodrome Report network for meteorology, the U.S. EPA’s Air Quality System networks for CAPs, and the interpolated Orbiting Carbon Observatory 3 for CO₂. These results are used to examine the impacts of CAPs and GHGs on meteorology through feedback mechanisms. WRF-Chem-GHG 1-km predictions are used to provide urban meteorological conditions and background concentrations for MUNICH street-scale simulations of PM_2.5 and CO₂.  The Vehicular Emission Inventory model is used to estimate on-road traffic emissions. PM_2.5 predictions are evaluated with observations at the monitoring stations from the nationwide measurement networks and low-cost sensors in Greater Boston. The simulation results at 1-km by WRF-Chem-GHG and on street-scale by MUNICH are compared to assess their ability and relative strengths in reproducing observations. The coupled WRF-Chem-GHG and MUNICH system provides a useful tool for synergetic impact assessment of the CAPs and GHGs to support regional-city decision-making.

This study conducted a comparative analysis of dispersion and receptor models to assess the impact of PM₂.₅ emissions from the sintering processes of a steel plant on nearby sensitive sites. Combining air quality monitoring and modeling approaches were utilized to evaluate the contributions of sintering processes under various meteoro-logical conditions. PM₂.₅ samples were collected at two sensitive sites for analyzing their chemical compositions, including water-soluble ions (WSIs), metallic elements, and carbonaceous content. A chemical mass balance (CMB) receptor model was applied to resolve potential sources and their contributions, while an air quality dispersion model was employed to simulate the transport and distribution of sintering emissions. The results revealed that the contributions of sintering emissions to PM₂.₅ levels at the sensitive sites varied significantly depending upon wind direction. High PM_2.5 contributions were observed at downwind locations, particularly during fall and winter when meteorological conditions favored pollutant accumulation. Chemical analysis indicated that sulfate (SO₄²⁻), nitrate (NO₃⁻), ammonium (NH₄⁺), iron (Fe), and calcium (Ca) were the dominant components in PM₂.₅, with seasonal variations in their relative abundance. Receptor modeling results estimated that the sintering process contributed less contribution of PM₂.₅ at the sensitive sites. By receptor model analysis with dispersion modeling, this study provided a comprehensive assessment of the spatiotemporal impact of sintering emissions on local ambient air quality. The findings highlight the importance of incorporating meteorological variations into emission control strategies. Future efforts should focus on adaptive emission management practices, such as adjusting sintering process emissions based on forecasted weather patterns, to minimize PM₂.₅ pollution at the sensitive sites near the steel plants.

Open biomass burning (BB) in Southeast Asia has significant adverse impacts on air quality in the region and in downwind areas. While previous studies have shown that many areas in China can be impacted, quantitative assessment of the contributions of BB on various parts of China has not been extensively reported. In addition, the local impacts of BB are affected by plume rise. In this study, we used a revised Community Multiscale Air Quality (CMAQ) model that includes the light absorption of brown carbon to model the long-range transport of precursors, ozone, and fine airborne particulate matter from spring (March to May 2022) open BB in southeast Asia to various parts of China, including the Tibetan Plateau, the Yungui Plateau, and the Pearl River Delta and the Yangtze River Delta regions. The biomass burning emissions will be estimated using an in-house emission processor that generates CMAQ ready-emissions directly from the MODIS and VIIRS satellite data. Different plume rise algorithms will be tested to assess the sensitivity of predicted local and long-range transport of BB burning emissions due to plume rise estimations. Population exposure and the associated adverse health effects will be quantified using the model results.

Biomass burning (BB)-derived brown carbon (BrC) undergoes aqueous-phase photooxidation in cloud/fog droplets, forming aqueous secondary organic aerosol (aqSOA) with significant climatic and aerosol physicochemical implications. However, the mechanistic pathways governing these transformations remain poorly constrained. This study investigates the direct photolysis and OH radical oxidation of four nitroaromatic BrC species—4-nitrocatechol (4NC), 4-methyl-5-nitrocatechol (4M5NC), 3-methyl-6-nitrocatechol (3M6NC), and 5-nitrosalicylic acid (5NSA)—under atmospherically relevant aqueous conditions. Their second-order OH rate constants were (5.0±0.2)×10⁹, (7.2±0.3)×10⁹, (5.4±0.2)×10⁹, and (5.7±0.4)×10⁹ M^-1s^-1, respectively, corresponding to atmospheric lifetimes of 11–16 hours. Time-resolved optical property analysis revealed dynamic light-absorption behavior during photooxidation: 4NC, 3M6NC, and 5NSA exhibited alternating photo-enhancement and bleaching at 420 nm, consistent with humic-like substance (HULIS) formation. In contrast, 4M5NC showed monotonic absorbance decay, underscoring substituent position and steric effects in modulating aqSOA optical evolution. Non-target screening via ultra-performance liquid chromatography coupled with high-resolution mass spectrometry (UPLC-HRMS) identified 55 transformation products, including 19 oligomers uniquely generated through aqueous-phase reactions. These BrC transformation products were also detected in real PM_2.5 samples, highlighting aqueous-phase processes as critical drivers of BrC aging. Notably, functionalized products were less abundant than oligomerized species, suggesting aqueous oligomerization as a key pathway for aqSOA formation. This work establishes that BrC-derived transformation products, particularly oligomers, are atmospherically persistent and significantly contribute to aqSOA burden, with implications for aerosol radiative forcing and atmospheric chemistry models.

Peat fires are a serious problem on the island of Sumatra in Southeast Asia. Peatlands are strata in which dead plants have carbonized and accumulated in water, and thus contain more carbon than other soils. The large quantities of smoke from these fires, which contain toxic substances, mix with exhaust gases and cause severe haze in the surrounding cities and sometimes even in neighboring countries.　Smoke pollution not only pollutes the air and causes respiratory and other problems, but also leads to economic losses by reducing visibility and severely limiting the use of airplanes and automobiles. Smoke is also an international problem, as it rides on air currents and harms neighboring countries across oceans.　Since biomass burning aerosols (BBAs) in smoke has the property of absorbing sunlight, near-ultraviolet (UV) data can be used to detect absorptive aerosols such as BBA.The second-generation global imager (SGLI) on board the Japan Aerospace Exploration Agency’s Global Change Observation Mission-Climate (JAXA/GCOM-C) is a 19-channel multispectral sensor with wavelengths ranging from UV to thermal infrared (IR), including red and near-IR polarization channels. Our recent work demonstrates that these features of the SGLI are useful for characterizing BBAs. The geostationary orbiter Himawari-8 can observe the Southeast Asian region with high temporal resolution. In this study, we will use ground data and satellite observation data derived from SGLI　and Himawari-8, to understand the smoke damage.In addition, transport of pollutants during large-scale peat fires will be discussed from simulations using a chemical transport model used the meteorological fields simulated by SCALE (Scalable Computing for Advanced Library and Environment) regional model for offline calculations.

The mixture of urban emissions and seasonal transported biomass burning (BB) emissions of tropical peat forest fires in Southeast Asia impact regional air quality. A soot-particle aerosol mass spectrometer (SP-AMS) was deployed in Singapore during the smoke-haze episode in 2019 to investigate the influence of transported BB on local air quality. The SP-AMS was operated with dual vaporization scheme, simultaneously characterizing non-refractory particulate matter and refractory black carbon (rBC) and alternating between bulk and single particle measurements.  Two periods predominantly affected by smoke haze were observed, coinciding with severe regional peat forest fires between September and November 2019. with emissions originated from peat-forest fires at Kalimantan and Sumatra in September and mostly from Sumatra in November. During October, a more typical urban influence period allows us to identify sources other than transported BB smoke (such as cooking and two different type of traffic) by positive matrix factorization. The secondary sources include regional background, locally formed and transported smoke haze-related organic aerosols (Haze-OAs). The organic-to-rBC ratios during the three periods vary from 1.9 (Oct.) to 5.2 (Sept), with the highest (median) rBC concentration reaching 6.1 μg m^-3 in November. The mass spectra of BB-smoke aerosols are characterized by different contribution of levoglucosan and potassium related fragments (K⁺ and K₃SO₄⁺), with K-rich particles identified through single particle clustering.  The clustering also evidenced various degrees of internal mixing between organic and sulphate (SO₄⁺) with mass contribution ranging from 25–50% (OA) and 30–50% (SO₄⁺), as well as larger (>300 nm) rBC-rich and hydrocarbon-like containing particles. The temporal evolution of OA sources and particle mixing state throughout these smoke haze episodes improve our understanding of reaction processes of fine particulates.  The implications on relevant optical and cloud properties will be provided during the presentation.

Biomass burning is a significant source of trace gases and aerosols, with substantial implications for air quality and climate. In particular, the impacts of biomass burning on tropospheric composition, including ozone and other reactive species, are not yet fully understood. This study aims to simulate the effects of biomass burning on global tropospheric composition, with a focus on ozone and other reactive species. We employed the CESM1/MAM3 model to simulate and quantify the impacts of biomass burning emissions on key tropospheric components, including aerosols, volatile organic compounds (VOCs), nitrogen oxides (NOx), carbon monoxide (CO), and ozone (O₃). To improve the reliability and accuracy of the simulations, the model results were constrained using observational data from the Atom and isotopic measurements. The results indicate that biomass burning significantly affects the concentrations of ozone and other reactive species, such as HOx and NOx, in the global troposphere. The inclusion of biomass burning emissions notably altered the chemical composition in regions influenced by fire emissions. These results provide valuable insights into the role of biomass combustion emissions in atmospheric chemistry and climate change, and provide scientific support for policy decisions aimed at reducing biomass combustion emissions to mitigate ozone pollution and climate impacts.

This study assessed the altitude dependence of biomass burning (BB) aerosol emission variability in continental Southeast Asia using five BB emission inventories: FEER (Fire Energetics and Emissions Research version 1.2), FINN (Fire INventory from NCAR version 2.5), GFAS (Global Fire Assimilation System version 1.2), GFED (Global Fire Emissions Database version 5), and QFED (Quick Fire Emissions Dataset version 2.6 release 1). We focused on PM2.5 emissions and categorized the study region into three altitude groups: lowland (below 500 m), midland (500-2000 m), and highland (above 2000 m). Our findings reveal a clear altitude dependence in BB emissions. The midland region exhibits the highest emission amounts during the main burning period (March-April). The lowland also shows significant emissions, particularly from February to April, while the highland experiences the lowest emissions, with a burning season spanning from January to April. The burning season emissions are the major contributors to the annual emission amounts for all altitude groups. Both midland and lowland regions demonstrate gradual reducing trends in annual emission amounts over the past two decades (2003-2022), whereas the highland region shows larger interannual variability. Additionally, there is a weak increasing trend in BB emissions during the summer months, particularly for the highland. The five BB datasets consistently support these findings. We propose a percentile-based emission classification to assess emission levels, highlighting that high levels of BB emissions dominate the interannual variability and long-term trends of the main burning season emissions, while lower levels contribute to the long-term increasing trend in summer.

Biomass burning occurs frequently in Peninsular Southeast Asia (PSEA) in spring, and has an important impact on air pollution and regional climate in southern China. This study quantitatively evaluates the impact of biomass burning in PSEA during springtime on PM2.5 and ozone formation, and extreme drought in southern China. The results show that: (1) Biomass burning in PSEA increased PM2.5 concentrations by 68.0% and 24.1% in Yunnan Province and downwind provinces, respectively. The increase in PM2.5 concentration in Yunnan Province was mainly caused by primary pollutants, while the increase in PM2.5 concentration in downstream provinces was mainly caused by secondary particulate matter. Biomass burning in PSEA increased O3 concentrations by 19.4% in Yunnan Province and decreased O3 concentrations by 5.3% in downwind provinces, respectively. The increase in O3 concentration in Yunnan Province was due to gaseous precursors, while the decrease in O3 concentration in downwind provinces was mainly due to the inhibition of photolysis rates. (2) From 2009 to 2010, the southwest region suffered a severe drought event that occurred once in a century. Sensitivity simulations showed that biomass burning increased the drought severity by 0.01–0.75 levels, enlarged drought areas by about 10%, and prolonged the drought duration mainly by one month in southwest China. Biomass burning mainly affects drought by reducing precipitation: on the one hand, it forms an inversion layer with warming in the upper layer and cooling in the lower layer; on the other hand, it further weakens the water vapor transport to the southwest region by reducing the temperature difference between land and sea. In addition, biomass burning warms the ground and atmosphere north of the Bay of Bengal, weakens the southern trough, and further reduces precipitation in the southwest region.

By-products of wildfires can have significant impact on public health by travelling long distances from the source, bringing pollutants into cities far away from the fire. This research focuses on nitrogen dioxide (NO₂), carbon monoxide (CO), and formaldehyde (HCHO) down-plume enhancements and emissions rates (ERs) using TROPOspheric Monitoring Instrument (TROPOMI) satellite measurements. A 1-D advection-diffusion model calculates ERs from discernable wildfire plumes, by creating a grid of 4-km binned TROPOMI background-subtracted vertical column densities (VCDs) downwind and fitting a Gaussian distribution along the bins to rotate the grid downwind. Emission rates are calculated for nine wildfires in Western Canada during the summer 2023 wildfires and for three fires observed by the Fire Influence on Regional to Global Environments and Air Quality (FIREX-AQ) campaign (summer 2019). ERs show CO emission near the source of the fire, with NO\textsubscript{2} and HCHO forming further down the plume and HCHO rates increasing as NO₂ rates decrease. Maximum ERs also showed a slight linear trend for all species versus daily-averaged fire radiative power (FRP). Canadian wildfire plume VCD enhancements exhibited positive correlations between NO₂ and CO, consistent with other studies, but the TROPOMI HCHO data product's sensitivity to noise made it difficult to obtain significant correlations with CO or NO₂. Finally, formaldehyde nitrogen (HCHO/NO₂ VCD) ratios (FNR) demonstrate the dependency of ozone (O₃) production on VOC and NO_x VCDs. FNR values down-plume show VOC-sensitive production mostly near the source of the fire, with NO_x-sensitive production more prominent further down the plume, raising health concerns for areas with high NO_x, such as cities. This study also observed a negative exponential for maximum FNR values versus daily-averaged FRP, suggesting a slight correlation between HCHO and NO₂ that decreases exponentially with higher FRP.

This study investigates aerosol characteristics during forest fires (2018-19) in the Uttarakhand region of India. It utilizes multi-satellite datasets, including the Visible Infrared Radiometer Suite (VIIRS), Moderate Resolution Imaging Spectroradiometer (MODIS), and Cloud and Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO), to analyze active fire points, aerosol columnar concentration, and vertical distribution. The study identifies the peak fire period from April to June, primarily due to high temperatures, low moisture, dried-up natural springs, and abundant forest fuel. The research reveals that May and June 2019 saw the highest number of fires (1573) compared to the major fire events of 2016 (1327). The worst-hit districts were Almora (360 fires), Nainital (325), Pauri Garhwal (183), and Tehri Garhwal (167), with the largest affected areas in Almora (719.9 hectares), Pauri Garhwal (304.8 hectares), and Nainital (294.15 hectares). The study of aerosol column concentration shows a significant increase in aerosol loading (AOD > 1) during and following fire events. Vertical extinction profiles reveal maximum aerosol concentrations near the surface, with variations extending up to 4-5 km in altitude. Vertical feature masks and aerosol subtype profiles indicate pollution from dust and elevated smoke spanning the surface to 5 km, with smoke concentrations reaching heights of 3-4 km during intense fires. Using the HYSPLIT trajectory model, the analysis identifies probable sink zones for pollutants, including Nepal, Tibet, Uttar Pradesh, Bihar, and northeastern states of India.

Polycyclic Aromatic Hydrocarbons (PAHs), a group of persistent organic pollutants (POPs) widely distributed in the environment, are mainly produced by incomplete combustion of organic materials. Monitoring alone is unable to conduct a long-term and broad regional analysis. Modeling is an effective tool to represent the chemical and physical processes of PAHs. In this study, we simulated Benzo[a]pyrene (BaP, one of the most toxic and highly carcinogenic PAHs) from global (1°×1°) to regional (0.33°× 0.33° and 0.11°×0.11°) scales in 2013 and 2018 by coupling the state-of-the-art physical and chemical modules associated with PAHs into a global-regional nested atmospheric aerosol and chemistry model (IAP-AACM). The model successfully reproduced the global and regional spatial distribution characteristics and seasonal variations of BaP concentrations.The results showed that the BaP annual mean concentrations are the highest in China followed by Europe and India, with high values exceeding the target values of 1ng m^-3 over some areas. In addition, we quantified both global changes in BaP concentrations and changes in health risks posed by BaP from 2013 to 2018. Compared with 2013, the most significant increases were seen in India, Europe, southeast Asia, and south Africa. By contrast, the concentration in Russia and the United States shows an decrease. The concentration of BaP in China decreased in 2018 due to emission reductions, with the largest decrease in eastern China. However, the concentration in the Sichuan Basin showed an inverse trend, reflecting the impact of meteorological conditions. Using the ROI-T that considers temperature and humidity, we simulate higher values for most regions. The health risk results indicated there was a slight decrease in 2018 compared to 2013and there remain potential risks in 2018.

As a national ecological civilization pilot zone with unprecedented strategic importance in China, Hainan Province requires precise identification of pollutant sources and PM_2.5 background concentrations to achieve world-leading air quality objectives. This study established a three-dimensional regional air quality modeling system (WRF-CMAQ) and integrated multi-source ground observation data for validation, systematically analyzing the spatiotemporal impacts of natural emissions, local anthropogenic activities, and regional transport on PM_2.5 concentration in Hainan. The findings revealed that local anthropogenic emissions constituted the predominant contributor to PM_2.5 concentrations (49.5%) in Hainan, while local biomass burning showed negligible impact (0.9%). Regional transport accounted for 29.3% of PM_2.5, comprising biomass burning (10.9%) and anthropogenic emissions (18.4%) from neighboring areas in Hainan. Natural sources contributed significantly, with biogenic volatile organic compounds (10.4%) and sea salt aerosols (1.4%) demonstrating measurable influences. After excluding local anthropogenic emissions, the annual mean PM_2.5 background concentration in Hainan was determined as 6.7 μg m^-³. Seasonal analysis indicated elevated background concentrations during spring, winter, and autumn. Comparative results from the literature indicated that Hainan's PM_2.5 background concentration was comparable to levels determined in other clean regions around the world. Compared to current ambient PM_2.5 levels, Hainan demonstrated substantial potential for additional pollution mitigation. To optimize air quality improvement, we proposed establishing interregional cooperation frameworks between Hainan and neighboring regions to implement targeted control strategies addressing both biomass combustion and anthropogenic emission source. These findings delivered critical scientific evidence to policymakers, supporting the formulation of data-driven regulatory measures and strategic roadmaps essential for realizing Hainan's 2035 vision of world-class air quality standards.

The concentration of roadside air pollutants varies spatially and temporally, influenced by traffic-related factors such as traffic volume, vehicle type, and driving speed, as well as meteorological conditions like wind speed, wind direction, and temperature. Additionally, the interaction between emission sources and meteorological conditions causes pollutants to disperse both horizontally and vertically, leading to concentration differences at different altitude. This study aims to measure the concentration of air pollutants occurring on highways using both ground monitoring stations and Unmanned Aerial Vehicles (UAVs). Based on ground measurement data, learning algorithms were applied to predict altitude-based concentration changes.The study was conducted on the Gyeongbu Expressway, South Korea’s busiest highway, with high-traffic sections selected as measurement sites. Arim Air OA200 (Arim Science, Korea) sensors were installed at these locations to measure PM_2.5, PM₁₀, O₃, NO₂, and CO concentrations. The same equipment was also mounted on DJI Matrice 300 RTK drone (DJI, China) to collect pollutant concentration data at altitudes ranging from 0 to 60 meters. Measurements were taken at 07:00, 13:00, and 17:00 in August 2024 and February 2025 to analyze seasonal and diurnal variation in air pollution levels.Machine learning algorithms including Light Gradient Boosting Model (LightGBM), Random Forest (RF), and eXtreme Gradient Boosting (XGBoost) were applied to develop predictive models. The model’s performance was evaluated using metrics such as Root Mean Squared Error (RMSE) and R-squared (R²), and the best-performing model was selected to predict altitude-based concentration changes from measurement data.This study provides a deeper understanding of the dynamic changes in roadside air quality and offers insights for developing effective roadside air pollution management strategies.

This study estimated and visualized the 3-D changes in Ultrafine Particle (PM_0.1) concentrations in an area with dense high-rise buildings in Taichung, Taiwan. We used an unmanned aerial vehicle (UAV) system equipped with high-precision instruments to measure PM_0.1 concentration data, further applied the Geo-AI model to capture the complex relationship between air quality data and land-use factors at different altitudes. The model performed excellently with an R² value of 0.95. External validation, ten-fold cross-validation, and stratified validation all confirmed the model's robust predictive capability across different sub-clusters. This study used the powerful 3-D mapping function of the geographic information system to display the PM_0.1 spatial prediction results, examined the 3-D interactive relationship between pollutant concentrations and surrounding land use. The results of this study can be applied to urban planning and public health, offering more targeted health recommendations and intervention strategies.

Selenium is a naturally occurring element that cycles in the environment via biogeochemical processes and can exist in different oxidation states. Dimethyl selenide (DMSe) and dimethyl diselenide (DMDSe) are the major volatile methyl derivatives present in the atmosphere. However, little is known about the secondary organic aerosol (SOA) formation processes and the role of selenium in the resulting SOA oxidative potential. In this study, we investigated the formation of SOA from the oxidation of DMSe and DMDSe in laboratory chamber experiments. A combination of online and offline mass spectrometric approaches was employed to characterize the chemical composition of DMSe- and DMDSe-derived SOA at the bulk and molecular levels. Reactive oxygen species (ROS) and changes in selenium oxidation states were characterized using electron paramagnetic resonance spectroscopy (EPR) and X-ray photoelectron spectroscopy (XPS). Our findings indicate that aqueous processing of DMSe- and DMDSe-derived SOA is key to the production of ROS, with the changes in selenium oxidation state observed before and after aqueous dissolution. These results provide mechanistic insight into the role of selenium in atmospheric chemistry and its potential impact on environmental processes.

The evidence linking exposure to fine particulate matter (PM2.5) with childhood allergic diseases is limited and inconsistent. In 2018, a population-based cross-sectional survey was carried out, which included 131,412 children aged 6-18 years from six cities in South China.. This study aimed to explore the association between acute PM2.5 mass concentration (PMC) and its oxidative potential (OPP) with childhood allergic diseases. The dithiothreitol (DTT) method was adopted to determine the OPP, expressed as DTTv and DTTm (normalized by air volume and mass, respectively). The prevalence of childhood allergic diseases was assessed using the International Study of Asthma and Allergy in Children (ISAAC) standardized questionnaire. Our findings revealed significant positive associations between childhood allergic diseases and both PMC and OPP. Notably, the associations of OPP were stronger than that of PMC. For example, the adjusted odds ratios (OR) of allergic rhinitis (1.043 vs. 1.038), conjunctivitis (1.012 vs. 1.004) and eczema (1.034 vs. 1.021) with per interquartile range (IQR) DTTm increase were higher in OPP, than PMC. Additionally, compared to their counterparts, the allergic diseases risk was higher among boys (OR of allergic rhinitis: 1.048 vs. 1.037) and children aged ≤12 years (1.046 vs. 1.039) for both PMC and OPP exposure than (P < 0.05). These results suggest that PMC and OPP are associated with an increased risk of childhood allergic diseases, with OPP demonstrating stronger associations than PMC.

The oxidative potential (OP) of particulate matter (PM) is a crucial indicator for assessing its capacity to induce oxidative stress, thereby impacting human health. However, research on the relationship between OP and the intrinsic composition and sources of inhalable particles remains insufficient. Existing studies often rely on single data sources and traditional modeling approaches, which are inadequate for uncovering the complex nonlinear relationships between PM composition and OP, and lack effective methods to balance predictive accuracy with causal analysis. Machine learning (ML), with its ability to capture high-dimensional data and nonlinear relationships, is considered a powerful tool for exploring the statistical connections between health outcomes and the intrinsic composition of inhalable particles, as well as for developing predictive models for atmospheric PM OP. Therefore, this study integrates research on the relationship between PM composition and sources with OP, and the application of ML modeling in atmospheric PM studies, systematically reviewing the current state of ML applications in this field and fully exploring the potential of ML in predicting PM OP. The study indicates that future research should focus on expanding data collection scope, overcoming the bottleneck of diversified data integration, promoting the collaborative application of hybrid ML methods and quantitative source apportionment techniques to identify fine pollutant source attributions, determining causal relationships between pollutants and diseases, utilizing Bayesian optimization and SHAP for model parameter tuning and interpretability, and integrating multi-source data to achieve big data-driven personalized pollution protection.

The oxidative potential (OP) of particulate matter (PM) is crucial for understanding its ability to generate reactive oxygen species. However, the major chemical drivers influencing OP still need to be better understood. This study investigated the seasonal variations of OP and identified key drivers and source mechanisms in the industrial city of Zibo, located in North China Plain. We used the XGBoost model and Positive Matrix Factorization (PMF) to identify key drivers and source mechanisms. In 2022, PM_2.5 samples were collected from an urban site in Zibo, and major chemical components were analyzed. OP was quantified using the dithiothreitol (DTT) method. The results revealed that the annual average DTTv in Zibo City for 2022 was 1.1 nmol/min/m³, with the highest DTTv levels observed in autumn, followed by spring, summer, and winter. Using the XGBoost model, we identified that metal elements such as Pb, Ba, and Cu, along with water-soluble ions NO₃⁻ and SO₄²⁻, significantly contributed to DTTv. Source apportionment analysis via PMF identified five major sources of PM_2.5. Throughout the study period, secondary particles were the predominant contributors to PM_2.5 (49 %), while coal combustion had the lowest contribution (7 %). To further elucidate the sources of OP in PM_2.5, we integrated the measured OP with source contributions derived from PMF. The findings indicated that secondary particles and industrial sources contributed the most to DTTv, accounting for 40 % and 21 %, respectively. The OP sources exhibited seasonal variations: secondary particles were the primary contributors in winter, while dust sources dominated in spring. In summer, vehicle emissions increased substantially, and industrial emissions became the major source in autumn. This study highlighted the critical drivers and source mechanisms of OP in industrial cities and would be beneficial for future air quality control and risk reduction.

Particulate matter (PM) emitted from road traffic causes adverse health effects upon inhalation and respiratory deposition. Non-exhaust emissions will eventually become the dominant source of traffic PM upon transition to electric vehicles; however, non-tailpipe PM is currently unregulated as its health impacts are still unclear. In this study, we generated brake wear particles (BWPs) with non-asbestos organic, ceramic, and semimetallic brake pads using custom dynamometers and measured aqueous-phase formation of reactive oxygen species (ROS). We found that BWPs do not contain environmentally persistent free radicals (EPFRs), and all types of BWPs generate exclusively ·OH radicals in water. BWPs generated by ceramic and semimetallic brakes during heavier braking lead to higher ·OH yields compared to gentle braking conditions, suggesting higher ·OH formation potential from ultrafine BWPs. Chemical characterization reveals that organic and elemental carbon correlated positively with ·OH formation while exhibiting negative correlations with abundant metals including Fe and Mn. We suggest that the source of ·OH is thermal decomposition of organic hydroperoxides derived from phenolic resin. PM oxidative potential quantified with the dithiothreitol (DTT) assay exhibited a positive correlation with the ·OH yield. These results provide critical insights into the toxicity and adverse health effects of BWPs.

Organic aerosol has significant impacts on human health. However, detailed mechanisms remain poorly understood due to their complex nature in the atmosphere. In this study, PM_2.5 samples were collected daily in Chongqing, a typical megacity located in Southwestern China, with molecular-level composition characterized by an ion mobility spectrometry-time of flight mass spectrometry. Nitrogen-containing organic species (CHNOs) were found to be important components of PM_2.5 during polluted episodes, with biomass burning and secondary transformation process identified as the major sources. Evidence from IMS-derived collision cross sections further illustrated organic nitrates/nitrites (ONs) as the main contributors to CHNOs. The formation of ONs was promoted by NH₄⁺ via its impact on aerosol liquid water content and aerosol acidity. The impact of NO₂ was predominant under high NO₂ and high relative humidity (RH) conditions, which is likely related with the heterogeneous of conversion of NO₂ to CHNOs. High levels of NH₄⁺ and NO₂ accompanied by high RH resulted in high levels of particulate CHNOs, which is not only crucial for haze formation but also important for adverse health effects. Significant correlations were observed between the acellular and cellular oxidative potential metrics (OP^AA, OP^GSH, OP^DTT, OP^EPR, and ROS) and CHNOs (e.g. C₈H₁₃NO₈). To the authors’ knowledge, this is the first study linking molecular-level CHNOs, with ONs further identified, and oxidative stress impacts characterized for PM_2.5 during wintertime haze episodes, which would be beneficial to future air quality regulation as well as mitigation of PM health impacts evolved under complex air pollution conditions.

Long-term exposure to fine particulate matter (PM_2.5) is associated with a range of respiratory and cardiovascular diseases. PM_2.5 consists of a myriad of organic and inorganic species, with its toxicity varying according to its chemical composition, sources, and other physical/chemical properties. In this project, we aim to investigate the oxidative potential (OP), cellular inflammation, and oxidative stress responses induced by five different chemical fractions of urban PM_2.5 samples (namely water-soluble total, water-soluble metals, water-soluble non-metals, lipid soluble, and total PM_2.5 extract). Additionally, we examined the synergistic and antagonistic interactions among these fractions in relation to overall PM_2.5 toxicity. A comprehensive chemical characterization of PM_2.5 samples were conducted. The OP of water-soluble PM_2.5 was assessed using DTT and 2-OHTA assays. Metals were found to be the predominant contributors to DTT consumption, whereas organics were the largest contributors to ∙OH generation. Notably, a significant antagonistic effect between water-soluble metals and non-metals fractions on ∙OH generation was observed in samples with high PM_2.5 concentrations. We monitored TNF-α secretion and the gene expression of pro-inflammation and oxidative stress makers (i.e., Cxcl2, Hmox-1, and Cyp1a1) in macrophages exposed to the five PM_2.5 chemical fractions. The water-soluble total fraction reduced the highest levels of TNF-α secretion and upregulation of Cxcl2, Hmox-1, and Cyp1a1, indicating a significant contribution of water-soluble components in PM_2.5-induced cytotoxicity. Additionally, water-soluble metals and non-metals fractions synergistically upregulated Hmox-1 expression but antagonistically modulated Cxcl2 expression. A general antagonistic effect between the lipid-soluble fraction and the water-soluble total fraction was observed in TNF-α secretion and oxidative stress gene expressions. These results underscore the significant role of water-soluble components to the toxicity of PM_2.5 and provide a comprehensive understanding of the interactions among different PM_2.5 chemical fractions in contributing to overall PM_2.5-induced toxicity.

Environmentally Persistent Free Radicals (EPFRs) represent a novel class of hazardous substances, posing risks to human health and the environment. In this study, we investigated the EPFRs in ambient fine, coarse, and total suspended particulate matter (PM_2.5, PM₁₀, TSP) in rural North China Plain, where local primary emissions of EPFRs were limited. We observed that the majority of EPFRs occurred in PM_2.5. Moreover, distinct seasonal patterns and higher g-factors of EPFRs were found compared to those in urban environments, suggesting unique characteristics of EPFRs in rural areas. The source apportionment analyses revealed atmospheric oxidation as the largest contributor (33.6%) to EPFRs. A large water-soluble fraction (35.2%) of EPFRs was determined, potentially resulting from the formation of more oxidized EPFRs through atmospheric oxidation processes during long-range/regional transport. Additionally, significant positive correlations were observed between EPFRs and the oxidative potential of water-soluble PM measured by dithiothreitol-depletion and hydroxyl-generation assays, likely attributable to the water-soluble fractions of EPFRs. Overall, our findings reveal the prevalence of water-soluble EPFRs in rural areas and underscore atmospheric oxidation processes can modify their properties, such as increasing their water solubility. This evolution may alter their roles in contributing to the oxidative potential of PM and potentially also influence their impact on climate-related cloud chemistry.

Atmospheric particulate matter (PM) exposure is implicated in respiratory pathologies, where the respiratory microbiota plays a pivotal role in maintaining pulmonary homeostasis. Oxidative stress, mediated by reactive oxygen species (ROS), is a central mechanism of aerosol toxicity, yet its interplay with microbial ROS dynamics and viability remains poorly characterized. This study investigates the effects of isoprene secondary organic aerosols (SOA), a dominant SOA component, and 9,10-phenanthrenequinone (PQN), a redox-active quinone abundant in PM, on two representative respiratory microorganisms: Staphylococcus aureus and Pseudomonas aeruginosa. Microbial responses were evaluated through growth inhibition, ROS, malondialdehyde (MDA), cell surface hydrophobicity (CSH), Fourier-transform infrared (FTIR) spectroscopy, lipidomics, and morphological imaging. Preliminary results showed that PQN induced significantly greater growth inhibition than isoprene SOA in both species, with S. aureus exhibiting higher sensitivity. ROS assays demonstrated that superoxide production in S. aureus exceeded that of P. aeruginosa under exposure, with PQN eliciting stronger oxidative stress. MDA quantification confirmed pronounced lipid peroxidation in S. aureus under PQN exposure. FTIR spectral analysis revealed that PQN drives membrane remodeling in S. aureus, marked by increased lipid saturation and reduced fluidity, whereas P. aeruginosa remained unaffected. Divergent CSH responses were observed: Gram-negative P. aeruginosa displayed an increase followed by decrease trends with rising PQN concentrations, contrasting the inverse pattern in Gram-positive S. aureus. Morphological analysis identified PQN/ isoprene SOA-induced cellular deformation, crumpling, and aggregation in both species. Lipidomics identified PQN dose-dependent increases in PC 34:1 in P. aeruginosa and decreases in 1,2-dipalmitoyl phosphatidylglycerol in S. aureus, underscoring species-specific membrane adaptations and proposing these lipids as biomarkers of PM stress. These findings implicate oxidative stress as a critical pathway mediating organic aerosol perturbations in respiratory microbiota. 

  The space-based cloud profiling radar (CPR) on CloudSat has been valuable in evaluating and improving cloud processes of numerical weather prediction (NWP) and climate models. Assimilating CPR will also be beneficial for improving accuracy in NWP analysis and forecasts. EarthCARE/CPR, which was launched in May 2024, is expected to bring more benefits because it provides more accurate observations with higher sensitivity to clouds. Successful assimilation of CPR observations requires a deep understanding of the characteristics of CPR observation and its simulation from the NWP model used in the data assimilation system. This study aims to evaluate simulation by comparing CPR observation and simulation made by the global model at Japan Meteorological Agency.   This comparison study started for CloudSat/CPR and now deals with EarthCARE/CPR. RTTOV ver13.0 is used as a radar simulator to simulate assimilation variables of radar reflectivity. EarthCARE/CPR reflectivity observations are obtained from L2a CPR One-sensor Echo Product and averaged to match an assimilation horizontal scale (~55 km). The comparison for three weeks in August 2024 shows that simulated reflectivity is smaller in its variability and slightly weaker in mean echo than observed reflectivity except at high altitudes.  The reflectivity departure of observation from the simulation is carefully examined because it is especially important for data assimilation. For example, its probability density function, situation dependency, and scale representativity are being examined. Based on these findings, the development of assimilation preprocessing such as quality control and observation error assignment is underway. In the meeting, we will present the latest findings and discuss how to effectively assimilate EarthCARE/CPR.

This study systematically addresses the bottleneck in the application of geostationary satellite infrared radiation assimilation in cloud-covered areas. Traditional clear-sky radiance assimilation methods fail to effectively handle the complex scattering effects in cloud regions, resulting in underutilization of cloud microphysical information, which directly affects the accuracy of initial field construction for convective systems. Although existing research has attempted to use all-sky radiation assimilation methods, the empirical large observation error schemes applied in thick cloud regions significantly limit the efficiency of data utilization. To overcome this technical bottleneck, this study innovatively proposes an adaptive regional partition assimilation framework based on the identification of cloud optical thickness thresholds: First, high-precision optical thickness products are obtained by FY-4A AGRI visible/infrared multi-channel synergistic inversion, and a three-dimensional cloud classification matrix is constructed to precisely partition thick clouds, thin clouds, and clear sky regions. A differentiated assimilation strategy system is then established - clear sky regions retain traditional radiation assimilation, thin cloud regions introduce an asymmetric observational error model and use covariance inflation to increase the weight of observational data, while thick cloud regions innovatively establish physical links between cloud water path (CWP), cloud ice path (CIP), and model condensate variables for indirect assimilation. By coupling the FY-4A AGRI clear-sky radiation and cloud condensate path assimilation modules in the WRFDA assimilation system, the first all-weather adaptive partitioning assimilation system for geostationary satellite infrared data is constructed. Numerical experiments show that this method significantly improves the results for three heavy rainfall events during the 2021 Jiang-Huai Mei-yu period with the 24-hour forecast TS score improves by 0.12-0.15, precipitation location hit rate increases by 18.7%. This study provides a novel solution for all-sky assimilation of geostationary satellite infrared data and has significant application value in enhancing the predictability of convective systems.

Assimilation of all-sky radiance (ASR) is expected to offer additional advantages over clear-sky radiance (CSR). For instance, it can increase data coverage and the number of assimilated observations.
In this study, we modified and extended the method from a previous study (Okamoto et al. 2023), which assimilated all-sky infrared radiances from the geostationary (GEO) satellite Himawari-8, to include other GEO satellites: GOES-16, Meteosat Second Generation (MSG) -1, and -4.
A preliminary experiment in which Okamoto et al. (2023) was applied to assimilate ASRs from GOES-16 and MSGs demonstrated the problems with the method. The forecast accuracies significantly worsened after two-day forecast for each element and from initial forecast time for the temperature over the Sahara Desert and the Arabian Peninsula. One approach to address these issues was to use the sun zenith angle as a predictor for bias correction. The bias correction method applied in Okamoto et al. (2023) assumes minimal model bias. However, the limited coverage of GEO satellites and the presence of regional diurnal variations in model bias invalidate this assumption. Thus, incorporating a predictor that reflects local time variation effectively maintained the Gaussian distribution of first-guess departures. The assimilation experiment results also indicated that it improved prediction accuracy in the later forecast period. The degradation in the temperature field was observed where cloud effects between the model and observations are discrepant. To address this issue, we developed a method to inflate the observation error based on the magnitude of the discrepancy. This discrepancy value was also used in a new quality control procedure to discard the data when the value exceeded a threshold. These improvements allowed us to achieve a positive impact of ASR assimilation that exceeds the performance of CSR assimilation in the tropics, while also mitigating regional degradations.

Data assimilation (DA) of satellite visible reflectance has shown great potential to improve the forecasting skills of numerical weather prediction (NWP) models. One of the most commonly used forward operators for satellite visible images is the Radiative Transfer for TOVS (RTTOV) software package. RTTOV provided a wide choice of cloud optical parameterizations, which leads to difficulties in deciding which might be optimal for DA applications. To solve this problem, different cloud optical parameterizations in RTTOV were evaluated by comparing the observed and synthetic visible images. The observed images (O) were provided by FY-4B and Himawari-9, the latest geostationary meteorological satellite in China and Japan, respectively. Synthetic images (B) were generated by RTTOV configured with two different solvers, including the Discrete Ordinate Method (DOM) and the Method for FAst Satellite Image Synthesis (MFASIS). MFASIS is a fast emulator for DOM, but the former accounts for some three-dimensional radiative effects and is therefore more accurate than the latter. The inputs of atmospheric state variables to RTTOV were provided by the China Meteorological Administration Mesoscale (CMA-MESO) model. In terms of assessment indices of the domain-averaged biases and the root mean squared errors of the O-B departure, the optimal cloud optical parameterization consists of the “Deff” scheme for liquid water cloud and the “Baran 2014” scheme for ice cloud, with the effective radius of ice clouds parameterized by ambient temperature and ice water content. The optimal cloud optical parameterization was tested by a set of DA experiments for a real-world extratropical cyclone system. Positive impacts on the analyses and forecasts of the weather system were reported. Most importantly, the mean sea level pressure was better tuned for the optimal cloud optical parameterization. This study provided much-needed guidance for optimizing cloud optical parameterization in RTTOV for the DA of satellite visible reflectance. 

High-quality satellite quantitative precipitation estimation (QPE) is crucial for theoretical studies and disaster monitoring. However, it remains unclear which information is effective or relatively less valuable. Accurately eliminating ineffective variables and applying effective ones as predictors can further enhance the accuracy and computational efficiency for QPE. In this study, an hourly QPE algorithm was developed using three machine learning (ML) models, including Random Forest, XGBoost and LightGBM. We focused on obtaining high-precision precipitation estimations and further analyzing the contribution of different input variables. Sensitivity experiments revealed that satellite visible channels and cloud properties are key factors for accurate QPE. In contrast, information provided solely by infrared channels and meteorological variables is relatively limited. Among three ML models, LightGBM achieved the best QPE, and was comparable to, or even slightly better than GPM IMERG, which may be attributed to its incorporation of more effective variables and training with ground rain gauge. However, it sometimes underestimates heavy precipitation compared to GPM IMERG, probably due to few training samples and saturation of satellite spectral signals. The analysis of Shapley Additive Explanations (SHAP) indicates that QPE are more sensitive to cloud properties (e.g., cloud water path), but some meteorological factors, such as relative humidity at different pressure levels are becoming more important as the environment becomes drier. Additionally, the performance of ML model and GPM IMERG deeply relies on cloud type. These findings are expected to provide valuable references for the construction of future satellite QPE algorithms in terms of feature selection and data processing.

Merging multiple precipitation datasets without relying on ground-based gauges has proven to be an effective strategy for improving estimation accuracy. However, conventional merging techniques often assume a single statistical distribution for both rainy and non-rainy conditions, leading to two major challenges: (i) uncertainty in distinguishing between rain and no-rain events and (ii) biased merging weights, which result in less accurate precipitation estimates. This study introduces a novel two-step merging framework, Generalised Signal-to-Noise Ratio Optimisation (G-SNR), to overcome these limitations, with a demonstration over the complex and data-scarce terrain of High Mountain Asia (HMA). In the first step, the Categorical Triple Collocation-Merging (CTC-M) method improves rain/no-rain classification, yielding a 15% increase in the Heidke Skill Score (HSS) and a Critical Success Index (CSI) of 0.48 compared to the parent datasets ERA5, SM2RAIN, and IMERG-Late. The second step employs Signal-to-Noise Ratio Optimisation (SNR-opt) to enhance precipitation magnitude estimates, achieving up to a 30% reduction in unbiased root-mean-square error (ubRMSE) relative to conventional SNR-based methods that do not differentiate between mixed distributions. Importantly, G-SNR also refines bias correction for extreme precipitation events, reducing bias by 20% at the 99th percentile. These results highlight the potential of G-SNR as a reliable framework for improving precipitation estimation in high-altitude regions, where data variability is significant and gauge observations are scarce.

Satellite-based precipitation estimation plays a crucial role in meteorological and hydrological applications, particularly in complex terrains and oceanic regions where ground-based observations are scarce or unavailable. Deep learning techniques have recently gained prominence in satellite precipitation estimation due to their capability of representing complex nonlinear relationships. However, existing deep learning methods are usually trained on gridded precipitation observation data interpolated with ground station measurements, which inevitably introduces extra errors. This study reframes the problem of estimating precipitation from satellite infrared data as an operator mapping problem between two function spaces. The multi-spectral infrared data from satellites and the ground-based rain gauge data are treated as samples from two distinct function spaces. Thus, estimating precipitation from satellite infrared data becomes equivalent to learning an operator that maps the source function space to the target function space. This approach avoids the errors associated with data alignment and enables resolution-independent rainfall estimation. To efficiently characterize the source function space and accelerate model convergence, the Representation-Enhanced Operator Network (RE-ONet) is proposed. RE-ONet employs a 3D fully-convolutional network to extract embedded features from multi-spectral infrared data before feeding them into the operator network, significantly reducing the number of parameters needed. Furthermore, autoencoder-based representation learning enhances the representational capacity of these embedded features. The proposed method was applied to the Sichuan Basin using data from the FY-4B satellite. Verification results show that the precipitation intensity estimation error significantly reduced, compared to the operational FY-4B precipitation product. The proposed model also outperforms a baseline deep learning model U-Net, providing better characterization of the spatiotemporal variability of precipitation fields.

Accurate satellite-based retrieval of boundary-layer water vapor over land is crucial for understanding the Earth-atmosphere system but remains challenging. This is because surface radiation can interfere with accurate readings of atmospheric conditions such as temperature and humidity. This study proposes a novel approach to explicitly extract the upwelling atmospheric radiance () from the total radiance () observed by the Infrared Atmospheric Sounding Interferometer (IASI), using a Residual Multi-Layer Perceptron model. A modified one-dimensional variational algorithm for surface-free radiances is also developed. The radiance extraction model, trained on simulated IASI radiances, is applied to IASI observations over mainland Australia in January and February of 2022. Validated against ERA5 and radiosonde observations, compared with the traditional -based method, the -based atmospheric profile retrievals show distinct improvements in boundary-layer humidity retrieval with at least 20% error reduction. This approach provides a new thought to enhance humidity retrievals from hyperspectral sounders and benefits other quantitative applications such as data assimilation.

Stratospheric Sudden Warming (SSW) is the most dramatic event in the winter polar stratosphere, characterized by a rapid temperature increase and a reversal of climatological westerly to easterly. Due to its significant and long-lasting impact on surface climate, SSW and its amplification at the surface has been extensively studied. However, the mechanism underlying the surface influence of SSW remains unclear. In this study, we show that the SSW-induced tropospheric geopotential increase, which is amplified at the surface, is primarily caused by an increase in Arctic surface pressure. The surface pressure increase is induced by a poleward mass flux across the Arctic circle, which is determined by the poleward propagation of planetary-scale waves in response to the mean state change in the lowermost stratosphere during SSW. This result suggests that planetary-scale waves are critical not only for the onset of SSW, but also for its amplification at the surface right after SSW onset.

To better understand the role of mean stratospheric circulation in Sudden Stratospheric Warmings (SSWs) and their associated anomalous signals, we conducted idealized simulations using the CESM Eulerian Spectral-transform Dry Dynamical Core. This simple AGCM is run by dry hydrostatic primitive equations with zonally symmetric boundary conditions and analytically specified physics. Following the approach of Polvani and Kushner (2002), we modified the dynamical core to ensure that the model adequately resolves stratospheric dynamics and produced realistic stratospheric variability. By adjusting parameters that define the equilibrium temperature field in the stratosphere, we controlled the strength of the stratospheric polar vortex (SPV), allowing us to steer the model circulation toward the desired state. These simulations were designed to quantify how the strength of the mean stratospheric circulation influences troposphere-stratosphere coupling and the simulation of SSWs within a controlled dynamical environment. Through statistical analysis and various dynamical diagnostics on the long- term simulations, we examined differences in stratosphere-troposphere coupling under varying strengths of the SPV. We found that in simulations with a stronger SPV, the long-term mean of the zonal-mean flow increased, accompanied by enhanced variability in the stratospheric wind field. Under such conditions, the frequency of stratospheric circulation breakdown decreased; however, when they did occur, they were associated with more intense fluctuations and larger deviations from the mean state. Additionally, in stronger-SPV simulation, we observed that, on long-term average, the upward EP flux entering the stratosphere from the tropopause increased, leading to greater flux convergence within the stratosphere.

The downward impact of sudden stratospheric warming events (SSWs) on the troposphere is still controversial. Using the reanalysis data, 52 SSWs are identified over the period from 1940–2022, and 33 downward-propagating (DW) SSWs with noticeable impacts on the troposphere are selected with the remaining 19 SSW non-downward-propagating (NDW). The DW events are further classified into three types that are followed by cold surges over the Eurasia (EA), over the North America (NA), and over both (BOTH), respectively. Although the stratospheric polar vortex is weakened and deformed for both DWs and NDWs, the formers are stronger and lead to more significant negative Northern Annular Mode (NAM) and North Atlantic Oscillation (NAO) response in the troposphere. For DWs, the anomalous high develops in the polar region, which deflects to lower latitudes, consistent with the frequent appearance of the polar high and the midlatitude blockings. The shape of the anomalous polar high varies with the DWs type, and the extension and deflection of the anomalous high lead to different surface responses. The DWs are also accompanied by a southward shift of the precipitation belt especially over the oceanic and coastal regions. The NDW SSWs show relatively weaker impact on the troposphere, which is primarily related to the weaker amplitude of the stratospheric disturbance. The differences among three types of DWs include diverse NAM structures in the stratosphere, various spatiotemporal evolutions of the NAO pattern in the sea level pressure, different forcing by planetary waves, and varying number ratios between displacement and split.

On 25 September 2002, the first major sudden stratospheric warming event (SSW02) was recorded in the Southern Hemisphere (SH) since routine upper-atmosphere observations commenced in 1957. This SSW event was marked by the splitting of the polar vortex, a phenomenon rarely observed even in the Northern Hemisphere. While previous studies primarily examined the role of tropospheric waves and vortex preconditioning in directing these waves toward the polar stratosphere, the role of in situ-excited planetary waves (PWs) remains unexplored. The current study addresses this gap by examining the impact of in situ-generated PWs on the vortex splitting during SSW02. As the onset neared, the displaced polar vortex elongated and ultimately split into two vortices. The explosive amplification of zonal wavenumber (ZWN) 2 PWs (PW2) at 10 hPa, which split the vortex, was not only attributed to upward-propagating PW2 from the lower stratosphere but also by westward-propagating PW2 excited in situ in the mid-to-upper stratosphere, which then descended to 10 hPa. The spontaneous generation of westward PW2 in the stratosphere was induced by barotropic–baroclinic instability, triggered as the stratosphere became dominated by easterlies descending from the lower mesosphere. An anomalous poleward shift of the polar vortex facilitated the development of easterly and vortex destabilization by confining ZWN1 PWs (PW1) into the polar stratosphere, where they deposited strong westward momentum. Instability amplified PW2 through two mechanisms: (1) the breaking of PW1 generated smaller-scale waves through energy cascading while inducing instability, which in turn amplified these waves, ultimately contributing to the local growth of PW2; and (2) over-reflection of upward-propagating PW2. While both mechanisms contributed to the enhancement of PW2, the latter became dominant as the onset approached.

Sudden stratospheric warming, as a violent dynamic phenomenon occurring in the winter hemisphere, has attracted extensive attention and been deeply studied. Recent research indicates that during sudden stratospheric warming, not only are the planetary wave perturbations in the middle and upper atmosphere of the winter hemisphere affected, but the planetary waves in the summer hemisphere also exhibit significant abnormal behavior, which is possibly due to the cross-equatorial coupling effect. Further analysis reveals that the cross-equatorial coupling effect triggered by the sudden stratospheric warming is also modulated by the stratospheric quasi-biennial oscillation. Specifically, when the quasi-biennial oscillation is in the eastward phase, the cross-equatorial coupling effect on planetary waves is more prominent.

In changing climate, the frequency of extreme wildfires has severely increased. Convective mechanisms uplift the smoke plumes of wildfires, bringing its content into the troposphere, and in some cases to the stratosphere. The problem was considered purely low-atmospherical until it was realized that, for some wildfires, smoke plumes can overshoot the tropopause and reach the stratosphere. While in the well mixed troposphere, there are removing mechanisms for the smoke particles brought, in the stratosphere the aerosols can stay for months, circulating around the globe. Next degree of alarm came when it was demonstrated that the effect on the atmosphere produced by some of these plumes is comparable to effects of moderate volcanic eruptions. We discuss two methods of detecting and characterising the wildfire plumes in the stratosphere, using the vertically resolved satellite data. We study the temperature's behavior during the plume's tropopause-crossing, plume's impact on distribution of biomass burning gazes, and calculate the masses of injected aerosols and gazes.

Anthropogenic exacerbation of soil moisture droughts has been widely investigated in single layers such as the surface or root zone. However, this exacerbation in their spatiotemporal evolution in terms of soil vertical structure remains unclear. Therefore, this report will focus on depth-dependent responses of soil moisture droughts to climate change.

Chiang Mai, Thailand, located in the northern highlands, is increasingly vulnerable to climate extremes, particularly prolonged droughts and intensifying wildfires. These interconnected hazards are driven by rising temperatures, declining precipitation, and soil moisture depletion, leading to heightened wildfire risks. Understanding these interactions is essential for effective climate risk management and disaster mitigation. This study applies a multi-indicator framework integrating climatic, hydrological, and remote sensing data to monitor, analyze, and predict extreme climate events in the region. The research aims to examine spatial and temporal trends of droughts and wildfires over the last decade using multivariate statistical analysis and geospatial techniques. The methodology employs multiple analytical approaches, including the Mann-Kendall Trend Test and Sen’s Slope Estimator to detect long-term climate trends, cross-correlation analysis to examine lagged relationships between droughts and wildfires, and the Generalized Extreme Value (GEV) Distribution to quantify the probability of extreme events. Principal Component Analysis (PCA) will help identify dominant climate variability patterns influencing drought-wildfire interactions, while remote sensing and GIS-based spatial analysis will delineate high-risk wildfire zones and forecast potential future hotspots. The results, to be presented at the conference, are expected to provide insights into historical and projected trends of climate extremes in Chiang Mai, identify high-risk wildfire-prone zones, and support the development of a climate monitoring framework for early warning systems, policy recommendations, and climate adaptation strategies. This research aims to enhance evidence-based decision-making for disaster risk reduction, resource management, and environmental conservation in northern Thailand.

Compound drought and heatwaves (CDHW) events are complex climate extremes influenced by global climate change. Estimating their comprehensive risk using univariate statistics is challenging. In this study, China was taken as a case study area, and a two-dimensional joint function was constructed based on the duration and severity of CDHWs to assess the joint occurrence probability of CDHWs under diverse scenarios. The results revealed that for China as a whole, the changes in the severity threshold had a greater impact on extreme CDHWs, and the joint occurrence probability of abnormal extreme CDHWs was more sensitive to the severity. The conditional probability rose more rapidly for short-term events (3 days) than for long-term events (5-7 days). When the duration exceeded 7 days and the severity exceeded different thresholds, the joint occurrence probability of CDHWs was within the range of 2% to 6%. Among the different regions, the duration and severity had varying impacts on the joint occurrence probability. The extreme CDHWs in North China, Northeast China, and western Northwest China were more strongly influenced by the duration. While Jianghuai, South China, Southwest China, eastern Northwest China, and the Tibetan Plateau were more prone to longer-lasting extreme CDHWs, and in these areas, when CDHWs lasting more than 7 days occur, changing the severity threshold had a greater impact on their extremity. In North China in 1997, the longest CDHW had joint return periods of over 1000 years when both duration and severity exceeded thresholds, and over 500 years when either exceeded thresholds. The findings demonstrate the variations in the impacts of duration and severity on the joint probability of CDHWs in different regions of China.

Floods and droughts are extreme hydrological events that significantly impact human society and the economy. These anomalies influence surface soil moisture (SSM), a key parameter in the water cycle that affects vegetation growth, surface-atmosphere interactions, and water resource management. Despite its importance, the role of SSM in extreme events remains understudied in Thailand. This study investigates daily SSM variations during flood and drought events in 2010 across ten provinces in Thailand, using the ERA5-Land reanalysis dataset. The analysis reveals that during the drought from late February to March, SSM remained consistently low (0.18–0.25) for more than a week due to prolonged dry conditions, primarily driven by the absence of precipitation. In contrast, the November flood in southern Thailand resulted from excessive rainfall on already saturated soil, preventing further infiltration and increasing surface runoff. By comparing SSM with precipitation data from the GSMaP gauge-calibrated dataset, we found that soil moisture responds rapidly to precipitation, reaching saturation within two days, while drying occurs over a longer period, typically taking weeks to a month depending on soil properties. Soil type also plays a crucial role in SSM dynamics. Areas with higher clay and silt content retain moisture longer, leading to regional variations in SSM despite similar precipitation patterns. Understanding these interactions can enhance preparedness and water resource management during extreme weather events in Thailand. For further in-depth analysis, future studies could explore the role of soil moisture in a broader context, including its interaction with other atmospheric variables like temperature, humidity, and wind speed during extreme events. Additionally, conducting high-resolution spatial and temporal studies with ground-based soil moisture measurements would improve the accuracy of reanalysis data and provide a more detailed understanding of SSM dynamics.

In recent years, frequent drought events have exacerbated forest mortality globally. Accurately identifying and quantifying the impact of drought on forest health and mortality distribution has become a critical issue in global ecological drought research. Existing remote sensing-based methods for forest mortality monitoring primarily rely on single meteorological datasets or vegetation indices, which, while effective in capturing external drought conditions and canopy optical responses, fail to comprehensively reflect the physiological changes within forests, resulting in insufficient accuracy in quantifying drought-induced forest mortality. To address this limitation, this study integrated key indicators, including meteorological and hydrological data, forest water content, greenness, and photosynthetic efficiency, and incorporated prior knowledge of drought-induced forest changes. We developed a global canopy mortality spatial prediction model by combining structural equation modeling (SEM) and random forest (RF) techniques. The results show that the model effectively identifies spatial patterns of drought-induced forest mortality, especially in hotspot areas. Specifically, the canopy mortality map generated by the hybrid model achieved an R² of 0.74, i.e., an 18% improvement over the RF model (R² = 0.63). Variables such as hydrometeorological conditions, forest water content, forest greenness, and forest photosynthetic efficiency were identified as highly important for predicting canopy mortality, explaining both their direct and indirect contributions to mortality. Among these, the relationship between forest water content and canopy mortality was the most prominent, showing a stronger influence compared to other factors. This finding highlights the potential value and applicability of satellite-derived vegetation water content (VWC) datasets in monitoring drought-induced forest mortality. This study provides an enhanced assessment approach that facilitates precise monitoring of drought-induced forest mortality, supporting ecosystem management under climate change.

Scientific assessments of precipitation extremes in response to climate change have become increasingly significant. However, conducting such assessments has always faced a dilemma. On the one hand, the widely-used storyline approach with pseudo-global-warming (PGW) experiment could capture precipitation extremes under climate change. Nevertheless, its capability of studying the impacts of multi-scale systems is restricted due to the limited-area simulation strategy. On the other hand, the global cloud-resolving simulation is an effective approach to study the performance of multi-scale systems under climate change. Yet, the current computing limitations pose a significant challenge conducting global ensemble experiments. Confronted with this dilemma. This study devised an innovative solution: The storyline approach with a novel global PGW experimental framework. This framework is grounded an advanced global model featuring variable-resolution modeling techniques coupled with data assimilation. by using the proposed framework with a 60-3 km variation-resolution global model, the researchers quantitatively unveiled the strong response of the 7•20 Henan precipitation extreme to global climate change. Moreover, they illustrated the impacts of multi-scale system interactions under climate change, spanning from the large-scale subtropical high and double typhoon down to the mesoscale convective system. This study not only provides a new approach to solve the long. standing dilemma in assessing the extreme events in response to climate change, but also opens avenues to study the impacts of climate change on extreme events from the perspective of multi- scale system interactions.

With growing need to fully understand the characteristics of daytime and nighttime heatwaves under the global warming, an overall investigation has been conducted comparing independent daytime, nighttime and day-night compound heatwaves over China. The findings reveal that the Yangtze–Huaihe River basin (YHRB) suffers from the most frequent compound heatwaves, which also exhibit a significant rise in both frequency and intensity during 1979-2020. Focused on the regional heatwaves over YHRB, independent daytime and compound heatwaves are mostly related to the atmospheric systems at middle and lower levels, including strengthened and westward Western North Pacific subtropical high (WNPSH) and the Pacific-Japan/East Asia – Pacific (PJ/EAP) pattern, which benefits the sunny weather and downdrafts over YHRB. On the other hand, independent nighttime and compound heatwaves are influenced by the synoptic systems at higher levels, including northward and cyclonic meandering East Asian subtropical jet (EASJ), strengthened and eastward South Asian high (SAH) and positive phase of circumglobal teleconnection (CGT) pattern, accompanied with cloudy weather and strengthened downward longwave radiation over north YHRB. Our findings would help improve the understanding of the mechanisms underlying various types of heatwaves and better respond to them.

The application of numerical weather prediction (NWP) models for estimating precipitation depths with extremely low exceedance probabilities, such as Probable Maximum Precipitation (PMP), has garnered significant attention due to its advantages. However, key challenges remain in this approach, including the incorporation of climate change effects and the probabilistic evaluation of estimates. This study addresses these challenges by employing a model-based approach to maximize extreme precipitation while considering the influence of climate change, with the ultimate goal of estimating PMP in the target watershed. Specifically, the study focuses on recent meso-scale convective systems, known as "Senjo-Kousuitai" in Japanese, which have occurred in northern Japan. The Weather Research and Forecasting (WRF) model (version 4.1.2) was utilized to simulate and enhance precipitation associated with these events over the Akagawa and Arakawa River Basins. The results demonstrate a successful amplification of precipitation within the basin, reaching depths corresponding to extremely low exceedance probabilities as inferred from radar-based observational records. Furthermore, this study successfully quantified the impact of climate change on maximum precipitation depths. A probabilistic assessment of the estimated maximum precipitation was conducted using atmospheric drivers of precipitation and a large-ensemble climate simulation database (d4PDF). Overall, the findings present a valuable framework for addressing critical challenges in model-based PMP estimation.

Climate change and human activities intensify extreme weather events, making trend analysis of temperature and precipitation vital for understanding water availability, hydrology, biodiversity, and livelihoods. This study aimed to quantify the interannual and seasonal variability in climate variables at the Beas River Basin in the Northwestern Himalayas for 43 years (1980-2023). Five meteorological stations were selected to observe trends using various statistical methods using the TerraClimate dataset with 4 km resolution. Applying the Mann-Kendall test, trend analysis was used to detect variations in precipitation and maximum and minimum temperatures across all the stations. The Pettitt test was applied to identify points of abrupt change. Results show that Tmax and Tmin have been increasing at a rate of 0.02°C and 0.04°C per year in the basin. A notable abrupt change was identified by Pettitt’s test in 1999 for temperature at five stations. The Long-Term Average (LTA) anomaly analysis for temperature showed a negative trend from 1980 to 2000 (ranging from 0 to -1.5°C), followed by an increase from 2001 to 2023 (ranging from 0 to +1.5°C). Rainfall trends showed no significant trend in all the stations. However, analysis revealed a decline in rainfall patterns post-2000, with slight upward trends in specific years. In terms of elevation, above 3000 meters, both Tmax and Tmin exhibited a more pronounced increase across all seasons, indicating greater warming at higher altitudes. In the JJAS season, areas above 3000 meters demonstrated a decreasing rainfall trend, while lower elevations showed greater variability across all seasons. The lack of local observatories highlighted the importance of using high-resolution datasets for accurate analysis. This study highlights the hydrological impacts of climate change in the Beas Basin, supporting water management, climate research, disaster planning, and stakeholder engagement for future resilience.

In recent decades, the Hindu Kush Himalaya (HKH) region has emerged as a hotspot for hydroclimate extremes, experiencing both more frequent intense precipitation events and droughts, largely linked to global warming. The highly complex orography and diverse climatic conditions of the HKH plays a crucial role in regional weather patterns, and there is considerable variability in projected changes to the hydroclimate of the region. Here, we use the Shared Socioeconomic Pathways (SSPs) emission scenarios of the IPCC derived by Coupled Model Intercomparison Project Phase 6  (CMIP6), to investigate future changes in hydroclimate extremes and examine the potential driving factors for those changes. Under a low-emission scenario (SSP1-2.6), the frequency of extreme rainfall is projected to increase over the Western and northeastern Indian region of the HKH. Additionally, under medium emission (SSP2-4.5), a notable rise in heavy rainfall intensity (14.3%) is observed in the major river basins of the HKH including the upper Ganga and Indus basins. Elsewhere in the HKH, the frequency of dry days examined was more concentrated over the extended part of the Karakoram of HKH region, spanning across Pakistan and India.  However the intensity of wet days dominated over the central part of the Himalayan region. Additionally, the mean local Hadley cell, acting as the primary driver, expands to 33.5° in the near future (2021–2040), indicating drier conditions. In contrast, its narrowing in the mid (2041–2070) to 25.6° and far future (2081–2100) to 27.7° suggests a transition toward wetter conditions under all SSPs. These findings emphasize the urgent need for a deeper understanding of global warming impacts and the development of adaptive strategies to enhance resilience and ensure sustainable socio-economic and environmental outcomes in the HKH.

The current study analyses the influence of Sea Surface Temperature (SST) in the simulation of Super Cyclonic Storm Amphan over the Bay of Bengal region using the Advanced Research Weather Research and Forecasting (WRF-ARW) Model. Experiments without SST (CNTL) and with SST (SST) were conducted to simulate the Amphan over a cloud-resolving scale with a horizontal resolution of 1.667 km. The study employed three two-way interactive nested domains with horizontal resolutions of 15 km, 5 km, and 1.667 km, respectively, and the innermost domains were configured on a moving nested platform. The initial and boundary conditions for the model simulations were obtained from the NCEP FNL analysis data, and the time-varying SST data were taken from the fifth generation of the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis. The model was executed for 96 hours from 00 UTC on 17 May to 00 UTC on 21 May, 2020. The estimated track, Maximum Surface Wind (MSW), and Minimum Sea Level Pressure (MSLP) were compared with India Meteorological Department data, Doppler Weather Radar (for maximum reflectivity), India Meteorological Department - National Centre for Medium Range Weather Forecasting (IMD-NCMRWF) data (for accumulated rainfall). The results indicate that the SST experiment has no effect on the track of the cyclone. The SST experiment precisely evaluated cyclone intensity with an improvement of about 35% in MSW and 60% in MSLP. Results demonstrate that SST experiment provide better estimates of cyclone structures in terms of maximum reflectivity, and accumulated rainfall. The study concluded that the SST-updated experiment using the WRF-ARW model on a cloud-resolving scale provided improved results in the intensity and structure of the Cyclone. Additional studies were conducted on a larger number of cyclone cases, and the results indicated that SST simulations provide a more accurate estimate.

Emission of anthropogenic greenhouse gases has resulted in greater Arctic warming compared to global warming, known as Arctic amplification (AA). From an energy‐balance perspective, the current Arctic climate is in radiative‐advective equilibrium (RAE) regime, in which radiative cooling is balanced by advective heat flux convergence. Exploiting a suite of climate model simulations with varying carbon dioxide (CO2) concentrations, we link the northern high‐latitude regime variation and transition to AA. The dominance of RAE regime in northern high‐latitudes under CO2 reduction relates to stronger AA, whereas the RAE regime transition to non‐RAE regime under CO2 increase corresponds to a weaker AA. Examinations on the spatial and seasonal structures reveal that lapse‐rate and sea‐ice processes are crucial mechanisms. Our findings suggest that if CO2 concentration continues to rise, the Arctic could transition into a non‐RAE regime accompanied with a weaker AA.

In this study, the performance of 24 Coupled Model Intercomparison Project Phase 6 (CMIP6) models in simulating the dynamic processes of Arctic sea ice concentration (SIC)- and El Niño-Southern Oscillation (ENSO)- forced teleconnection during winter is subjectively and objectively evaluated. The Arctic SIC-forced teleconnection is associated with a warm Arctic-cold Eurasian pattern of surface temperature (T2m), a low Arctic-high Eurasian pattern of sea level pressure (SLP), and a southeastward propagating wave-train originating from Arctic in the upper troposphere. The ENSO-forced teleconnection is associated with a poleward propagating wave-train originating from tropical Pacific in the upper troposphere, a low North Pacific-high Arctic pattern of SLP, and a cold North Pacific-warm Greenland pattern of T2m. The metrics of Taylor skill scores and Distance between indices of simulation and observation (DISO) are used to objectively and quantitatively evaluate the performance of models. The results of subjective and objective evaluation are essentially consistent. The CanESM5, MPI-ESM1-2-HR, EC-Earth3, and MRI-ESM2-0 models have the best performance in simulating the Arctic SIC-forced teleconnection. The CESM2, ACCESS-CM2, NESM3, NorESM2-MM, CAS-ESM2-0, MRI-ESM2-0 models have the best performance in simulating the ENSO-forced teleconnection. The two best-performing multi-model ensembles well reproduce the dynamic processes of the Arctic SIC- and ENSO- forced teleconnection. The diversity of model performance is attributed to the different skills of different models in simulating the interannual variability of Arctic SIC, the anomalous deep warm high over the Barents-Kara Seas, the interannual variability of tropical Pacific SSTs, and the wave number of poleward propagating Rossby waves.

In the past several decades, summer Arctic sea ice has been experiencing dramatic decline.Imposed on the significant linear decling trend cuased by anthropogenic forcing, Arctic sea ice also shows an obvious decadal variations. Attributions of the Arctic sea ice variations to the tropical SST forcing are well established especially in wintertime, however, the linkage between tropics and Arctic in summertime is inherently uncertain and unstable. On the other hand, the summertime Arctic sea ice decline is not spatially uniform but varies significantly with locations. By separating the externally forced and internally generated sea ice variations, we find that there is a strong linkage between tropical pacific meridional mode (TPMM) and Arctic sea ice tripole mode on quasi decadal timescales. The influece of TPMM on Arctic is mainly through the Atlantic pathway: First, the diabatic heating of TMPP favors a teleconnection pattern resembling the NAO in the northe Atlantic, the circulation pattern would force an NAT like SST pattern in the North Atlantic. Secondly, the meridional gradient of SSTAs plays an important role in the formation of Rossby wave propagation originating from the NA along the Polar front jet, forming a zonal wavenumer-3 pattern in the mid-high latitudes, and leads to the tripole mode of Arcric sea ice decline through thermodynamical effects. This finding can potentially contribute to improving the skill of predicting Arctic sea ice.

Under global warming, Arctic sea ice has diminished at an alarming rate, profoundly impacting the global climate system. September 2012 witnessed a record-breaking minimum in Arctic sea ice extent, with August exhibiting unprecedented rapid ice loss. Satellite observations and reanalysis data reveal that the Arctic experienced an intense Beaufort Sea storm in August 2012, which disrupted local atmospheric and oceanic systems, causing widespread sea ice melt and drift. Although existing studies have largely attributed this event to local weather and climatic factors, the role of remote drivers has remained unexplored. This study, however, sheds light on a significant yet overlooked remote influence. It reveals that the extreme Beaufort Sea cyclone was amplified by a preceding intense convective event over the western North Pacific. The tropical convection triggered a Rossby wave train that propagated along a great - circle path three to five days prior to the mature of the cyclone. This wave train resonated with the Arctic cyclone over the Chukchi - Beaufort Seas, significantly enhancing its intensity. Concurrently, the North Pacific Oscillation (NPO) pattern established a conducive large-scale circulation background. The NPO intensified the polar front jet, thereby enhancing the meridional potential vorticity gradient and high-latitude baroclinic instability. This pattern not only facilitated the genesis of the cyclone but also promoted the poleward propagation of Rossby waves from tropical convection. Our findings underscore the critical yet overlooked remote influence of North Pacific convection on Arctic sea ice extremes, offering novel insights into tropical-polar interactions and advancing the mechanistic understanding of extreme climate events for improved climate modeling and prediction.

Winter warming over the Barents-Kara Seas (BKS) has received extensive attention over the past two decades because it is closely associated with Arctic sea ice loss, winter Eurasian cooling, and extreme cold events over East Asia. However, the role of mid-latitude atmospheric circulation anomalies in resulting winter BKS warming is unclear. This study investigates the relationship between autumn (October-November) East Asian Trough (EAT) and the BKS warming in the subsequent winter (December-February) for the period 1979-2022. The result shows that when the autumn EAT weakens, warming, increased moisture, and sea ice loss are observed in the BKS during winter. The weakened EAT promotes increased sea surface temperatures (SSTs) in the mid-latitude North Pacific through increasing solar radiation and reducing cold air activities. Then, the positive SST anomalies persist into winter. These continuous warm SSTs from autumn to winter trigger winter Rossby waves downstream, which favors the occurrence of combination of a positive phase of the North Atlantic Oscillation (NAO)-like and high pressure over the Ural region. This combined circulation pattern facilitates the transport of warm and moist air over the North Atlantic, as well as the advection of warm air from lower latitudes, leading to warming in the BKS region.

Supercell tornado formation requires both intensification (via stretching) and organization (into a symmetric vortex) of vertical vorticity maxima during the transition to a tornado-characteristic flow from surface vorticity patches (Fischer et al., 2024). The final and crucial step in the tornadogenesis process is the precise spatiotemporal alignment of surface vorticity with the low-level mesocyclone. However, due to limitations in spatiotemporal resolution, operational S-band radars often fail to capture the formation process of surface vertical vorticity, which is smaller than mesocyclones and occurs at lower altitudes (below 200 m). The Shanghai X-band phased array radar (PAR) network has proven to be highly effective in capturing the structure, evolution, and dynamics of local severe storms. With volume scans conducted every 30-60 seconds and a radial resolution of 30 meters, the PAR provides far more detailed insights than S-band radars.Using high-resolution PAR data, this study investigates several tornado cases in the Yangtze River Delta, including both classical supercell and mini-supercell tornadoes. In one autumn case, as the rear-flank downdraft (RFD) surge merged into the hook echo nose region, a descending reflectivity core (DRC) emerged. A micro-scale vortex, with a diameter of less than 2 km, eventually formed between the DRC and the updraft motion within the weak echo region (WER). The surface micro-vortex was initially located approximately 3-5 km southeast of the mesocyclone at an altitude of 4.5 km. As a result, the vorticity structure extending from the surface to the 4.5 km-high mesocyclone formed a vortex tube that tilted upward from southeast to northwest. Subsequent continuous PAR observations revealed that the surface vorticity gradually moved beneath the low-level mesocyclone, leading to the contraction and stretching of the vortices, which rapidly intensified near the surface. This behavior suggests a potential precursor to tornadogenesis.

Tornadoes are local, small-scale, sudden and severe convective weather with a short period of occurrence and disappearance. They are one of the most violent and destructive weather phenomena in the Earth's atmosphere. At present, the monitoring and early warning of tornadoes mainly rely on Doppler weather radar observations, and early warning and forecasting are achieved by detecting the velocity characteristics of the parent storm and tornado. In China, the new generation of Doppler weather radars are widely used in tornado monitoring and identification tasks, and the dual polarization update technology has significantly improved the prediction and forecasting capabilities of tornadoes. This paper proposes an innovative tornado detection algorithm based on convolutional neural networks, which is deeply optimized for the characteristics of dual-polarization radar data. The specific improvement is that we have constructed a six-channel data processing network, integrating multi-dimensional polarization parameters such as velocity, reflectivity, velocity spectrum width, differential reflectivity and correlation coefficient, so that the network can capture the characteristics of tornadoes more comprehensively. At the same time, the algorithm is trained and tested using a self-built tornado dataset, and the algorithm performance is tested and evaluated on actual tornado cases. In addition, the algorithm combines artificial intelligence-assisted learning to optimize traditional algorithms based on tornado vortex signatures (TVS) and tornado debris signatures (TDS), further improving detection results.

In collaboration with 22 institutions, including the Guangdong Meteorological Bureau, Peking University, the Meteorological Observation Center of the China Meteorological Administration (CMA), Nanjing University, Sun Yat-sen University, the Hong Kong Observatory, and the Beijing Institute of Urban Meteorology, the Tornadic WInd STorm ExpeRiment in the Greater Bay Area (TWISTER) was launched by the CMA Tornado Key Laboratory. The experiment aims to capture the near-storm environments of supercells and tornadoes prior to their formation, as well as the thermodynamic and microphysical structures following storm development, over complex underly surface condition in the background of subtropical monsoon and typhoons.The first phase of the experiment (TWISTER-I) took place from June to August 2024 in the Guangdong-Hong Kong-Macao Greater Bay Area. Leveraging a dense modern operational radar network in the Greater Bay Area (including 12 S-band dual-polarization radars and 54 X-band phased array radars), along with a C-band radar at Baiyun Airport, TWISTER-I deployed additional observational systems. These include two C-band dual-polarization radars on mobile platforms, six microwave radiometers, two temperature and humidity-Raman lidar systems, four Doppler lidar wind profilers, and two vehicle-based wind profiling radars in central Guangdong, a region with the highest frequency of supercell and tornado activity since 2000.During the experiment, the Guangdong 3-km mesoscale operational model, a 1-km all-sky satellite-based WRF-EnKF ensemble data assimilation and forecast system, the 1-km WRF deterministic forecast system, and the Beijing Urban Meteorology Institute's 43-m high-resolution ensemble 4D-VAR assimilation system, an AI-based Tornado Identification System were simultaneously run in real time to evaluate tornado forecasting methods. During TWISTER-I, valuable meteorological data was collected, including one land tornado, nine coastal waterspouts, 15 supercells, two squall lines, 17 mesocyclones, three downbursts, and 17 tornado vortex signatures (TVS). As a follow-up, TWISTER-II will be launched in the spring and summer of 2025.

Tornadoes are highly destructive weather events, and their timely detection is crucial for mitigating damage and saving lives. Doppler weather radar serves as the primary operation tool for tornado detection. However, existing detection methods mainly struggle with high false alarms. In addition, the potential mismatch between tornado reports and the corresponding radar signatures further compromise the detection accuracy. To tackle these issues, we propose TorDet, a novel two-stage deep learning approach designed for tornado detection using Doppler weather radar data. TorDet introduces a quantitative tornado labeling scheme that integrates the prominence of tornado-related radar features and tornado reports, explicitly categorizing samples (TOR, WRN, WEK, and NUL) to reduce ambiguity in both manual labeling and model training. TorDet's two-stage design divides the detection process into distinct yet complementary tasks. In the Detection stage, it generates lowresolution feature maps to identify potential tornado-related regions by capturing large-scale information. The Positioning stage refines these detections, accurately locating tornado centers and classifying their categories. To make better use of limited data, we introduce deep supervision at intermediate layers in both stages, improving performance by enhancing the loss function and enriching gradient for backpropagation.We conduct experiments using radar data collected between 2017 and 2024. Experimental results demonstrate that TorDet, benefiting from its unique design and training strategy, outperforms existing models by significantly reducing the False Alarm Ratio (FAR) and improving the Critical Success Index (CSI), while maintaining a high Probability of Detection (POD).

The Greater Bay Area, recognized as a region with a high frequency of tornado occurrences in South China, has established the world's largest X-band polarimetric phased array radar (PAR) network. The PARs show great potential in the surveillance of fast-evolving tornado storms. Through an analysis of the EF2 tornado event on June 16, 2022, this study presents the PAR network's superior performance in detecting the tornado vortex and capturing fine horizontal and vertical structures (TVS, TDS, ZDR columns, etc.) of convective storms when compared to nearby S-band operational radars. Furthermore, the assimilation of supplementary PAR data using the Ensemble Kalman Filter (EnKF) has successfully simulated discernible vortex circulation evolution for tornadic storms. The study identifies the primary mechanisms driving tornado formation and intensification as the tilting of low-level horizontal vorticity and the stretching of vertical vorticity. Based on the accurately simulated fine-scale tornadic storm vortex derived from the EnKF analysis field through the assimilation of PAR observations, the deterministic forecast can produce the vortex circulation of the tornadic storm about six minutes in advance of the tornado touchdown, which is not realized by assimilating S-band polarimetric radars only. With the operational implementation of the PAR network and its application in data assimilation systems, significant advancements are being made in both the scientific understanding and nowcasting of tornadoes in South China.

China faces the Northwest Pacific with the world’s most active tropical cyclones (TCs). Under global warming, whether and how the warming of the “Roof of the World”, the Tibetan Plateau (TP), influences the environmental fields and precipitation of landfalling tropical cyclones (LTCs) in China remains unclear. This study examines the evolution of the precipitation rate of LTCs over 43 years and the driving factors. A data-driven objective classification reveals that the TP warming drives interactions between the South Asian High (SAH) and the Western Pacific Subtropical High (WPSH), forming the key environmental fields influencing LTC precipitation in China. Under global warming, the precipitation rate of LTCs over China exhibits an overall increasing trend. However, the TP warming exerts both positive and negative effects on precipitation rate of LTCs under different environmental fields.

Typhoon activity in the Northwest Pacific region is most active during the summer and autumn. In summer, typhoon movement is primarily guided by the circulation of the Northwest Pacific Subtropical High, steering them westward toward Taiwan or northeastward toward Japan. In autumn, as the subtropical high weakens and retreats eastward, the guiding airflow becomes less distinct. Typhoons tend to move toward Japan or Korea, follow irregular paths, and are more likely to interact with high-latitude weather systems.This study primarily utilizes typhoon data from the Central Weather Administration’s typhoon database to analyze the main trajectory categories of typhoons making landfall in Taiwan during the summer and autumn months. It also examines the genesis locations of typhoons for each trajectory category and their spatial relationship with the Northwest Pacific Subtropical High.Additionally, the study explores the relationship between typhoons affecting Taiwan in these two seasons and investigates its connection to the interannual variations of large-scale sea surface temperatures in the tropical central and eastern Pacific.

One hot spot for the genesis of cyclonic systems is the Southern Indian Ocean is to the south of Indonesia. These cyclonic systems could lead to significant impact to Indonesia due to their potential destructiveness. Consequently, a better understanding of their climatological spatial characteristics and variability could provide usable information for disaster risk reduction and mitigation purposes. To achieve this, we firstly apply an objective cyclone tracking algorithm, namely the University of Melbourne cyclone detection and tracking algorithm. The event identification is based on reanalysis (1975-2023) to get a robust understanding of the climatological distribution and variability of the entirety of cyclonic systems, from weak, shallow disturbances to extreme tropical cyclones. The results of the analysis show that cyclonic systems generally occurred in the period from December to April with a maximum frequency in the southwest of Sumatra. Large-scale climate variability modes, such as El Niño-Southern Oscillation (ENSO), are shown to have a significant impact on the presence of cyclones and their precursors in the IMC. There are twice of the number of cyclonic systems identified in non-El Niño phase in comparison to the El Niño phase. In the same period, the highest activity of cyclonic systems is observed during MJO phases 4 and 5 (the maritime continent) (from phase 1 to 8) with 37% of all system passage. Furthermore, the relationship between equatorial waves such as Convectively Coupled Kelvin Waves (CCKW) and cyclonic systems is also investigated. 

Despite the concerns about the damage caused by autumnal tropical storms, their risk has not been quantitatively defined, and the geographical distribution of this risk remains unclear. This study uses a localized index of tropical storm activity and attempts to examine the geographical risk patterns of the autumnal tropical storms that local residents observe in the East Asian region. The regional risks are effectively captured by examining the differences in localized storm activity indices between August and September. This study shows that the risk of autumnal storms increases over most of East Asia, and the seasonal response appears the highest around the Ryukyu Islands. In particular, Japan and Vietnam experience increased tropical storm risks due to the higher number of stronger and slower storm cases. On the other hand, in most coastal regions of China, Taiwan, the Ryukyu Islands, the Korean Peninsula, and the Philippines, increased risks by the intensified storms offset the reduced storm risks from the lower number of cases. The findings in this study are expected to help local residents make more informed and efficient decisions to mitigate the damages from the autumnal tropical storms.

Biomass-burning organic aerosol(s) (BBOA) are rich in brown carbon (BrC), which significantly absorbs solar irradiation and potentially accelerates global warming. Despite its importance, the multiphase photochemistry of BBOA after light absorption remains poorly understood due to challenges in determining the oxidant concentrations and the reaction kinetics within aerosol particles. In this study, we explored the photochemical reactivity of BBOA particles in multiphase S(IV) oxidation to sulfate. We found that sulfate formation in BBOA particles is predominantly driven by photosensitization involving the triplet excited states (³BBOA^*) instead of iron, nitrate, and S(IV) photochemistry. Rates in BBOA particles are three orders of magnitude higher than those observed in the bulk solution, primarily due to the fast interfacial reactions. Our results highlight that the chemistry of ³BBOA^* in particles can greatly contribute to the formation of sulfate, as an example of the secondary pollutants. Photosensitization of BBOA will likely become increasingly crucial due to the intensified global wildfires.

Intermediate volatility organic compounds can contribute to secondary organic aerosol formation, especially in urban areas. The kinetics rates and aerosol yield of oxidation of IVOCs are useful to evaluate the contribution in the ambient environment. In our lab research, we measured the rate constants of different structures of IVOCs with OH radicals and Cl atoms at room temperature, the gas phase mechanism was proposed according to the gas phase products and quantum calculations. In the meanwhile, the particle phase products were measured and analyzed. Also, temperature is another key factor in secondary aerosol formation, which would change in a wide range over the world.

Isoprene-derived secondary organic aerosol (iSOA) represents one of the most abundant biogenic sources of atmospheric organic particulate matter and its formation is profoundly influenced by anthropogenic emissions. However, the long-term measurement of iSOA, particularly in polluted urban regions, remains scarce, hindering the quantitative understanding of the anthropogenic influence on iSOA formation in the atmosphere. Here we present a field and modelling combined study of iSOA in atmospheric fine particle matter (PM_2.5) in urban Shanghai, China in three summers and winters during 2015-2021, aiming to understand the response of iSOA formation to the strong anthropogenic emission reductions over this period. The particulate concentrations of iSOA species formed by reactive uptake of different epoxide intermediates, such as isoprene epoxydiol (IEPOX), hydroxymethylmethyl-α-lactone (HMML), and/or methacrylic acid epoxide (MAE), on aqueous aerosols were measured by different mass spectrometric techniques. The complementary simulations were also performed with the Community Multiscale Air Quality (CMAQ) model and the results are compared to the measurements. The inter-annual and seasonal variations of IEPOX-SOA and HMML/MAE-SOA as well as the major driving factors are characterized. The pathway-specific response mechanisms of iSOA formation to anthropogenic emission reductions are identified. Implications for the mitigation of biogenic SOA formation through anthropogenic emission controls are discussed.

Viscosity is a key property of organic aerosols (OA) that influences the rates of heterogeneous and multiphase reactions, photochemistry, and the uptake of gaseous pollutants. However, the phase state of organic aerosols is often not considered in current chemical transport models (CTMs). We previously developed a method to estimate the glass transition temperature (T_g) of an organic compound based on its volatility, which has been successfully applied to field observations of volatility distributions to predict viscosity (Li et al., 2020). In this study, we apply this method to predict the phase state and viscosities of secondary organic aerosols (SOA) across China using the Weather Research and Forecasting model coupled to Chemistry (WRF-Chem). Simulations show that SOA particles are primarily liquid or have low viscosity (<10⁴ Pa s) in the southeast, semi-solid (viscosity ranging from 10⁵ to 10⁸ Pa s) in central and northeastern China, and highly viscous or amorphous solid in the northwest. The phase state of SOA particles is influenced by ambient conditions and the particle chemical composition. High relative humidity (RH > 60 %) is the main factor determining the phase state of SOA particles in the planetary boundary layer. When RH is below approximately 60 %, predicted viscosity is influenced by RH, temperature, and also chemical composition.  In addition, we calculate the bulk diffusion coefficient, an important parameter for determining mass transport and mixing rates, using the fractional Stokes-Einstein relation. We demonstrate how the viscosity of an organic shell impacts atmospheric multiphase chemistry  by examining the uptake of isoprene epoxydiol (IEPOX) and the hydrolysis of N₂O₅.

The frequency and intensity of fires have increased globally over the past several decades and are projected to continue rising in the future. The emissions from biomass burning present a significant threat to air quality. During the Fire Influence on Regional to Global Environments and Air Quality (FIREX-AQ) aircraft study, the chemical composition of fire-emitted submicron particles was quantified with an AMS. Analysis of particles from wildfires in the western US revealed a consistent composition dominated by OA across plumes. In contrast, particles from agricultural fires sampled in the eastern US exhibited greater variability, with a higher proportion of inorganic species, notably Cl and K. Rapid measurements of K in fire plumes, conducted at frequencies up to 5 Hz, demonstrated strong correlation with concurrent IC filter measurements, allowing a quantitative closure of the particle ion balance. While laboratory experiments indicated potential variability in the AMS instrumental response to K inorganic salts, field observations suggested a consistent response for freshly emitted particles from fires predominantly composed of OA. Ultrahigh-resolution analysis of filter samples allowed for the identification of organosulfur species, including organosulfonates and organosulfates, in some fresh biomass burning plumes, where they made significant contributions to AMS sulfate levels compared to the typically low percentages found in regional background air. The agreement between AMS inorganic-only sulfate and IC sulfate was found. Furthermore, thermodynamic models were employed to estimate aerosol pH, a crucial parameter influencing various particulate physical and chemical processes, revealing that freshly emitted submicron particles from western biomass burning had near-neutral pH levels (averaging ~6-7), likely due to strong NH₃ buffering. This contrasts with regional background particles, which exhibited moderate acidity (pH ~2-3), and some fairly-aged sulfate particles with even lower pH values (pH ~1). These findings underscore the significant chemical distinctions between fire-emitted plumes and the regional background aerosols.

The liquid-liquid phase separation (LLPS) of atmospheric aerosols is a pivotal process that profoundly influences their physicochemical properties, including chemical reactivity, cloud condensation nuclei and ice nuclei activity, as well as optical effects. Prior to LLPS, a liquid phase transition occurs, accompanied by complex chemical reactions. However, capturing the in-situ chemical changes during this period presents a significant challenge. Recently, an aerosol optical tweezer (AOT) has been employed to investigate the key factors affecting LLPS and the chemical transformations during the phase transitions of single droplet containing (NH₄)₂SO₄ and organic surfactants. Our findings reveal that increasing acidity can enhance the miscibility between the inorganic (AS) and organic (C7-C10) phases, thereby inhibiting LLPS. Additionally, the carbon chain length emerges as a crucial factor influencing LLPS, with longer carbon chains exhibiting stronger hydrophobicity and triggering the LLPS phenomenon more readily. Furthermore, the acidity of AS and C₈H₁₇SO₃Na mixture droplet was monitored in-situ throughout the entire LLPS process. The disappearance of whispering gallery modes (WGM) indicates the formation of a partially engulfed morphology during separation. Concurrently, a red shift of SO₄^2- (ΔRS=0.39-0.52 cm⁻¹) was observed, suggesting the formation of new phases of (NH₄)₂SO₄. Moreover, the acidity of droplet was found to be sensitive to the phase state, with the pH decreasing gradually  (ΔpH=-0.25) before the phase state change and undergoing a sharp decline (ΔpH=-1.5) during the transition from liquid to solid. This phenomenon indicates that the acidity of droplets in the partially engulfed morphology exhibits a more pronounced feedback on ambient conditions compared to that in the homogeneous mmorphology. Our results underscore the role of aerosol acidity in inhibiting phase separation and highlight the significant impact of the phase state on acidity variations.

Solubility largely determines the impacts of aerosol Fe on marine ecosystems and human health. Currently, modeling studies have large uncertainties in aerosol Fe solubility due to inadequate understanding of the sources of dissolved Fe. This work investigated seasonal variations of Fe solubility in coarse and fine aerosols in Qingdao, a coastal city in the Northwest Pacific, and utilized a receptor model for source apportionment of total and dissolved aerosol Fe. Desert dust was found to be the main source of total Fe, contributing 65 and 81% annually to total Fe in coarse and fine particles, respectively; in contrast, dissolved aerosol Fe originated primarily from combustion, industrial, and secondary sources. The annual average contributions to dissolved Fe in coarse and fine particles were 68 and 47% for the secondary source and 32 and 33% for the combustion source, respectively. Aerosol Fe solubility was found to be highest in summer and lowest in spring, showing seasonal patterns similar to those of aerosol acidity. Increase in Fe solubility in atmospheric particles, when compared to desert dust, was mainly caused by secondary processing and combustion emission, and the effect of secondary processes was dictated by aerosol acidity and liquid water content.

The Indian summer monsoon is crucial to over a billion people since it supplies over 75% of the country’s annual precipitation.  Significant intraseasonal variability causes flooding and water shortages, affecting agriculture, water supplies, and damaging infrastructure. We use observations and the ERA5 reanalysis to understand the dynamics and thermodynamics of active and break periods in the monsoon over 1940–2023.  We perform a trajectory analysis, in which air parcels are tracked backwards from the core monsoon zone of central India, at various pressure heights in the lower and mid-troposphere, for ten days.  The trajectories are performed for each June-September day for the whole data period, generating a unique dataset.  The trajectory dataset is then decomposed into active and break periods, based on gridded rainfall observations from the India Meteorological Department.  Trajectory density maps demonstrate clear structural differences; while the low-level flow towards the core monsoon zone is dominated by the recognisable C-shaped monsoon circulation during active events, the flow of dry air from sub-tropical and mid-latitude regions to the west becomes dominant even at low levels during breaks.  This Lagrangian analysis is complemented by a Eulerian study of the role of regional and synoptic-scale weather features. In the second part of the work, we focus specifically on the role played by dry intrusions as precursors to monsoon breaks, which is not well understood.  We develop an index based on moisture deficit and find that most breaks are associated with dry intrusions, which begin to enter India around a week prior to the middle day of breaks.  As breaks evolve, these dry intrusions deepen throughout their horizontal extent and descend into the country, stabilising the troposphere and creating an unfavourable environment for deep convection.  We also find that extended breaks have stronger dry intrusions as precursors.

This study examines how the Madden-Julian Oscillation (MJO) phases affect afternoon diurnal convection (ADC) patterns in Sri Lanka during 2001-2020. Sri Lanka experiences the highest frequency of ADC events in the Indian subcontinent region while located in a pivotal position within the propagation pathway of the MJO. To address the research gap regarding the MJO’s impact on seasonal diurnal rainfall in Sri Lanka, we analyze both the spring and autumn seasons, which are the two seasons with greater diurnal rainfall variability, focusing on strong MJO phases (P1-P8). Our findings show that daily rainfall increases during the P2-to-P3 phases and decreases during the P6-to-P7 phases in both seasons. The diurnal rainfall patterns, however, show seasonal differences. In spring, the diurnal rainfall amplitude peaks during P2-to-P3 phases, while in autumn, it peaks during P8-to-P1 phases. ADC events are more frequent and intense during these respective phases. The MJO's effect on both diurnal rainfall amplitude and ADC events is stronger in autumn compared to spring. During active MJO phases, we observe enhanced westward propagation of diurnal rainfall linked to ADC events, driven by moisture convergence and increased upward motion. The combination of mid-to-upper level easterly winds and deep convection over Sri Lanka leads to more distinct westward propagation during P2-to-P3 phases in spring and P8-to-P1 phases in autumn. These findings enhance our understanding of how the MJO influences local rainfall patterns and can aid in improving regional weather forecasting.

Tropical cyclones (TCs) that undergo rapid intensification (RI), defined as wind speed increases of at least 13m/s within 24 hours, present substantial forecasting challenges and risks. Current predictive models for RI TCs demonstrate limited accuracy, with an 82.6% detection rate and 27.2% false alarm frequency. To overcome these limitations, we introduce RITCF-contrastive, an innovative forecasting model that employs contrastive learning techniques. This model integrates satellite infrared images with comprehensive atmospheric and oceanic data, while specifically addressing dataset imbalances and incorporating critical TC structural characteristics. Evaluated on 1149 TC events in the Northwest Pacific (2020-2021), the RITCF-contrastive model achieves superior performance, with a 92.3% detection rate and only 8.9% false alarms. Compared to conventional deep learning approaches, our model shows an 11.7% enhancement in detection capability and 3 times reduction in FARate. This advancement not only significantly improves RI TC prediction accuracy but also establishes a new methodological framework for forecasting these high-impact meteorological phenomena.

Generally, the interaction between the East Asian subtropical westerly jet (EASWJ) and the East Asian summer monsoon (EASM) is regarded as a critical dynamic factor in the evolution of precipitation patterns in China. Using simulation results from the transient climate evolution since the last glacial maximum, this study applies the multivariate empirical orthogonal function (MVEOF) analysis method to investigate the co-evolution relationships of the EASWJ and the EASM during the early, middle and late Holocene, as well as their influence on precipitation patterns in China. The results indicate that all the first MVEOF modes in different time periods of the Holocene display an out-of-phase relationship in the intensity anomaly between the EASWJ and the EASM. However, the jet stream is wider and more tilted in the late Holocene. Under the influence of secondary circulations formed during their co-evolution, a north-south dipolar precipitation pattern in eastern China (“flood in the south and drought in the north” or “drought in the south and flood in the north”) and an east-west dipolar pattern in northern China (“wet in the west and dry in the east” or “wet in the east and dry in the west”) are found in the early and middle Holocene, while in the late Holocene a regionally-consistent precipitation pattern is witnessed across the whole region. In this mode, the precipitation in the middle and late Holocene is primarily dominated by trend changes. The second MVEOF mode reveals that the EASM weakens when the EASWJ shifts eastward and the jet steam axis shortens during the early Holocene, resulting in a north-south dipolar precipitation pattern in eastern China and a regionally-consistent pattern in northern China. In the middle Holocene, dipolar precipitation patterns are also observed in both eastern China and northern China when the EASWJ moves northward and the EASM strengthens, while the moisture condition in North China is less pronounced, and vice versa. In the late Holocene, the intensity anomalies of the EASWJ and the EASM exhibit an out-of-phase relationship in the temperate zone and an in-phase relationship in the subtropical zone, leading to a tripolar precipitation pattern in eastern China and a dipolar pattern in northern China. In this mode, the precipitation during the middle and late Holocene is primarily dominated by centennial oscillations. The precipitation patterns influenced by the co-evolution relationship between EASWJ and EASM correspond well with the reconstructed precipitation data, providing an explanation for the precipitation patterns observed in the reconstructed data from the perspective of dynamical mechanisms.

Abstract
This study investigates the distinct response of the Asian Summer Monsoon during the first and third years of triple La Niña events, with a focus on precipitation anomalies in Southern China and Pakistan. By analyzing sea surface temperature (SST) patterns and atmospheric circulation changes, we reveal how the persistence of La Niña across three years affects regional rainfall. Our results show that during the first year of a triple La Niña, Southern China experiences above-average precipitation while Pakistan faces reduced rainfall due to weakened monsoon activity. However, by the third year, Southern China suffers from significant drought with greatly reduced precipitation, while Pakistan experiences an increase in rainfall. These findings demonstrate the cumulative effects of triple La Niña events on the Asian Summer Monsoon, revealing contrasting precipitation responses between the first and third years. The study emphasizes the importance of considering multi-year La Niña events in long-term climate prediction and water resource management strategies for the affected regions.

The present study, for the first time, reports a significant increase in the frequency of Monsoon depressions (MDs) over the northern Arabian Sea in the past four decades. The analysis reveals that increased frequency of MDs due to the substantial variations in both dynamic and thermodynamic parameters across the observation array. Notably, there has been a noteworthy upswing in the Genesis Potential Parameter (GPP) within the northern Arabian Sea sector of the region, shedding light on the increased likelihood of MDs forming in this area during recent monsoon seasons in contrast to the decreasing MDs in Bay of Bengal. However, this finding strongly underscores the increased risk of the emergence and expansion of MDs in the Arabian Sea region over the past two decades, because of rising mid-tropospheric moisture, dynamical instability, augmented relative vorticity at 850 hPa, and weakened shear between upper and lower tropospheric winds. Therefore, it provides absolute assurance of their occurrence with the increasing dynamical process of its formation seems to be due to a combination of barotropic and dynamical instability. The evidence points to a heightened potential for MDs development in this area. Certainly, this is one of the significant contributors to the increased rainfall over northwestern India (NWI) in recent decades.

Using the Japanese Atmospheric General circulation model for the Upper Atmosphere Research-Data Assimilation System (JAGUAR-DAS), a whole neutral atmosphere reanalysis dataset (JAWARA) over about 19 years from September 2004 to December 2023 is produced. JAWARA is the first long-period reanalysis covering the height region from the surface to the lower thermosphere (~ 110 km). This wide height coverage is a notable advantage over other operational reanalysis datasets, which cover up to the middle mesosphere. Key dynamical characteristics are compared between JAWARA and two satellite observations and three other operational reanalysis datasets in their covered height regions. The seasonal variations of zonal mean temperature and zonal wind are similar between JAWARA and the datasets used for comparison. The climatologies of zonal mean temperature, zonal wind, residual-mean circulation, and E-P flux in the meridional cross section are also broadly consistent with other reanalysis datasets. The analyzed residual-mean vertical flow in the northern high latitudes in the middle atmosphere exhibits the well-known patterns of upwelling in summer and downwelling in winter. JAWARA also shows a prominent feature of strong downward propagating anomalies from the lower thermosphere to the upper stratosphere after sudden stratospheric warmings. This analysis takes full advantage of the JAWARA data, which cannot be made using satellite observations and other reanalysis datasets. This reanalysis product is expected to contribute broadly to future research on the characteristics of observed mesospheric phenomena, thermosphere–ionosphere coupling, space weather, and improvement of middle atmospheric meteorological systems, including their interannual and decadal scale variability.The JAWARA dataset is available online: https://doi.org/10.17592/002.2025010407

The quasi-decadal variability of the climate system in the Northern Hemisphere can be identified both by fluctuations in the large-scale anomalies of Pacific Ocean mid-latitudes sea surface temperature, so called Pacific decadal oscillation (PDO), and by fluctuations in the intensity of the Northern Hemisphere stratospheric polar vortex (iSPV). The existence of these regimes and the study of the peculiarities of the stratosphere-troposphere interaction during various phases of this quasi-decadal oscillation can significantly improve meteorological forecasts on different time scales.In our work we analyse how PDO and the Northern Hemisphere stratosphere quasi-decadal variability relate to each other and we try to restore the cause-and-effect relationship in the mechanism of formation of this variability. Based on analysis of observational data (JRA-55) and model experiments (GCM-INM-CM5) it is shown that the transition from one regime to another in the stratosphere precedes the transition in the ocean by 5–6 years. Since the stratospheric dynamics are influenced by the intensity of the vertical propagation of wave activity from the troposphere to the stratosphere, we can conclude that phase transition starts in the troposphere and manifests in quasi-stationary planetary waves structure and dynamics changes. When a phase changes, stratospheric processes immediately react, this can be seen from the SPV intensity, and then over the period of 5–6 years the ocean relaxing to this transition. So, the iSPV can be considered as indicator of the PDO phase change. The difference in the SPV intensity between the PDO phases is significant: a positive phase characterized by the SSW frequency of occurrence – 0.55, and a negative phase – 0.74. This difference is more significant than the difference between ENSO phases, during El Niño events – 0.72, and during La Niña events – 0.68.

Since the discovery of the Antarctic ozone hole, the variability, causes, and climate impacts of stratospheric ozone have received attention. Stratospheric ozone depletion was expected to recover within this century because of the Montreal Protocol. However, unlike the relatively consistent conclusions regarding ozone changes in the Southern Hemisphere, the recent trends of Arctic stratospheric ozone are still under debate. The work focuses on the long-term variability of Arctic stratospheric ozone, its dominant factors, and its climate impacts on East Asia. The key findings are as follows: 1) Arctic lower stratospheric ozone has exhibited a continued decline since the 2000s, suggesting that ozone depletion remains ongoing in this century. 2) The underlying dynamical mechanism governing Arctic ozone variability are examined, revealing that the North Pacific sea surface temperature anomalies modulate the polar stratospheric ozone concentrations via regulating the planetary wave activity and Brewer-Dobson circulation. 3) A physical linkage is identified between Arctic stratospheric ozone and spring precipitation in eastern China, providing a potential predictor for seasonal precipitation forecasts in the region.

Solar eclipse events (SEs) provide an excellent natural experiment for the response of the middle atmosphere to short-term changes in solar radiation. This study examines the averaged response of ozone (O₃) in the middle and upper atmosphere to among 31 pieces of solar eclipse events (SEs) since July, 2024 using Aura MLS satellite observations. O₃ in the upper stratosphere and mesosphere has short chemical lifetime (~100 s) thus sensitive to the radiation changes even of during the hourly SEs. Observational data derived from Aura MLS show that ozone mixing ratios in the lower mesosphere exhibit significant increases, reaching increase ratio of 95% at 0.1 hPa and 65% at 0.14 hPa, approximately 50% and 70% of the diurnal difference, respectively. The gradient of ozone increase during SEs diminishes with increasing obscuration (η), particularly when η exceeds 70% due to reduced water vapor photodecomposition and the dominance of the Chapman cycle. On the other hand, O₃ increase more than nighttime in the upper stratosphere are observed, possibly driven by dynamical effects such as localized upward air flow rather than O₃ production processes alone. This study identifies a possible correlation between the rapid solar radiation decrease during SEs and localized updrafts in the upper stratosphere, as evidenced by a 40% reduction in long-lived carbon monoxide (CO) near 0.46 hPa when η exceeds 70%.

Using various observations and a chemistry-climate model, this study investigates the impact of wintertime total column ozone (TCO) on snow cover over the Tibetan Plateau (TP). The results indicate that during anomalously high TP TCO events, the snow cover on the TP is anomalously high, and vice versa. Further analysis reveals that high TP TCO leads to a warmer stratosphere and a colder troposphere. The stratospheric warming induces an upwelling of air masses, resulting in reduced geopotential height in the upper troposphere and lower stratosphere (UTLS) over the TP. The cyclonic circulation in the UTLS associated with this low-pressure leads to high-potential vorticity (PV) stratospheric air being transported to the western TP. To maintain the conservation of PV, cyclonic vorticity forms above the western TP in the UTLS and can extend into the troposphere. The southerlies associated with this cyclonic circulation, and positive vertical gradient of zonal wind in the troposphere, induce ascending motions over the TP. The tropospheric cyclonic circulation over the western TP also transports water vapor from the Bay of Bengal and the Arabian Sea to the TP. Ascending motions and increased water vapor lead to increased snowfall over the TP. Additionally, ascending motions and increased water vapor are conducive to cloud formation, which amplifies the cooling of the TP caused by TCO increase. The snow-albedo positive feedback further strengthens the cooling and prolongs duration of snow cover. Model simulations suggest that a change of 10 DU in TCO can result in a 1.7% change in snow cover.

Air pollution, particularly fine particulates (PM_2.5), poses a significant threat to human health and ecosystems, and its impact is amplified in sensitive mountainous regions like the central Himalayas. This work showcases a comprehensive investigation of PM_2.5 pollution in Uttarakhand, India, leveraging a unique network of Compact Useful PM_2.5 Instruments (CUPI) deployed across all 13 districts of the state. Initial campaign measurements using low volume samplers provided crucial baseline data on physical and chemical characteristics of Particulate matter (PM), revealing distinct source properties and apportionment across diverse terrains (peaks/slopes, valleys, and plains). The subsequent high-resolution CUPI network data unveils the dynamic evolution of PM_2.5 levels, including diurnal and seasonal patterns, and their transport across these distinct complex terrains. Further, the sensitivity of PM_2.5 concentrations to local meteorological parameters, identifying key drivers of pollution variability in this area is also explored. Bias-corrected ERA and MERRA-2 reanalysis datasets are integrated to assess long-term PM_2.5 trends and possible future projections. Event-based analyses, such as dust storms and winter biomass burning transport from North India, highlight the complex interplay of regional and local factors. Finally, we present preliminary WRF-Chem modeling efforts to simulate these events, validating model performance against ground truth and reanalysis data. This integrated approach provides critical insights into PM_2.5 pollution dynamics in the Himalayas, informing mitigation strategies and contributing to improved air quality management in this vital region.

Tropopause folds (TFs) are regarded as the main mechanism for stratosphere-troposphere exchange (STE) in the midlatitude region and play an important role in the redistribution of gases and aerosols between the troposphere and stratosphere. However, studies characterizing what fraction of STE is associated with tropopause folds are lacking. In this study, we use Lagrangian trajectories to estimate STE mass flux and the TF detection algorithm by Škerlak et al. (2015) to determine whether trajectories associated with STE are related to TFs. Trajectories and TF detection are performed on the ERA5 model-level data at 0.5° spatial and 3-hourly temporal resolution. By matching exchange events to TFs using a temporal and spatial separation threshold of 1 hr and 300 km (based on the method of Sprenger et al., 2003), we find that in the Northern Hemisphere during year 2022 the annual average stratosphere-to-troposphere (STT) mass transport of all transport events is ~1.14×10¹⁰ kg s⁻¹, while those associated with TFs account for roughly 34.8% of it. For troposphere-to-stratosphere (TST) transport, the average transport rate is ~5.91×10⁹ kg s⁻¹, with TFs accounting for 27.4%. We plan to perform this analysis for the period 1980–2023 and assess the presence of trends in transport and the role of TFs in both STT and TST.

Heat and air pollution extremes are two leading global health stressors, both of which are particularly serious in China and India. It is well recognized that exposure to co-occurrence of heat and air pollution extremes will cause amplified health outcomes, yet century‐long understanding of future co‐occurrence is still lacking. On the basis of sophisticated regional coupled climate-chemistry modeling, we predict future individual and joint occurrences of heat and air pollution extremes in China and India in 2096–2100 relative to 2010–2014. We find intensified co-occurrences of heat and air pollution extremes in both China and India, despite reductions in projected emissions and improved air quality. Under the medium air pollution control of SSP245, the frequency of T_w&PM&O₃ joint hazard increases by 382% in North India, and 729% in Beijing by the end of this century. Given the significant role of temperature changes in the co-occurrence and larger compounding health impacts, actions are urgently needed to reduce exposure to co-extreme events.

In the past decades, the Tibetan Plateau has experienced rapid warming, accompanied by marked differences in seasonal extreme precipitation change. However, research focusing on the impact of human activities on the seasonal extreme precipitation is still limited. In this study, we employ different observations and CMIP6 models to investigate the relevant issue based on an optimal fingerprinting method. We found that both station data and ERA5 reanalysis data show similar detection results. In one-signal analyses, the anthropogenic signal can be detected in the increasing trends of maximum one-day precipitation amount (Rx1day) and maximum consecutive 5-day precipitation amount (Rx5day) in spring and winter but not in other seasons, whereas the signal of natural forcing cannot be detected in Rx1day and Rx5day changes during these two seasons. In two-signal analyses, human activities on extreme precipitation changes over the Tibetan Plateau is noticeable during both spring and winter, with the human influence being more pronounced in spring. This study thus provides important evidence for human influences on the seasonal change of extreme precipitation over the Tibetan Plateau.

Recent observations have unveiled a remarkable transformation in the spatial distribution of heatwaves across East Asia. Traditionally, such events have predominantly impacted southeastern China. However, an in-depth analysis of CPC/NOAA land surface temperature data has pinpointed the middle and lower reaches of the Yangtze River Basin as the primary center for extreme heatwave events in the 21st century. Sophisticated threshold analysis techniques have revealed that the frequency and spatial coverage of these extreme heatwaves in the Yangtze Basin are escalating at a pace significantly quicker than that of conventional heatwave events. The westward expansion of the western Pacific subtropical high (WPSH) stands out as the pivotal force driving this phenomenon. Rigorous statistical analysis has revealed a profound link between atmospheric conditions and temperature extremes: an increment of 10-gpm in geopotential height over the basin is associated with a surge of 0.43°C in regional maximum temperatures, accompanied by a 4% increase in the area affected by heatwaves. In contrast to the relatively stable eastern boundary of the North Africa high, the WPSH has undergone a notable westward shift in recent decades, exacerbating the heat-dome effect over the Yangtze Basin. Climate projections derived from 29 CMIP6 models under the SSP2-4.5 scenario indicate a persistent intensification of atmospheric conditions conducive to extreme heat in the Yangtze Basin throughout this century. These findings underscore the basin’s increasing importance as a central hub for intense heatwave activity within East Asia’s monsoon system, with significant implications for developing regional climate adaptation strategies.

Southeast Asia, home to nearly 700 million people and comprising mostly least developed and developing countries, is one of the most exposed and vulnerable regions to the impacts of climate change, particularly extreme climate events. Given the uncertainty surrounding the success of limiting global warming under the Paris Agreement, it is crucial for countries in the region to advance adaptation measures to minimize the impacts of future climate extremes and enhance climate resilience. For climate adaptation efforts to be effective and to prevent maladaptation, these measures must be evidence-based, relying on assessments of future climatic extremes and hazards. High-resolution climate projections play a vital role in providing information on how future climate extremes may change in response to varying levels of greenhouse gas emissions (emission scenarios) or global warming. However, generating robust, multi-model, multi-scenario, high-resolution climate projections is highly technical, expensive, and beyond the capabilities of most countries in the region. This is where CORDEX Southeast Asia has stepped in, addressing this gap by conducting coordinated regional climate downscaling simulations since 2013. CORDEX Southeast Asia has produced multi-model projections at a 25 km resolution using CMIP5 GCMs for the Southeast Asian domain, further downscaled to a 5 km resolution over several sub-domains. Currently, it is completing multi-model projections using CMIP6 GCMs. Additionally, CORDEX Southeast Asia is undertaking a project called CARE for Southeast Asia Megacities, aimed at generating climate projections at the city scale for common climate extremes in five megacities across the region. This talk will highlight key advancements, insights, challenges, and gaps in understanding future climate extremes in Southeast Asia.

Although compound drought and heatwave extremes have recently drawn much attention globally, there exist three interesting issues to explore as follows: First, how can we perform event detection as accurately as possible? Second, whether droughts are always concurrent with heat waves remains unknown. Third, whether droughts can be reproduced directly utilizing atmospheric dynamics.To explore the three issues, our recent achievements are as follows: First, regarding accurate event-oriented detection, we generated a global-scale set of seasonal-scale meteorological drought events following the recently proposed 3D DBSCAN-based workflow of event detection. [see algorithm cases (https://doi.org/10.1016/j.aosl.2022.100324) and global drought detection (https://spj.science.org/doi/10.34133/olar.0016 ) ] Second, we investigated diversity of temperature extremes compounded with droughts. We found associated global-scale HOT-, COLD-, NORMAL-, and HYBRID-type event-oriented droughtS account for approximately 40%, 10%, 30%, and 20%, respectively, during 1980–2020. The HOT and NORMAL types appeared mostly within 45°S~45°N in warm seasons, with Cold types over mid-high latitudes in cold seasons. The achievements may provide robust event-oriented insights into the physical mechanisms behind global droughts and concurrent temperature anomalies. [see diversity of temperature anomalies (https://spj.science.org/doi/10.34133/olar.0017 ) ] Third, regarding dynamically-based reconstruction of compound droughts and heatwaves, we employ three kinds of dynamic features (i.e., vertical velocity, relative vorticity, and horizontal divergence) for hydrometeorological reconstruction (e.g., precipitation and near-surface air temperature) under drought situations through a so-called XGBoost method. The study adopts the reconstruction scheme on the interannual variability and finds dynamically based reconstruction feasible. More importantly, from interpretable perspectives, global-scale analysis of dynamic contributions helps discover unexpected dynamic drought-inducing roles and associated latitudinal modulation. For instance, low-level cyclonic/anticyclonic anomalies contribute to drought development in the northern middle and high latitudes. These achievements could provide guidance for dynamically based drought monitoring and prediction across the globe. [see paper (https://doi.org/10.1175/JHM-D-22-0006.1)]

Against the background of global warming, the frequent occurrence of drought events, particularly extreme droughts, has emerged as the most critical bottleneck constraining economic development, livelihood improvement, and ecological security in China. Understanding the historical evolution characteristics of extreme drought in China under climate warming, quantifying the contributions of various influencing factors, and projecting future development trends constitute not only crucial scientific issues requiring immediate resolution but also represent a significant national demand for enhancing disaster prevention and mitigation capabilities in the new era. Our findings reveal that strong El Niño events do not necessarily lead to extreme summer droughts in northern China, with the positive Eurasian mid-latitude teleconnection pattern (EU pattern) being the critical factor. The CFSv2 dynamic prediction model can only accurately predict extreme drought events in northern China when it effectively captures the positive phase of the EU pattern. Furthermore, our newly developed dynamic-statistical prediction model, incorporating both central-eastern Pacific sea surface temperatures and spring Eurasian snow cover, demonstrates superior performance over the CFSv2 dynamic model at longer lead times. Anthropogenic activities have significantly increased the risk of extreme high-temperature and low-precipitation events in northeastern China during spring and summer, particularly over the past three decades, including the enhanced probability of extreme Baikal high-pressure events similar to the 2017 case. Notably, human influence has increased the likelihood of extreme drought in southwestern China by approximately 50% and extreme high-temperature events by 170%, with a 60% increase in concurrent extreme heat and drought events over the past 30 years. Additionally, anthropogenic forcing has substantially elevated the risk of Bay of Bengal anticyclone events during spring and summer in northeastern China.

Under global warming, the escalating frequency and intensity of hot extreme events, coupled with a global decline in relative humidity, have led to an increased likelihood of hot-dry extremes. We find that the probability of hot-dry coupling in more and more dry extremes increases (“drier get hotter” phenomenon). By exploring the coupling relationship between global surface air temperature and atmospheric humidity, the historical and future spatiotemporal trend of “drier get hotter” phenomenon are investigated and predicted, and the physical mechanisms underlying the global hot-dry coupling are analyzed, encompassing heavy precipitation patterns, surface energy distribution, and land-atmosphere interactions, both in historical and future periods. Finally, the contribution of anthropogenic climate change signal to the trend of hot-dry extremes coupling is quantitatively attributed.

Compound heatwaves, characterized by sustained high temperatures both day and night, have profound impacts on ecosystems, human health, and socio-economic systems. This study employs ERA5 reanalysis data and applies spectral clustering to classify the circulation patterns associated with regional compound heatwave events along the South China coast during the summers of 1981–2023, and investigates their dynamic and thermodynamic mechanisms. The results indicate that compound heatwaves in this region can be categorized into two types: the subtropical high type and the cyclonic type, which account for 73% and 27% of all events, respectively.The subtropical high type is driven by the westward extension of the Western Pacific Subtropical High, accompanied by deep subsidence, leading to longer-lasting heatwaves with an average duration of 4.32 days. In contrast, the cyclonic type is closely associated with subsidence in the periphery of tropical cyclones, exhibiting higher regional average daily maximum temperatures (33.95°C) and affecting a broader area, with high-temperature grid points covering 35.0% of the region. However, these heatwaves are shorter in duration, averaging 3.90 days. Both types of heatwaves are strongly linked to enhanced radiative forcing and adiabatic heating due to vertical motion. The subtropical high type relies more on the stable maintenance of large-scale circulation, while the cyclonic type is closely related to the movement and relative position of tropical cyclones. This study provides new insights into the spatiotemporal characteristics and formation mechanisms of compound heatwaves, offering significant implications for improving the forecasting of extreme high-temperature events.

The frequent occurrences of extremely hot days (EHD) in summer poses significant threats to society, ecosystems, the economy, and human health. While the summer of 2024 marked the hottest meteorological summer on record for the Northern Hemisphere, Northern China experienced its highest frequency of EHD in the summer of 2023 since the year 1979. This study investigates the impacts of interannual (IA) and interdecadal climate variabilities, including global warming (ID+GW), on these extreme heat events. Through observational analysis, we reveal that a quasi-tropical high-pressure anomaly, linked to a mid-latitude Rossby wave train, induces downward motion over the Northern China region, resulting in positive vertical adiabatic heating and thermodynamic heating in the lower atmosphere, despite negative horizontal heat advection. Notably, the Rossby waves associated with both IA and ID+GW variabilities primarily originate from the North Atlantic, with additional contributions from the Pacific-Japan teleconnection pattern for IA variability. To further validate these findings, idealized numerical experiments were conducted using an atmospheric general circulation model (AGCM), confirming the proposed mechanisms.

Atmospheric energetics deals with atmospheric energy transport, including latent heat, sensitive heat, radiation, and mechanical energies. The various mechanical components of energy, their interconversion, generation, and dissipation, are interlinked and described by the Lorenz Energy Cycle. Available potential energy(APE) is generated when the thermal structure of the atmosphere alters by redistribution and relocation of energy. The theoretical framework and the current datasets allow for studying long-term temporal changes in global atmospheric energetics. Lorenz energy cycle describes how solar heating generates APE and its transformation into kinetic energy to drive atmospheric dynamics. The primary energy components are Mean Available Potential Energy(MAPE) and Mean Kinetic Energy(MKE). In this work, we calculated the variation in atmospheric energetics over the Indian continent and nearby regions owing to El Nino. The energy distribution is compared with non-El Niño years to assess the changes in energy components during the El Niño. Since 2000, El Niño events have been observed in 2002–03, 2004–05, 2006–07, 2009–10, 2014–16, 2018–19, and 2023-24. El Nino years after the year 2000 are considered in this study. The results are compared with non-El Nino years, which are 2001-02, 2004-06, 2009, 2013-15, and 2017-20. The study is performed for two seasons: pre-monsoon and monsoon. Mar-Apr-May months defined the Pre-Monsoon season, and Jun-Jul-Aug months defined the Monsoon season. Significant changes were observed globally from an El Niño year to a non-El Niño year, and similar changes were observed over the Indian region as well. MKE was significantly high(~2*10⁵ J/m²) over the Arabian region and significantly low (~0.5-1.5*10⁵ J/m²) over the Bay of Bengal region during the El Niño year in the pre-monsoon season. This causes dynamical changes in the atmosphere, which could be responsible for decreasing Indian monsoon rainfall.

The Himalayan region is experiencing rapid environmental changes because of climate change and human interventions, significantly affecting the hydropower development, agricultural patterns, and water resources. This study integrates findings from three research’s that analyse the impacts of the hydropower projects, the shifting altitude of apple orchards, and the disappearance of natural springs. Using a combination of primary and secondary data sources, including long-term meteorological datasets (1955–2021), field surveys, and hydrological modelling, the studies offer a thorough evaluation of these critical issues.The hydropower impact study uses rainfall, temperature, and soil moisture data from 1955 to 2019, analysed using the M-K test, Linear Regression Model, and Sen’s slope method. Field surveys were conducted at 12 hydropower sites to assess local perceptions. The findings exhibit declining rainfall, increasing temperatures, and groundwater depletion, causing river fragmentation, biodiversity loss, and higher disaster risks such as landslides and flash floods.The study on shifting apple cultivation examines temperature, rainfall, and chilling hour data (1975–2014) utilizing trend analysis and field surveys across major apple-growing regions. Results indicate a significant decline in chilling hours, forcing farmers to shift apple orchards from 1200–1500m to above 3500m. The study also highlights increasing pest infestations and soil degradation in lower altitudes, leading to economic losses.The spring disappearance study uses rainfall-runoff modelling (MIKE 11 NAM model) and geotagging of 276 springs to analyse the decline in groundwater recharge. Results indicate that over 60% of springs have dried up, primarily due to climate change, groundwater over-extraction, and hydropower development, creating a severe water crisis in many Himalayan villages.The findings emphasize the need for sustainable hydropower policies, climate-resilient agriculture, and integrated water conservation strategies to safeguard ecological and socio-economic stability in the Himalayas.

Accurate prediction of extreme precipitation information is crucial for disaster risk management, social and economic development security, and climate change research. However, state-of-the-art regional climate models still have difficulties in simulating extreme weather. In order to take appropriate measures to reduce the risk of water-related disasters, which are expected to become more severe with climate change, there is an urgent need to develop technologies that can accurately represent predict localized weather patterns by regional weather models.  First of all, we have investigated the predictability of extreme rainfall event in Japan with utilizing two global reanalysis products (JRA-55, ERA-5) which are widely used in regional weather modelling studies. As compared daily precipitation on two reanalysis products with observation data at Tokyo metropolitan area, the results indicated that JRA-55 tended to overestimate daily precipitation whereas ERA-5 tended to underestimate it. The reasons causing these over- and under- estimations would be due to their rough spatial resolution of grid cells which could not capture the land surface information of specific point where a monitoring equipment were installed.In this presentation, we utilized an additional regional reanalysis product (Long-Term Regional Reanalysis for Japan with Assimilating Conventional Observations, called as RRJ-Conv) (Fukui et al., 2018; Hunt et al., 2007; Kobayashi et al., 2015; Saito et al., 2007) to be compared with two global products to clarify the predictability of summertime extreme rainfall events in regional weather models. Acknowledgement:This work was supported by JST Grant Number JPMJPF2013.

As a relatively new atmospheric model, the Model for Prediction Across Scales–Atmosphere (MPAS-A) has been widely applied for large-scale simulations. However, its performance in mesoscale simulations, particularly in tropical regions, remains less explored. MPAS utilized unstructured centroidal Voronoi meshes with C-grid staggering and geometric-height vertical coordinates. It supports both global and limited-area (regional) simulations, as well as uniform and variable horizontal resolutions, enabling a smooth transition between mesh resolutions. This study aims to evaluate the performance of MPAS-A in simulating urban-scale phenomena, with a focus on extreme surface air temperature in Jakarta (Indonesia) during the hot spells of October 2023. Eight simulations were conducted using different initialization/boundary conditions (IBCs), terrestrial datasets, and simulation domains to test the flexibility of MPAS-A in modifying each of those inputs. Various validation metrics were applied to assess near-surface meteorological variables, land surface temperature (LST), and the vertical atmospheric profile. MPAS-A simulated diurnal patterns of the near-surface variables well, except for wind direction. The model also performed well in simulating LST. Moreover, the biases in the vertical profiles varied with height and were sensitive to the choice of IBCs. MPAS-A successfully simulated the extreme event, capturing higher air temperatures in the southern part of Jakarta compared to the northern region. Negative temperature advection by sea breeze affected the lower air temperature in the northern area. This study highlights the role of sea breezes as natural cooling mechanisms in coastal cities. Additionally, MPAS-A is feasible for several applications for urban climate studies and climate projection such as dynamical downscaling of CMIP6 models for major cities in Indonesia to assess the effects of urbanization and global warming.

In this study, a record-breaking surface gust wind event of over 45 m s-1, which occurred in the coastal region of East China during the early evening hours of 30 April 2021, is examined. The dynamical characteristics of this event is explored by using a high-resolution mesonet comprised of eight radar wind profilers (RWPs), surface observations, radar and satellite data. Observation-al analyses show the development of several cloud clusters ahead of the axis of a midlevel trough with pronounced baroclinicity, and the subsequent organization into a comma-shaped squall sys-tem with a leading convective line over land and a trailing stratiform region moved offshore. The latter is embedded by a mesovortex with intense northerly rear inflows descending to the surface, accounting for the generation of the gusty winds. Results indicate the different roles of multi-scale processes in accelerating the surface winds to extreme intensity. Specifically, the large-scale baroclinic trough provides intense background rear inflows that are enhanced by the formation of the mesovortex, while moist downdrafts in the rear inflows account for the downward transport of horizontal momentum, leading to the generation of intense cold outflows and gusty winds close to the leading convective line. Despite the lack of sufficient observations for quantitative analysis, this study provides a qualitative analysis that offers valuable insights into the dynamics of extreme gusty winds. Moreover, the above results underscore the value of RWP mesonet ob-servations in enhancing our understanding of extreme wind events and in improving the nowcast-ing and prediction efforts in the future.

The post-landfall decay of tropical cyclones (TC) is often closely linked to the magnitude of damage to the environment, properties, and the loss of human lives. Despite growing interest in how climate change affects TC decay, data uncertainties still prevent a consensus on changes in TC decay rates and related precipitation. Here, after strict data-quality control, we show that the rate of decay of TCs after making landfall in China has significantly slowed down by 45% from 1967 to 2018. We find that, except the warmer sea surface temperature, the eastward shift of TC landfall locations also contributes to the slowdown of TC decay over China. That is TCs making landfall in eastern mainland China (EC) decay slower than that in southern mainland China (SC), and the eastward shift of TCs landfall locations causes more TCs landfalling in EC with slower decay rate. TCs making landfall in EC last longer at sea, carry more moisture upon landfall, and have more favorable dynamic and thermodynamic conditions sustaining them after landfall. Observational evidence shows that the decay of TC-induced precipitation amount and intensity within 48 h of landfall is positively related to the decay rate of landfalling TCs. The significant increase in TC-induced precipitation over the long term, due to the slower decay of landfalling TCs, increases flood risks in China’s coastal areas. Our results highlight evidence of a slowdown in TC decay rates at the regional scale. These findings provide scientific support for the need for better flood management and adaptation strategies in coastal areas under the threat of greater TC-induced precipitation.

Eastward propagation is an essential feature of the Madden-Julian Oscillation (MJO). Yet, it remains a challenge to realistically simulate it by global climate system models, and the reasons are not fully understood. This study evaluates the capability of 20 Coupled Model Intercomparison Project Phase 6 (CMIP6) models in simulating MJO’s eastward propagation and its intrinsic links with the dynamic/thermodynamic structures and the background mean states, aiming at better understanding the sources of the simulation errors. The metrics to evaluate the MJO internal dynamics consists of six parameters: (1) the east-west asymmetry in the low-level circulation, (2) the boundary layer moisture convergence propagation, (3) the vertical tilt of equivalent potential temperature or moist static energy, the vertical structures of (4) diabatic heating and (5) available potential energy generation, and (6) upper-level diabatic heating and divergence. We also gauge the performance of three MJO-related background mean-state fields, including precipitation, sea surface temperature, and low-level moist static energy. It is argued that these parameters are relevant internal and external factors that could affect MJO eastward propagation. We find that the boundary layer moisture convergence is most tightly coupled with the eastward propagation of MJO and controls the pre-moistening, destabilization, and the leading low-level diabatic heating and available potential energy generation. The CMIP6 models exhibit significant improvements against CMIP5 models in simulating MJO dynamic/thermodynamic structures and the mean states. The diagnostics in this study could help to identify the possible processes related to CMIP6 models’ shortcomings and shed light on how to improve simulation of MJO eastward propagation in the future.

This study reveals an interdecadal change in boreal winter dominant ENSO atmospheric teleconnection and demonstrates its underlying physical mechanism. During the period 1960–2020, the ENSO teleconnection was characterized by the Western Pacific (WP) pattern and the Pacific–North American (PNA) pattern, but the dominant pattern of ENSO teleconnection witnessed an alternation at around early 1980s. In the first epoch of 1960–1979 (E1), the ENSO teleconnection was mainly dominated by the WP pattern with an inhibited PNA pattern, whereas in the second epoch of 1987–2020 (E2), the ENSO teleconnection was dominated by the PNA pattern without a prominent WP pattern. It is further suggested that the changing strength of the westerly jet, rather than the pattern of the ENSO-related tropical convection, played the critical role in the alternation of the winter dominant ENSO teleconnection. During E1, because the westerly jet stream in the East Asia–North Pacific sector was relatively weak, poleward propagation of the Rossby wave train from the western Pacific was evident, forming the dominant WP pattern related to ENSO. In E2, because the westerly jet in the region largely strengthened, the PNA pattern was dominant owing to the enhanced zonal waveguide of the westerly jet. Evidence from a set of numerical experiments and the simulations of CMIP6 models supports the important role of the westerly jet strength in shaping the ENSO teleconnection.

Since the original formulation of the rate of information transfer between observables has been formulated by Liang and Kleeman (2005, PRL, https://doi.org/10.1103/PhysRevLett.95.244101), considerable progress has been made in the development of the theory (e.g. Pires et al, 2024, Physica D, https://doi.org/10.1016/j.physd.2023.133988), together with its application in various fields in particular in Atmospheric and Climate Sciences. In the present work, we will briefly introduce the methodology developed for 20 years now, and we will show its power in the context of different applications: first on a simplified reduced-order atmospheric model for which the rate of information transfer and the information entropy budget can be explicitly computed (Vannitsem et al, 2024a, QJRMS, https://doi.org/https://doi.org/10.1002/qj.4805), and the recent application of the methodology to the causal dependencies between climate indices at monthly time scales which reflect the large scale dynamics of the climate system over the North Atlantic and Pacific (Vannitsem et al, 2024b, https://doi.org/10.5194/egusphere-2024-3308). The method allows in particular to isolate key nonlinear causal dependencies among climate indices present at yearly to decadal time scales.

The impact of the tropical Atlantic Ocean on the Antarctic sea ice variation is well-established, but quantifying this impact on observed sea ice change remains challenging due to the interplay of multiple factors. To isolate this effect, we conduct an Atlantic Sea Surface Temperature (SST) pacemaker experiment using the fully coupled Norwegian Earth System Model. In this experiment, the prescribed Atlantic SST anomaly is nudged while allowing other components to evolve freely (PCMK), and the results are compared to a historical simulation where all components evolve freely (HIST). The difference between PCMK and HIST reveals the isolated influence of tropical Atlantic SST on Antarctic sea ice. Our findings show that constraining tropical Atlantic SST in PCMK leads to a more realistic sea ice variability in the Weddell Sea during austral autumn (March–April–May). Atlantic-induced Antarctic sea ice variation accounts for ~40% of the observed variability. This arises from a more realistic ascending branch of the Hadley Cell, which enhances upper-level zonal winds up to 40°S. The large-scale meridional circulation then transports this improvement southward, modulating lower-level atmospheric temperatures over the Weddell Sea and ultimately influencing sea ice variability through atmospheric thermodynamic processes. Overall, our study indicates a reliable Antarctic sea ice prediction requires an accurate Atlantic SST initialization and prediction.

During the modern satellite era, sea ice in the Southern Ocean has been experiencing substantial multiscale variability. On the interannual time scale, El Niño–Southern Oscillation (ENSO), as the strongest air-sea coupled mode occurring in the tropical Pacific, is a key factor affecting the Antarctic sea ice. Based on observational analyses and model experiments, the study investigated the asymmetric impacts of weak and strong La Nina (i.e., the negative phase of ENSO) on the Antarctic sea ice concentration (SIC) in austral summer (December to the following February). The amplitude of negative SST anomalies in the tropical Pacific associated with strong La Nina is approximately twice that associated with weak La Nina. The results suggest that the strong La Nina induces a positive phase of the Southern Annular Mode (SAM) in the Southern Hemisphere, resulting in a significant increase in sea ice in the Indian Ocean sector through dynamic and thermodynamic processes. In comparison, the weak La Nina causes a Pacific–South American (PSA) pattern. No obvious SIC anomalies can be seen around the Antarctic. To verify such asymmetric responses of the Southern Hemispheric atmospheric circulation and Antarctic sea ice to the weak and strong La Nina, several numerical experiments are conducted using the National Center for Atmospheric Research (NCAR) Community Atmosphere Model, version 5 (CAM5). The result showed that the strong event can remotely induce a positive-phase SAM-like response in the Southern Hemisphere, whereas the weak event generates a PSA response, exhibiting an asymmetric impact.

The Australian summer monsoon (AUSM) is the strongest monsoon in the Southern Hemisphere and it is greatly influenced by the climate conditions in the Indo-Pacific and adjacent regions. Inspite of the substantial studies of the monsoon, the linkage between the AUSM and the high-latitude Southern Ocean climate has not been fully understood. This study investigates how AUSM rainfall is affected by the Antarctic Circumpolar Current (ACC) by simulating scenarios with a closed and an opened Drake Passage with the Community Earth System Model. It is found that AUSM precipitation decreases as a result of reduced local humidity caused by a strengthening ACC. An opened Drake Passage leads to strengthening ACC and Atlantic Meridional Overturning Circulation, which in turn creates an El Niño-like state in the Pacific. This weakens the Walker Circulation, resulting in reduced local moisture, anomalous subsidence over the AUSM region, and a subsequent decrease in monsoon precipitation.

In recent decades, Antarctica has undergone significant climate change, with most studies focusing on the impact of oceanic multiscale variability on Antarctica, especially on West Antarctica. However, our research reveals that Indian summer monsoon (ISM) rainfall strongly influences the austral winter (June–August) Antarctic climate and sea ice concentration (SIC) through atmospheric teleconnections. Diabatic heating from ISM rainfall shifts the Hadley cell northward, triggering a Rossby wave train from the Mascarene Islands into Antarctica. This alters sea level pressure and induces warm advection to both East and West Antarctica, leading to widespread warming. Consequently, Antarctic SIC undergoes a tripole redistribution, with increases in the Ross and Weddell Seas, and decreases in the Amundsen and Bellingshausen Seas. These findings emphasize the importance of ISM rainfall in shaping Antarctic climate and SIC, offering a comprehensive explanation for the 2023 austral winter's historically lowest Antarctic SIC, with significant implications for climate change research.

 Severe convective weather forecasting is one of the challenging tasks in operational forecasting, with tornado monitoring and forecasting being particularly difficult within this field. This presentation introduces the advances in tornado and thunderstorm-wind-gust monitoring and forecasting technologies at the National Meteorological Centre (NMC) of China. The NMC has developed: 1) 0-3 day tornado and thunderstorm-wind-gust forecasting techniques based on global numerical weather prediction (NWP) model forecasts; 2) 0-6 hour short-term forecasting techniques for tornadoes and thunderstorm wind gusts combining high resolution NWP model forecasts, updraft helicity, and AI forecasting methods; 3) 0-2 hour monitoring and warning techniques integrating convective-storm structural characteristics, AI-based identification methods, and the “Fenglei” AI large model for nowcasting; along with a tornado and thunderstorm-wind-gust disaster field survey system. Based on the favorable environmental conditions and physical mechanisms, NMC has developed over 50 severe convective physical parameters including the significant tornado parameter (STP) and supercell parameter (SCP), where STP and maximum updraft helicity (UH) prove effective for short-range (0-72 hour) and short-term (0-6 hour) forecasting tornado respectively. Tornado and thunderstorm-wind-gust monitoring relies on tornado vortex signatures and dual-polarization parameter evolution observed by dual-polarization Doppler weather radars, leading to the development of a multi-task deep learning tornado identification model based on radar data and a deep learning thunderstorm wind gust identification model. These technologies have been integrated into the SWAN3.0 system - the operational short-term forecasting and nowcasting platform for severe convective weather of China Meteorological Administration. The work is funded by the China National Natural Science Foundation project (Granted No. U2342204).

Tornadoes, as a devastating wind hazard, have the nature of high wind speeds, nonuniform and nonstationary wind flow, and significant turbulence. This study employs a coupled CM1 (storm-scale Cloud Model 1) and fine-scale LES model to generate a nonstationary, translating tornado vortex and then investigate the influence of terrain on tornadic fields near ground. Six terrains (a flat ground, the escarpment terrains with two slopes and hill terrains with two slopes in ASCE 7-22, and a real-world terrain of Carney, OK) are utilized to study the effects of terrain on near-ground tornadic wind field.  Results suggest that the presence of terrain increases the central pressure deficit and increases the peak wind speed and the width of high-speed region in the tornado swath, enhancing tornado intensity, and causes path deviations. Specifically, the horizontal and vertical velocities at 10-m AGL are greater with terrain and the location of the maximum pressure deficit occurs along the uphill segment for all idealized cases except the steep hill. It is also noted that the precise location (with respect to the terrain) of the maximum wind velocities and pressure deficits varies with the terrain shape and slope. The 3D real terrain simulation shows similarities with the idealized terrain simulations to a certain extent; however, the vertical velocities are lower, and the strongest winds occur over a smaller region demonstrating the complexity of the tornado-terrain relationship.

Investigating the relationship between land cover type and tornado intensity can inspire innovative nature-based mitigation strategies to civil structures facing tornado threats. This study employs a coupled simulation of fine-resolution CFD simulation with minimum grid resolution of 0.01 m and a storm-scale modeling (Cloud Model 1-CM1). Five types of land covers from 16 National Land Cover Dataset (NLCD) are considered in the coupled model, which are smooth, grass, shrub, built-up and mixed forest. The land cover is modeled on the bottom boundary through wall function by using surface roughness length. Two control runs with no additional surface roughness are operated: one no-slip boundary, the other having a free slip boundary. The effects of each type of land cover only induce some fluctuations on the tornado path by comparing with the control run, indicating the negligible effects of land cover on tornadic translating motion. However, for tornado intensity, land cover enhances the tornado’s central pressure but reduces maximum horizontal velocity at 10-m AGL significantly. The relationship between land cover and tornado intensity is complicated and not simple monotonicity: there is a “sweet spot” that results in the maximum intensity of the tornado. The variation of core radius and maximum horizontal velocity along height suggests that the range of influence of surface roughness is essentially concentrated below 50 m. Moreover, comparisons between the two control runs indicate that the effect of the no-slip boundary on horizontal velocity is concentrated below 0.8 m, while its influence on peak vertical winds remains significant at a height of 10 m.

This study also took a step forward by considering more realistic ground condition with both land cover and terrain. The results indicate that the role of terrain is more significant than land cover in varying tornadic characteristics at near-ground regions.

The Pearl River Delta (PRD), a tornado hotspot, forms a distinct trumpet-shaped coastline that concaves toward the South China Sea. During the summer monsoon season, low-level southwesterlies over the PRD’s sea surface tend to be turned toward the west coast, constituting a convergent wind field along with the landward-side southwesterlies, which influences regional convective weather. This study explores the roles of the trumpet-shaped coastline in the formation of a tornadic mesovortex within monsoonal flows in this region. The focused rotating storm developed in a low-shear environment (not ideal for a supercell) under the interactions of three types of air masses under the influence of the land–sea contrast, monsoon, and storm cold outflows. Based on rapid-scan X-band phased-array radars and numerical simulations, this intersection zone is characterized by local enhancements of ambient vertical vorticity and convergence. The modeling reproduced two mesovortices that are in close proximity in time and space to the realistic mesovortices. Results from sensitivity experiments suggest that the unique topography plays an essential role in modifying the vorticity budget during the mesovortex formation. While there is a high likelihood of an upcoming storm evolving into a rotating storm over the intersection zone, the simulation's accuracy is sensitive to the local environmental details and storm dynamics. The strengths of cold pool surges from upstream storms may influence the stretching of low-level vertically oriented vortex and thus the mesovortex formation. These findings suggest that the trumpet-shaped coastline is an important component of mesovortex production during the active monsoon season.

The Dabie Vortex (DBV) is a critical mesoscale vortex frequently impacting central and eastern China, closely linked to heavy rainfall and severe convective weather. Under favorable conditions, DBV can spawn tornadoes, causing significant damage. However, the characteristics and formation mechanisms of tornadoes within DBV contexts remain poorly understood. This study investigates historical DBV activity and tornado events from 2006 to 2020. Results reveal that 19 tornadoes occurred amid DBV background, primarily concentrated in Henan, Hubei, Anhui, and Jiangsu provinces. DBV-related tornadoes accounted for up to 40.0% of total tornadoes in northern Anhui and over 20% in central and northern Jiangsu. Approximately 84.2% of these tornadoes occurred during the development and maintenance phases of DBV, with 73.7% forming in the southeastern quadrant. Compared to typhoon-induced tornadoes, DBV tornadoes exhibited higher convective available potential energy and stronger vertical wind shear, with larger Supercell Composite Parameter (SCP) but similar Significant Tornado Parameter (STP) values. Tornadoes during the DBV development phase occurred under weaker thermal but stronger dynamic instability, while those in the maintenance phase displayed the opposite pattern. These findings highlight that DBV can provide favorable thermal and dynamic environments for tornado formation, with significant variations across different DBV life stages.

Recent studies have pointed out the two primary mechanisms by which significant near-ground vertical vorticity forms that is intensified through stretching to reach tornado intensity near-ground rotation. They are, respectively, (1) the downdraft mechanism, where the tilting of horizontal vorticity in descending air occurs to create vertical vorticity before the nadir of the descending parcel is reached; and (2) the in-and-up mechanism, where near-ground horizontal vorticity is abruptly tilted into the vertical at the bottom of tornado vortex by a strong updraft gradient, and the vertical vorticity thus created is then intensified through stretching. The primary source of the horizontal vorticity to be tilted into the vertical is another important question; baroclinic generation of horizontal vorticity by horizontal buoyancy gradients within the storm and the generation of horizontal vorticity by surface drag are the two primary possible sources. In an attempt to answer the above questions, this study analyzes a tornadic supercell storm simulation in which four tornadoes form successively and evolve through their life cycles. Surface drag typical of land surfaces is included in the simulation with a 50 m horizontal grid spacing. The results show that both the downdraft and in-and-up mechanisms contribute to vorticity generation in the pre-tornadic and mature stages of the first tornado, while only the in-and-up mechanism is observed during the pre-tornadic stage of the first tornado. As the simulated tornado evolves, continuous parcel tracking reveals a shift in dominance from the in-and-up mechanism to the downdraft mechanism, likely due to the development of stronger post-tornadogenesis near-surface convergent flows that draw in more compensating downdraft air from higher altitudes. Our results indicate that the horizontal vorticity tilted into the vertical direction and stretched to reach tornado intensity is predominantly generated by surface drag during both pre-tornadic and mature stages of all four simulated tornadoes.

Tornadoes associated with tropical cyclones cause significant damage in various countries, making their understanding and numerical forecast are a crucial challenge. This study aims to forecast tornadoes by a numerical model and conducts high-resolution numerical simulations at a horizontal resolution of 80 m in across the entire tropical cyclone using large-scale parallel computing on the Fugaku supercomputer.The case study focuses on the tornado outbreak that occurred in the Kyushu region of Japan in association with Typhoon ShanShan 2024. This event was one of the most significant tornado outbreaks in Japan, with over ten instances of severe wind damages, including five tornadoes classified as the Japan Enhanced Fujita Scale (JEF) 2, with estimated wind speeds exceeding 65 m/s.Numerical simulations were conducted using the Cloud Resolving Storm Simulator (CReSS, Tsuboki 2023), including a horizontal 500 m grid simulation and a one-way nested horizontal 80 m grid simulation, both of simulation covering the entire typhoon including outer rainband.For this experiment, 8,192 nodes of the Fugaku supercomputer (approximately 5% of the total nodes) were utilized. To accelerate computation in large-scale parallel processing, simulation processing mapping for Fugaku's server network structure was optimized and overlapped execution of computation and file output were implemented. As a result, mini-supercells continuously formed in the typhoon’s outer rainbands, and strong vertical vorticity (>0.5 /s) on the tornado scale was detected in the hook echo regions of multiple mini-supercells. Additionally, the 4-hour forecast time simulation and united files output was completed in approximately 1.5 hours elapsed time. These results indicate that predicting tornado forecast in associated with tropical cyclones are feasible with a sufficient lead time of more than two hour.

Recently, a new large-scale Ward-type tornado simulator had been built at Central South University, CSU-TW5. Three-dimensional velocity and surface pressure on the ground were measured to investigate the vortex characteristics under different fan speeds and swirl ratios. The measured vortices were compared with those of real tornado vortices and numerical models proposed by previous researchers to check the performance of the simulator. Results showed that CSU-TW5 successfully reproduced the main characteristics of tornado-like vortices as other simulators, such as WindEEE, VorTECH, and ISU simulator. The fan speed mainly impacts wind speed magnitude, scarcely affecting vortex structures. The swirl ratio has a significant influence on the vortex structures as reported by previous researchers. The simulated results of velocity and surface pressure are similar to those of real tornadoes. The vortex structure apparently transforms from a narrow one-celled vortex to a two-celled vortex when the swirl ratio rises from 0.18 to 0.87. A downdraft appears at the upper vortex core when the swirl ratio is 0.42 and moves downward when the swirl ratio increases. It reaches the location near the ground as the swirl ratio increases to 0.87. For the tornado-like vortex of CSU-TW5, the distribution of tangential velocity at the height of the maximum value is relatively closer to the Burgers-Rott and Sullivan models; the distribution of radial velocity at the height of the maximum value is generally closer to the Baker model, and it becomes closer to the Kuo-Wen model inside the core region when the swirl ratio is 0.87; the axial velocity at the height of the maximum value matches well with the Baker model under smallest swirl ratio. However, as the swirl ratio increases, the downdraft region gradually becomes larger, causing an obvious difference in the velocity distribution between the experimental results and the numerical models.

The North Atlantic Oscillation (NAO), the dominant mode of atmospheric variability in the North Atlantic region, plays a crucial role in weather and climate. To investigate when the NAO emerged and how it evolved over geological timescales, we analyzed time-slice paleoclimate simulations during the breakup of the supercontinent Pangea, starting 160 million years ago (Ma). Our findings indicate that a present-day-like NAO mode gradually formed between 80 Ma and 60 Ma, driven by the expansion of the North Atlantic Ocean and the enhanced land-ocean contrast. This expansion led to a regime transition in Northern Hemisphere winter circulation, characterized by a westward shift of the North Atlantic jet, a strengthening of the North Atlantic high pressure and storm track, and the emergence of NAO-like variability. The confluence of orographic effects of the Rocky Mountains also contributed to the strengthening of the NAO. This study depicts the evolutionary history of the NAO over geological time and reveals its coherent relationship with the evolution of continents and orography.

In this study, we compare the performance of climate models from the two most recent CMIP phases (CMIP5 and CMIP6) in simulating the Atlantic Meridional Mode (AMM). The focus of our presentation is on the ability of the models to correctly simulate the air-sea coupling during the AMM events and the external teleconnections that influence the development of the AMM. The analyses show the performance of the CMIP5 and CMIP6 models and discuss the improvements. The results do not show a significant difference between best models ensemble (BMME) and worst models ensemble (WMME) in terms of air-sea coupling but BMME tends to have stronger forcing from the wind to SST than worst models, particularly for CMIP6. Regarding external influence, there is a slight improvement in all MME for CMIP6 compared to CMIP5 in simulating the influence of the North Atlantic Oscillation (NAO) on AMM development. The modeling results from both CMIPs confirm the importance of NAO as an important external driver of the AMM.

The transition of a wildfire from a wind driven to plume driven fire presents numerous risks to firefighters, supporting emergency services and the public, including the potential for loss of infrastructure, buildings and lives, as well as significant environmental devastation. Plume driven wildfires typically exhibit extreme fire behaviour including deep-flaming, long-range ember transport and spotting, extreme winds that may include fire tornados, and lightning. The transition to a plume driven wildfire has been found to often occur when fire plumes connect with elevated regions of higher ambient humidity above a well-mixed surface layer. Here we report on the use of the Weather Research and Forecasting (WRF) model to show through a case study approach of a wildfire that forced channelling and topographic lifting of surface winds by mountainous terrain likely coupled a wildfire plume to an elevated layer of higher ambient humidity. This resulted in pyro-cumulus development with ember transport and spotting leading to the rapid growth of the wildfire and causing deep-flaming. This enhanced further the wildfire-atmosphere coupling with deep pyro-cumulus observed during the passage of a trough over the fireground. Results highlight the importance of understanding the influence of mountainous terrain on wildfire - atmospheric interactions, and its ability to trigger extreme plume driven wildfire behaviour.            

A warming climate can alter the atmospheric circulation across the globe, including the low-latitudes due to changing intensification and extensification of Hadley circulation. For instance, the perturbation in descending branch of Hadley circulation can lead to significant change in the climatic conditions, especially in the arid regions under the subtropical high-pressure cell. The perturbed atmospheric circulation could intensify extreme climatic conditions, particularly in Central Asia, one of the driest areas in the world. The Central Asia faces significant threats to water resources, agriculture, and livelihoods due to climate change. Despite these growing concerns, the knowledge of changing climate impacts on the Central Asian climates has been built on few studies. This study addressed this gap by analyzing long-term climate trends (1991–2024) in Central Asia, including Uzbekistan and its surrounding regions, with ERA5 reanalysis data. During the boreal summer (June through August, JJA), significantly increasing temperature and decreasing precipitation were prominent in the western Uzbekistan compared to the surrounding region. The results revealed that rising temperatures were significantly associated with increasing geopotential heights at both the 850 hPa and 500 hPa levels, suggesting a strengthening of the subtropical high-pressure system. Additionally, the strengthened subtropical high was unfavorable condition for precipitation process, resulting in hotter and drier climates in the western Uzbekistan. A vertical velocity of winds further supported this finding by revealing enhanced subsidence in the lower troposphere, intensified by the subtropical high as a key driver of the hotter and drier conditions. These results suggested that the intensification of the subtropical high can amplify summer warming and dryness in the western Uzbekistan through the increased descending air under enhanced geopotential height over the Central Asian region. This study emphasized the need for sustainable adaptation strategies to mitigate socioeconomic impacts under the expected hotter and drier conditions in Central Asia.

A prevailing hypothesis for generating multi-year La Niña compared to Single-year La Niña is the strong discharge of Ocean heat Content associated with preceding super El Niño. And this hypothesis is incomplete because more than half of the multi-year La Niña events are not followed by a super El Niño. Here wo demonstrate that fundamental cause of two types lies on the decay rate of equatorial SSTA in the boreal spring.

Column relative humidity (CRH) is a crucial factor in the initiation of convection and precipitation. In this study, we introduce a new indicator, integrated relative humidity (IRH), to quantify the saturation levels from the surface to a specified pressure level. We find that IRH can elucidate the relationship between the humidity in different tropospheric layers and convection during precipitating events. We have analyzed two 20-year datasets collected from collocated radiosonde and rain gauge stations in the Hong Kong region, categorizing the data into four states, i.e., no-rain, before-rain (6 hours prior), during-rain, and after-rain (6 hours post) and used kernel density estimation to compute the probability density functions of IRHs for each of the state. We found that upper tropospheric IRH is equivalent to CRH and ~0.7 is a critical threshold to discriminate rain and no rain events, consistent with previous research. Middle tropospheric IRH emerges as a more suitable indicator compared to other IRHs, particularly when set to ~0.85, as it includes 81% of during-rain state while excluding 84%, 46%, and 48% of the other three states, respectively. Upper free tropospheric IRHs all exhibit a cyclical pattern of ascent and subsequent descent across the no-, before-, during-, and after-rain states, reaching peak saturation levels during rain events. Notably, upper free tropospheric IRHs of before-rain are consistently higher than those of after-rain, the former being moistened by convective activity and the latter being dried by precipitation. Lower tropospheric IRHs of before-rain are evidently lower than those of after-rain, particularly near surface, suggesting that the moistened states of post-rain are attributed to boundary layer processes (e.g., evaporation) following precipitation. The variations of IRH across no-, before-, during-, and after-rain states indicate that the close coupling of atmospheric moisture and precipitation. The IRH serves as a useful marker of precipitation-moisture interactions.

Air pollution is a key global environmental problem raising human health concern. It is essential to comprehensively assess the long-term characteristics of air pollution and the resultant health impacts. We first assessed the global trends of fine particulate matter (PM_2.5) during 1980-2020 using monthly global PM_2.5 reanalysis, and evaluated their association with climate variability. We then estimated the PM_2.5-attributable premature deaths using integrated exposure-response functions. Results show a significant positive increasing trend of ambient PM_2.5 in the four decades due to increases in anthropogenic emissions. Ambient PM_2.5 caused a total of ~135 million premature deaths globally during the four decades. Occurrence of air pollution episodes was strongly associated with climate variability which are associated with up to 14% increase in annual global PM_2.5-attributable premature deaths.

Exposure to air pollution and excessive heat during hot-and-polluted episodes (HPEs) may synergistically cause higher health risks globally. Nevertheless, long-term global spatiotemporal characteristics of HPEs and their health impacts remain unclear. Herein, we conducted statistical analyses using reanalysis data of fine particulate matter (PM_2.5) and climate together with our derived concentration-response function for HPEs to assess global HPE variations from 1990 to 2019, and to estimate the PM_2.5-associated premature mortality during HPEs. Our results reveal that HPE frequency increased significantly globally. HPE PM_2.5 intensity in Global North continuously increased, overpassing Global South after 2010, indicating a recurred risk of air pollution under climate change in Global North after several years of emission control endeavors. Globally, we estimated approximately 694,440 (95% CI: 687,996–715,311) total mortalities associated with acute PM_2.5 exposure during HPEs from 1990 to 2019, with the Global South accounting for around 80% of these deaths. Among the most vulnerable 15 countries, India had by far the highest mortality burden, and the United States, Russia, Japan, and Germany were particularly highlighted as having higher burdens within the Global North. Our findings highlight the importance of considering environmental inequality between Global North and Global South, and co-benefits of air pollution-climate change mitigation during policymaking processes.

Recent evidence has shown the increasing trend of tropospheric ozone (O₃) in Southeast Asia (SE Asia). Mitigating O₃ pollution in SE Asia has become important and urgent. While the nonlinear O₃ chemistry makes policy-making complicated, the O₃ formation regime and O₃ response to different emissions have rarely been assessed in SE Asia. Furthermore, the O₃-attributable health impacts in SE Asia under future emission scenarios have yet to be quantified. Herein, we applied the regional chemical transport model with the High-order Decoupled Direct Method (HDDM) to simulate the O₃ sensitivity to precursor emissions in SE Asia, and then projected the health benefits under future Shared Socioeconomic Pathways (SSP) emission scenarios, providing policy suggestions for mitigating O₃ pollution and its health impacts. Our results show O₃ in urban areas (i.e., Singapore, Jakarta, Kuala Lumpur, Bangkok, and Ho Chi Minh City) was sensitive to both nitrogen oxides (NO_x) and volatile organic compounds (VOCs) emissions, and synergistic NO_x and VOCs control is thus essential. Suburban, rural, and sea areas were under a NO_x-limited regime, suggesting the high effectiveness of controlling NO_x over these areas. Compared with the health impacts in baseline year (2019), the annual total O₃-attributed premature mortality under the business-as-usual emission scenario (SSP245) is projected to reduce by 22k (47%) by 2050 due to the future NO_x emission reductions in power generation, industrial process, and transportation. Most of the health benefits will happen in Indonesia, Philippines, Vietnam, and Thailand. The sustainable emission scenario (SSP126) is projected to avoid 36k annual O₃-attributed premature mortalities by 2050 due to its more stringent NO_x reductions in shipping, transportation, and industrial process. SSP370 and SSP585 are projected to increase the O₃-attributable premature mortality by up to 33k because of the rising NO_x emissions.

Air pollution has become a significant environmental and public health concern in tropical urban areas, influenced by rapid urbanization, industrial activities, and changing weather patterns. Accurately predicting air pollutant concentrations is essential for reducing health risks, guiding policy decisions, and supporting sustainable urban planning. However, traditional models often struggle to account for the complex spatial and temporal variations in air quality data. This study introduced an LSTM-Diffusion model to enhance air pollution forecasting and create high-spatial resolution local air quality maps. The model integrated historical air quality data, meteorological factors, rainfall, and other pollutants to improve prediction accuracy. By combining Long Short-Term Memory (LSTM) networks for time-series analysis with diffusion-based techniques for spatial interpolation, this approach effectively captured local air pollution dynamics while addressing key challenges such as data gaps, seasonal fluctuations, and spatial variability. To ensure reliability, the model’s performance was evaluated using standard metrics, including Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and R-squared (R²). Additionally, explainable AI techniques were applied to enhance transparency, making it easier for policymakers and urban planners to interpret the results and implement data-driven solutions. The spatial prediction outcomes were visualized as high-resolution pollution maps, supporting real-time air quality monitoring and targeted intervention efforts. This study demonstrates the potential of LSTM-Diffusion models as a practical solution for real-time air quality prediction and spatial mapping in urban environments worldwide. By providing a scalable, AI-powered framework for air quality forecasting, this research is expected to contribute to broader environmental and public health strategies. The findings offer valuable insights into optimizing pollution control measures, refining air quality regulations, and developing long-term mitigation strategies.

Accurate air pollution prediction at high spatiotemporal resolution is crucial for urban air quality management and public health assessment, particularly in high-density cities. This study developed an integrated modeling framework that combines land-use regression (LUR) with machine learning to predict multiple air pollutants (PM_2.5, O₃, NO₂) at daily, 500m resolution across urban cities including Hong Kong, Seoul, Taipei, etc. These cities exhibit diverse geographical, meteorological, and emission patterns, which complicate pollution modeling. By incorporating a wide range of geospatial, building morphology, meteorological, and traffic-related variables, our model adapted to different urban environments and improved predictive accuracy. Additionally, SHapley Additive exPlanations (SHAP) were applied to quantify source contributions and assess predictor importance in each city, providing a detailed evaluation of pollution drivers. The proposed framework enhances the interpretability of machine learning-based air pollution models and offers a scalable solution for multi-city air quality assessment. These advancements are particularly valuable for epidemiological research, enabling more precise exposure assessments to support public health policies and disease burden analysis.

Air pollution was reported to adversely affect human health. Recent evidence indicated rising ground-level ozone (O₃) levels across Southeast Asia (SE Asia). To mitigate O₃ pollution that is formed through complex photochemical reactions, a full understanding of the sensitivity of O₃ to precursor emissions [i.e., volatile organic compounds (VOCs) and nitrogen oxides (NO_x)] and the relative contribution of precursors is required but remains unclear. This study applied an adjoint sensitivity model to assess the source-receptor (S-R) relationship between O₃ concentration, NOx, and VOCs, and each emission source's contribution over different receptor regions in SE Asia. The process analysis technique was utilized to further characterize O₃ formation. Our findings reveal that vertical diffusion and gas-phase chemistry were the primary contributors to ground-level O₃ formation, while significant depletion occurred due to strong dry deposition and aqueous and cloud processes. Subsequent analysis indicates a predominant NO_x-limited O₃ formation regime across SE Asia due to substantial biogenic VOC emissions, with notable exceptions in Singapore, Jakarta, Bangkok, and the Malacca Straits, where anthropogenic NO_x emissions were significant. In SE Asia, NO_x is identified as the primary contributor to the surface O₃, whereas VOCs contributed positively to VOC-limited regions but negatively to NO_x-limited areas. Additionally, our results indicate that regional and super-regional transboundary air pollution significantly contributed to O₃ levels in SE Asia, accounting for more than 50% of the O₃ concentration in each country in SE Asia. These insights are critical for formulating more effective regional O₃ mitigation strategies, suggesting a joint effort between regional and local policymakers in SE Asia for the effective reduction of NO_x and VOC emissions. 

Climate change and associated human response are supposed to greatly alter surface ozone (O₃), an air pollutant generated through photochemical reactions involving both anthropogenic and biogenic precursors. However, a comprehensive evaluation of China’s O₃ response to these multiple changes has been lacking. We present a modelling framework under Shared Socioeconomic Pathways (SSP2-45), incorporating future changes in local and foreign anthropogenic emissions, meteorological conditions, and BVOCs emissions. From the 2020s to 2060s, daily maximum 8-hour average (MDA8) O₃ concentration is simulated to decline by 7.7 ppb in the warm season (April-September) and 1.1 ppb in non-warm season (October-March) over the country, with a substantial reduction in exceedances of national O₃ standards. Notably, O₃ decreases are more pronounced in developed regions such as BTH, YRD, and PRD during warm season, with reductions of 9.7, 14.8, and 12.5 ppb, respectively. Conversely, in non-warm season, the MDA8 O₃ in BTH and YRD will increase by 5.5 and 3.3 ppb, partly attributed to reduced NO_x emissions and thereby weakened titration effect. O₃ pollution will thus expand into the non-warm season in the future. Sensitivity analyses reveal that local emission change will predominantly influence future O₃ distribution and magnitude, with contributions from other factors within ±25 %. Furthermore, the joint impact of multiple factors on O₃ reduction will be larger than the sum of individual factors, due to changes in the O₃ formation regime. This study highlights the necessity of region-specific emission control strategies to mitigate potential O₃ increases during non-warm season and under climate penalty.

The onset of the South China Sea (SCS) Summer Monsoon (SCSSM) is usually in May, strongly influencing the beginning of the wet season over the surrounding regions, such as South China and the Philippine Islands. The onset date is significantly affected by the El Niño-Southern Oscillation (ENSO), which manifests as two major types: the Eastern-Pacific (EP) and Central-Pacific (CP) patterns. However, the impacts of these patterns on the SCSSM onset await quantification. Here, using a novel statistical metric, binary combined linear regression, we find a dominant role of the CP pattern rather than the EP pattern, as a moderate CP El Niño/La Niña tends to cause SCSSM onset delay/advance by around 4.7/5.1 days in its decaying year. The anomalous SST structure of CP El Niño leads to persistent anomalous anti-cyclonic winds over the western North Pacific from April to June, suppressing the local convection and impeding the SCSSM winds. In comparison, the EP El Niño impact decays quickly in May and hardly affects the SCSSM onset. This, hence, weakens the relationship between the SCSSM onset and ENSO. These results help improve the SCSSM onset prediction by involving diverse ENSO impacts. 

This study investigates the dependence of influence of environmental factors on the intensification of tropical cyclones (TCs) upon the TC intensification rate (IR) and location in the western North Pacific (WNP) during 1980-2021. Slow intensification (SI) and rapid intensification (RI) TCs are classified into three clusters using the K-means clustering method. The center location of cluster 1 to 3 TCs shifts from west to east. The impacts of vertical wind shear change little with IR and location and from summer to autumn. The impacts of mid-level humidity and low-level vorticity are generally larger for RI than SI TCs in both summer and autumn. The impacts of mid-level humidity are greater in autumn than in summer for RI TCs. The three clusters of SI and RI TCs are associated with different tropical Indo-Pacific sea surface temperature (SST) anomalies. Southeastern Indian Ocean (SEIO) SST anomalies mainly influence intensification of TCs in the western WNP in both summer and autumn. Equatorial central Pacific (ECP) SST anomalies play a large role in the intensification of TCs in the eastern WNP in both summer and autumn. Equatorial eastern Pacific (EEP) SST anomalies contribute to the intensification of RI TCs in the western WNP in autumn. WNP SST anomalies play a role in the intensification of RI TCs in autumn. RI tends to occur when SST anomalies in two regions work together. Present results have important implication for understanding the intensification of TCs in the WNP in different locations and during different seasons.

The tropical cyclones (TCs) often cause intense rain and destructive winds. While these catastrophic weather conditions capture our attention, the less-known impact of TCs remains overlooked. This study reveals that TCs have a notable suppressive effect on monsoonal rainfall in southern China when they traverse the South China Sea. This phenomenon can be attributed to the influence of these mesoscale disturbances on the quasi-stationary, large-scale monsoonal circulation, which alters the moisture pathway. On the other hand, TCs in the tropical western North Pacific can significantly suppress rainfall over the Maritime Continent and its surrounding seas. The rainfall suppression is attributed to the lower tropospheric southwesterly anomalies to the south of TCs, which result in moisture divergence over the Maritime Continent. Additionally, the upper-tropospheric equatorward outflows of TCs also promote subsidence and suppress rainfall.

This study identifies 86 active events and 66 break events of the western North Pacific summer monsoon (WNPSM) from 1979 to 2020. These active and break events exhibit sharp contrast in large-scale convection and circulation fields. During the active period, deep convection almost covers the entire WNP domain (0°–25°N, 120°–170°E), accompanied by the strengthening of a local monsoon trough and monsoon westerlies extending from the Arabian Sea to the WNP. In contrast, during the break period, along with weakened monsoon westerlies, the subtropical high replaces the monsoon trough to dominate the WNP domain, leading to the disappearance of deep convection. Results about the convection evolution at multiple timescales indicate that active/break events are primarily contributed to by the wet/dry phases of intraseasonal oscillations (ISOs). The phase transitions of ISOs are nearly symmetric between active and break events. However, these two types of events present notable asymmetric features in convection anomalies relative to climatology, which can be attributable to the modulation of the seasonal mean component. Specifically, both active and break events correspond to anomalously strong seasonal mean convection, which amplifies (weakens) the convection enhancement (suppression) during the active (break) period and results in normal (enhanced) convection during adjacent periods. The preferred occurrence of active and break events in strong summer monsoons is because related mean backgrounds, including increased low-level positive vorticity, specific humidity, and easterly vertical shear, facilitate the intensification of ISOs. These mean backgrounds are likely induced by SST cooling in the north Indian Ocean and warming in the equatorial central Pacific during the simultaneous summer.

The East Asian monsoon cycle is identified by the western Pacific-Indian Ocean regional low-level circulation weather types (WTs) of daily 850hPa winds from 1979 to 2024 based on the ERA5 data. The intensity of the East Asian summer and winter monsoons are well represented by different kinds of WT evolution. We found that a strong winter monsoon tends to follow a strong summer monsoon. However, a strong or weak summer monsoon has no clear relationship with its preceding winter monsoon. The in-phase relationship between the summer and winter monsoon is driven by the ENSO. Furthermore, about one-third of 45 years show a clear biennial type of oscillation in the summer monsoon, which means the monsoon intensity swings between strong and weak during two successive summers. The transition from weak to strong summer monsoons is associated to El Niño, whereas the transition from strong to weak summer monsoons has little to do with La Niña.

The monsoon trough (MT) in the western North Pacific (WNP) exhibits strong synoptic and intraseasonal variabilities. The origins of the variabilities were investigated through idealized numerical model experiments and a theoretical model. To demonstrate to what extent the synoptic and intraseasonal variabilities in MT arise from signals outside of the region, idealized numerical model experiments in the presence and the absence of propagating synoptic and intraseasonal signals from the outside of the region were carried out. The simulation results indicate little change in the intensity and structure of the synoptic and intraseasonal variabilities over MT. This implies that the origin of these variabilities arises from internal atmospheric dynamics in the region. A 2.5-layer theoretical model was further constructed, in which an idealized background mean state derived from the observed moisture and zonal wind profiles in the MT region is specified. The model extends the traditional 2-level quasi-geostrophic model framework by including a prognostic moisture tendency equation and an interactive planetary boundary layer. The eigenvalue analysis of this theoretical model shows two most unstable modes. The first has a preferred zonal wavelength of 2700 km and a westward phase speed of 1.5 m s^-1, consistent with the observed synoptic mode characteristics. The second has a much larger (12000 km) zonal wavelength and near-zero zonal phase speed, resembling the observed intraseasonal mode characteristics.

Double low-level jets (DLLJs) are defined as the simultaneous presence of boundary layer jets (BLJs) and synoptic-system-related LLJs (SLLJs) in a region. Previous case studies have shown that DLLJs play crucial roles in coastal heavy rainfall in South China. However, the climatological behaviors of DLLJs are not well understood. This study examines the diversity and multi-scale mechanisms of DLLJs in South China during the warm season and their impacts on precipitation. Two high-incidence DLLJ regions are identified: one in the Beibu Gulf, mostly in April, and another in the northern South China Sea (NSCS), primarily in June. Both regions exhibit nocturnal peaks, though the NSCS peak is delayed by approximately 2 hours due to a stronger westerly jet component. Nocturnal peaks of SLLJs are mainly driven by global pressure tides and large-scale land winds due to sea-land temperature contrasts, while BLJs nocturnal peaks are associated with inertial oscillations and interactions between nighttime pressure tides and boundary layer thermodynamics. DLLJs are classified into three types based on the spatial relationship between BLJs and SLLJs: displaced type (Type-D), connected type (Type-C), and overlapped type (Type-O). The occurrence of these types varies throughout the warm season, influenced by westerly troughs, monsoonal flows, and tropical systems. In Type-O DLLJs, weaker daytime friction over oceanic regions contributes to an afternoon sub-peak. Key systems influencing DLLJs include anomalous high pressure over the South China Sea or Southeast China and land low pressure. Specifically, Type-D is associated with abnormal high pressure, Type-C is governed by both land low pressure and high pressure over the NSCS, and Type-O is influenced by land low pressure. DLLJs play an important role in shaping regional precipitation distribution in South China by regulating mid-low-level moisture transport, boundary layer convergence, and positive vorticity. Notably, variations in DLLJ types contribute to regional precipitation differences.

Dust storms are increasingly frequent due to climate change and altered land-use patterns. However, our current understanding of their impacts to the stability of sandy lands or deserts in source areas during historical periods is limited. The Mu Us desert, situated at the northern margin of the East Asian summer monsoon (EASM), delineates the boundary between inhospitable desert and human habitation. The expansion and contraction of the Mu Us desert significantly influence the livability of the surrounding environment. Therefore, understanding the dynamic of sand dune activities is of great scientific and social interest. Previous studies proposed that improved vegetation, benefited from strengthened EASM, is a critical factor for sand dune to be stabilized during a warm climate period. Conversely, vegetation collapse is believed to trigger the reactivation of sand dunes.. Here, based on directly dating results from the hinterland of the Mu Us desert, we unexpectedly find that sand dunes accumulated during the Little Ice Age and have remained stable, with no advancing mobility in the past hundred years. The current landscape of the Mu Us desert, characterized by rolling barchan dune chains, has taken its modern form under the influence of briefly prevailing westerlies. Climate model simulation suggest that, despite reduced summer precipitation, cold climate intervals favor the preservation of sand moisture due to high winter snowfall and low summer evaporation, significantly contributing to the stabilization of sand dunes within the Mu Us desert. Therefore, we conclude that large-scale landscape changes in the Mu Us Desert are not driven by variations in the EASM as previously thought.

The International Methane Emissions Observatory (IMEO) was launched in 2021 at the G20 summit by the United Nations Environment Program (UNEP). UNEP’s IMEO exists to provide open, reliable, public, policy-relevant data to facilitate actions to reduce methane emissions. UNEP, through IMEO, aims to fill gaps in knowledge and refine global understanding of the location and magnitude of methane emissions across different anthropogenic sectors. As countries and industry establish ambitious mitigation targets, accurate and measurement-based emission estimates are critical to accelerate emission reductions and assess progress by tracking changes in emissions over time. UNEP’s IMEO is collecting and integrating diverse methane emissions data streams, including from satellites, science studies and measurement-based industry reporting to establish a global, centralized public record of empirically verified methane emissions. In this presentation, we will provide insights from measurement campaigns across the world with a focus upon coal mine methane studies. Preliminary results will be presented from the evaluation of methane emissions from underground coal mining in Poland and surface coal mining in Australia.

The Asia-Pacific region is one of the world’s largest liquefied natural gas (LNG) supply chains with gas from Australia and South East Asia transported primarily to buyer countries such as Japan, South Korea and China. The life cycle emissions of this supply chain are dependent on the extent of methane emissions in the supplier countries. Estimates of methane emissions from liquefaction terminals have not been well-characterized through measurement studies, leading to large uncertainties in life cycle analyses. Here, in work conducted under UNEP’s International Methane Emissions Observatory (IMEO), we present results from a comprehensive measurement-based analysis of methane emissions from LNG terminals in Australia, the largest exporter in the region. We will discuss how airborne measurement approaches can be used to evaluate industry-reported emissions. These site-level emission results can then be used to determine the country-wide methane emissions intensity of LNG liquefaction to better inform both buyer and supplier countries. Finally, we will discuss how this work ties into the broader set of science studies and initiatives of UNEP’s IMEO and the need for greater use of measurement-based methods to determine methane emissions across the Asia-Pacific region.

Methane contributes about 45% to current warming, and reducing methane emissions is essential to keeping global temperatures below 1.5C and minimizing overshoot. Recently, a wealth of new methane satellite data is coming online. This presentation will discuss the impact of these satellites, and discuss useful data products in the context of a roadmap for reducing methane emissions. 

Accurate detection and quantification of methane point sources are essential for mitigating climate change. Prompt detection of abnormal methane (CH4) emissions in Oil & Gas (O&G) field, coal mine and liquefied natural gas terminal would enable action for climate change mitigation. CH4 emissions from O&G facilities predominantly emanate from key infrastructures, including wellheads, compressor stations, tank batteries, pipelines and flares, forming easily recognizable "point-source emissions". Spaceborne hyperspectral imaging spectrometers have recently emerged as a promising approach for mapping methane point-source emissions. These instruments, utilizing the radiance in the SWIR range, can discern subtle signal changes from methane absorption. Recent advancements in hyperspectral satellites have demonstrated their potential to map and quantify point-source methane emissions. Here we demonstrate the ability of a few Chinese hyperspectral satellites (GF5-01A, GF5-02, and ZY1-02E) to map methane point-source emissions from Oil & Gas sector in multiple basins of North America and Middle East, and from coal mine in China. Meanwhile, we compare the spectral and radiometric performance of these satellites instrument with other existing spaceborne hyperspectral imaging spectroscopy for methane point sources mapping, including PRISMA, EnMAP, and EMIT. Our study highlights the importance of considering the trade-offs between spectral resolution, SNR, and spectral calibration stability when designing and selecting hyperspectral imaging missions for methane point-source emissions detection. These results highlight the potential value of hyperspectral imaging spectroscopy in enhancing bottom-up inventories, refining regional methane estimates, and reducing uncertainties.

Monitoring greenhouse gas (GHG) emissions is critical for informing policies aimed at mitigating global warming. Spaceborne remote sensing offers a powerful solution by providing global coverage of GHG concentrations. Hong Kong University of Science and Technology’s Multi-Spectral Imaging Carbon Observatory (MUSICO) is scheduled to be deployed on the Chinese Space Station (CSS) in May 2026. MUSICO is designed to measure carbon dioxide (CO₂) and methane (CH₄) at a spatial resolution of 100 meters, covering a field size of 50 km. It targets a nominal column concentration measurement precision of 2 ppmv for CO₂ and 20 ppbv for CH₄, enabling the detection and monitoring of emissions from point sources. The instrument features three static Fabry-Perot imaging spectrometers with a spectral resolution of 0.01 nm and a nominal SNR no less than 200. These spectrometers cover an oxygen absorption band at 0.76 µm and two near-infrared channels at 1.61 µm and 1.64 µm for CO₂ and CH₄ measurements, respectively. To ensure accurate retrieval of GHG concentrations, MUSICO also includes aerosol measurements using bands at 0.38 µm, 0.44 µm, and 0.865 µm. Once deployed on the CSS, MUSICO will be capable of observing targets within latitudes of ±42 degrees. This presentation introduces the mission concept, outlines the instrument's design and capabilities, and reports on its status with test results.

Imperfectness of the state-of-the-art intensity forecasting of tropical cyclones (TCs) necessitates independent rapid intensification (RI) prediction schemes. Here we report one derived with an explainable artificial intelligence Wide Learning (WL). The scheme, named Wide Learning-based TC Rapid intensification Prediction Scheme (WRPS) Version 1 (WRPS1), predicts RI in the Western North Pacific by using twelve predictor variables representing environmental conditions and the state of TCs. Its prediction is based on a score that is a linear combination of whether or not (1 or 0) joint conditions on ranges of multiple variables are met, which is reproducible without WL. Relying on joint conditions allows WRPS to handle nonlinearity and inter-dependence among predictors, and the simpleness of the conditions provides explainability. A method to map an RI-prediction score to its probability is proposed and is used in WRPS. It is suggested that handling predictors favorable to RI when having moderate values, such as the current intensity, is a key for good RI prediction. It is demonstrated that quantifying the contribution of each predictor to the WRPS score helps one elucidate how the predictors jointly facilitated or hindered RI for each prediction case. The performance of WRPS1 is compared with RI predictions using the linear discriminant analysis, and WRPS1 is shown to perform well without using track predictions. The multiple linear regression analysis, which is customarily used for intensity prediction but not for RI prediction, is shown to perform well if the fraction of RI cases is increased when conducting regression.

Accurate prediction of tropical cyclones is crucial yet remains a challenging task. Innovative satellite observing systems provide essential data for advancing hurricane research and forecasting. This talk will present an updated progress in satellite data assimilation by the author's research team, with a particular focus on enhancing numerical simulation and prediction of tropical cyclones.The presentation will highlight several key developments, including: 1) All-sky assimilation of GOES-R (NOAA's Geostationary Operational Environmental Satellites) radiances, incorporating bias correction and optimized channel selection. 2) Joint assimilation of data from two NASA-supported small satellite missions—CYGNSS (Cyclone Global Navigation Satellite System) and TROPICS (Time-Resolved Observations of Precipitation Structure and Storm Intensity with a Constellation of Smallsats)—using various configurations and integrated data assimilation techniques.The talk will discuss the significant impact of these advancements on improving the accuracy of numerical simulations of tropical cyclones. Additionally, it will cover the latest developments in advanced data assimilation methods, as well as the critical roles of observational error characteristics, background error covariances, and bias corrections in satellite data assimilation.

Deterministic TC track forecasts have become so accurate in the medium range that I expect forecasting to move in new directions with different emphases.Two new possible directions are: 1) a more explicit (and better) prediction of the surface wind field structure (e.g., intensity and the wind radii); and 2) ‘completing the forecast’ with predictions of genesis (and dissipation).Current best track data sets have several deficiencies that will limit development in these new directions; specifically, they lack:•  position/intensity/structure data in the pre/potential TC (pTC or 9X)    genesis stage.•  observational analyses of the surface wind field •  a ‘diagnostic file,’ with environmental variables known to be related to intensity change,•  track forecasts for both pTCs and TCs that include TC structureThe foundation of the super Best Track (sBT) is the ‘final’ best tracks of the two US operational forecast centers – the JTWC and the NHC. The sBT is thus global and unique in three important ways by including:•   curated pTC tracks from JTWC/NHC since 2006•   dynamical environmental variables (analysis and forecasts) come from the twice daily ECMWF ERA5 10-d forecasts.•   two satellite precipitation analyses: IMERG (NASA) and GsMAP (Japan JAXA)The initial version of the sBT covers a 16-year period 2007-2022.  Two application examples are given: 1) formation rate (9X->NN) between basins and 2) differences in vertical wind shear and precipitation between developing v non-developing pTCs.The mean formation is about 3 d in all the basins, i.e., it takes about 3 d for an INVEST to become a numbered NN TC.  Furthermore, ~ 48 h before formation the wind shear and precipitation of 9Xdev v 9Xnon depart which implies potential for genesis forecasting.beta version at:   https://github.com/tenkiman/superBT-V04and two scientific applications:  https://surperbt.blogspot.com/2023/12/intro-to-superbt.html

Tropical cyclones (TCs) occurring in the western North Pacific (WNP) bring heavy rainfall and strong winds, causing significant damage to East and Southeast Asia. To mitigate the human and material losses induced by TCs, it is crucial to improve their predictability. However, errors in TC simulations using numerical models can arise due to uncertainties in the initial conditions. This study aims to improve the initial conditions of the TC prediction model by developing the previously suggested dynamical initialization (DI) scheme through data assimilation and applying it to TC simulations in the WNP. The proposed DI scheme with data assimilation (DIDA) follows a methodology similar to the previous DI scheme, wherein the axisymmetric component of the TC vortex is spun up through a six-hourly cycle run until the second cycle run. However, compared to the DI scheme, the DIDA scheme applies additional corrections before the third cycle run by adjusting the size and intensity of the vortex in the analysis field within the cycle run to match observations. This correction is implemented using the Weather Research and Forecasting (WRF) four-dimensional data assimilation (FDDA). In this study, we conducted a 72-hour TC forecasting experiment comparing the DIDA with the DI scheme using the WRF model. In terms of improving initial TC intensity, the DI scheme required multiple cycle runs to sufficiently strengthen the initial TC intensity. In contrast, the DIDA scheme efficiently enhanced the intensity within three cycle runs. Additionally, the TC simulation results demonstrate that the performance of the DIDA scheme was comparable to that of the DI scheme. These findings indicate that the DIDA scheme effectively enhances the initial TC intensity with fewer cycle runs while maintaining accuracy close to the DI scheme.

This study assimilates the ground-based radar observation for analysis and forecast of typhoon Chanthu (2021) using the WRF-local ensemble transform Kalman filter radar assimilation system (WLRAS). To explore the optimal configuration, the radar data assimilation (DA) experiments are conducted with various setups including different model horizontal resolutions and DA frequencies. The impact of these setups on the analysis and prediction of Chanthu is examined. According to an observational study by Fang et al. (2024), Chanthu experienced rapid intensification (RI) during its northward movement off the east coast of Taiwan. They identified several factors influencing Chanthu's RI, including TC-terrain interaction and vertical wind shear. The RI feature of Chanthu is successfully captured by the forecast initialized from the WLRAS analysis, demonstrating the high capability of the RDA system. However, the RI behaviors vary with different experiment configurations. Therefore, these DA experiments are analyzed and compared with radar observations to jointly validate the factors contributing to Chanthu's RI.

Predicting the rapid intensification (RI) of tropical cyclones (TCs) is a significant global challenge. In this study, we proposed a quantitative index, Symmetric Ratio, derived from satellite observations to depict the TC’s inner-core symmetry. Using the Symmetric Ratio and other traditional environmental factors as predictors, we developed RI prediction models for the Northwestern Pacific (WNP) and North Atlantic (NA) basins using machine learning algorithms of Decision Tree, Random Forest, Light Gradient Boosting Machine, and Adaptive Boosting. Additionally, we created an ensemble RI prediction model by combining the four individual models. The independent samples of TC real-time and forecasted data from the Automated Tropical Cyclone Forecast (ATCF) and environmental data from Global Forecast System (GFS) during 2021-2023 were used to test our models. The results showed that the ensemble model exhibited detection probability (POD) values of 0.24 for both 12-hour and 24-hour RI predictions, with false alarm rate (FAR) values of 0.47 and 0.33 in the NA basin. Compared with the best deterministic model in the National Hurricane Center with a 21% POD and a 50% FAR to forecast RI in the NA basin during 2016-2020, our model achieved higher POD and lower FAR. In the WNP basin, the ensemble model achieved POD values of 0.3 and 0.41 for 12-hour and 24-hour RI predictions, with corresponding FAR values of 0.46 and 0.45. And the new index, Symmetric Ratio, was a significant variable identified by models, underscoring the importance of TC’s inner-core structure in RI processes.

The MLT vertical wind plays a crucial role in the vertical transmission of momentum, energy, and heat among different atmosphere layers. However, the vertical wind is difficult to observe due to its small magnitude of typically cm/s. Multistatic meteor radar systems can obtain spatially resolved information, such as the horizontal divergence, because they can provide much higher meteor count rates and additional viewing angles. This can then be used to calculate the vertical wind. In this work, using data from the Chinese multistatic meteor radar system observations, we compare the newly developed VVP and the 3DVAR-DIV methods. Both are 3-dimensional spatially resolved wind retrieval methods, utilizing the same routine for vertical wind retrieval from horizontal divergence. The performances of the two methods for the estimation of horizontal wind and vertical wind are similar, indicating an hourly vertical wind generally less than 1 m/s hourly vertical wind. The vertical winds from both methods are highly correlated and exhibit similar diurnal and semidiurnal features, which in turn verifies the reliability of the vertical wind retrieval method.

Buckland Park (BP) is a University of Adelaide field site located approximately 40 km North of the city of Adelaide, South Australia.  ATRAD in conjunction with the University of Adelaide operate both a Stratospheric Tropospheric (ST) radar and a meteor detection radar which share the same hardware.  Both radars operate in the VHF band at 55 MHz, with 48 kW peak power.The meteor detection radar operates as a single ‘all-sky’ folded crossed dipole transmit antenna, and 5 single crossed dipole Yagi antennas arranged in a Jones cross pattern for reception.  Angle of arrival techniques on the Jones cross pattern are well established and allow unambiguous calculation of the sky position of the returned radar signal.It has been shown VHF radars are also capable of detecting satellites in low Earth orbit, resolving the satellite’s aliased range, aliased velocity and acceleration.  We have recently extended the angle of arrival technique to apply to satellite detections, thus allowing full determination of the satellites’ position.  Determining the angular position is important as the full detection can then be used to calculate the satellites’ orbit.  In addition, the satellite detection technique can also be utilised in resolving meteor head echoes collected in the same data set.  This paper will present and discuss the adapted processing techniques for angle of arrival and meteor head echo detection.

A unique project will be undertaken in Spring 2025 called SolarMaX. This project will be a campaign taking place over a 3-5 day window, when the astronauts in the Fram2 crew will become the first humans in polar orbit. They will use their ideal vantage point to record optical observations of the aurora from space, while scientific observatories and citizen scientists take images and data from the ground.The astronauts and citizen scientists will particularly focus on photographing fragmented aurora-like emissions (fragments), streaks, and continuous emissions such as STEVE and it's recently discovered poleward counterpart. To raise awareness of these phenomena, the International Space Science Institute team ARCTICS created resources such as the Aurora Handbook and Field Guide. Aurora researchers and citizen scientists worked together to create these free online books, which provide a wealth of information on how to identify different types of aurora, photograph them, and report them for scientists to use in their research. In addition, guidelines on collaboration are provided for both scientists and citizen scientists.If these features appear during the 3-5 day orbit window, then we can compare the images from Fram2 to images of the same feature from at least 2 members of the public and/or scientific observatories on the ground. This allows for triangulation of the altitude of the features and 3D reconstruction of the structures. Observing how the features evolve from space and the ground will also allow us to piece together how the systems relate to each other and how they are coupled together. Lastly, we are interested in using RGB values from the images to calculate how much energy input to the atmosphere when these features appear, which can potentially cause atmospheric expansion and satellite drag.

Polar winter descent of NOy produced by energetic particle precipitation (EPP) in the mesosphere and lower thermosphere affects polar stratospheric ozone by catalytic reactions. This, in turn, may affect regional climate via radiative and dynamical feedbacks. NOy observations by MIPAS/Envisat during 2002--2012 have provided observational constraints on the solar-activity modulated variability of stratospheric EPP-NOy.ESA’s Earth Explorer 11 candidate Changing Atmosphere Infra-Red Tomography (CAIRT) will observe the atmosphere from about 5 to 115 km with an across-track resolution of 30 to 50 km within a 500 km wide field of view. CAIRT will provide NOy and tracer observations from the upper troposphere to the lower thermosphere with unprecedented spatial resolution. We present the science studies using WACCM-X high resolution model runs simulating a Sudden Stratospheric Warming event to assess its potential to advance our understanding of the EPP-climate link and to improve upon the aforementioned constraints in the future.

  Understanding the thermodynamic structure of the atmospheric boundary layer is crucial for enhancing weather forecasting and elucidating heavy rain processes that lead to water-related disasters. In 2024, continuous monitoring of water vapor and temperature profiles in the lower troposphere using Raman lidar began on an isolated island off the west coast of Kyushu, located upwind of a linear precipitation band that caused serious water disasters. A noteworthy feature of our lidar system is the use of a laser wavelength of 266 nm in the UV-C wavelength region. This technique offers the advantage of minimal daytime background noise due to ozone layer absorption in the stratosphere. Consequently, the accuracy of UV-C lidar remains consistent between day and night, providing highly quantitative and continuous thermodynamic profile data. These lidar systems are operated and supported partially by programs for Bridging the gap between R&D and the IDeal society (society 5.0) and Generating Economic and social value (BRIDGE). The project "Advanced Flood Forecasting Based on Innovative Integrated Meteorological Data" (Principal Investigator: Prof. Yuji Sugihara, Kyushu University), is entrusted by the Ministry of Land, Infrastructure, Transportation, and Tourism. In this study, we present the performance of LiDAR-derived thermodynamic profiles based on a comparison with collocated radiosonde measurements and discuss the variability characteristics of the temperature and water vapor profiles through several precipitation events.  

Moisture recycling is an important source of precipitation in the tropical forests of South America and Africa. Moisture is partly recycled from the tropical forests themselves (forest rainfall self-reliance) and is therefore subject to deforestation, which reduces evaporation. During the dry season, when water is already scarce, a further reduction in precipitation due to decreasing moisture recycling rates could potentially be fatal for already vulnerable ecosystems. It is therefore important to better understand the self-reliance of precipitation in tropical forests. For this reason, we present climatologies of precipitation dependence on evaporation in and from tropical forests using WAM2layers driven by ERA5 data. We find that forest rainfall self-reliance increases during the dry season in both the Amazon and Congo rain forests.

Atmospheric Rivers (ARs) are long, narrow corridors of concentrated moisture transport from the ocean to the coast. They play a crucial role in global precipitation and extreme weather. This study performs a complex network-based analysis to explore the spatial structure and connectivity of ARs. In particular, the clustering coefficient is used as a network measure to identify the spatial linkages in moisture transport patterns in the Integrated Vapor Transport (IVT) along the West Coast of North America over a 25-year period (1995–2019). The clustering coefficient quantifies the network’s tendency to cluster. The analysis is performed for three different categories of ARs based on the IVT magnitude—Type-1, Type-2, and Type-3. Each grid is treated as a node, and network links are established using a correlation-based threshold to assess relationships in IVT variability. The clustering coefficient values are found to range between 0.4 and 0.8. These reasonably high clustering coefficient values indicate good spatial connectivity, which could be pivotal in the rapid distribution of moisture. From Type-1 to Type-3 ARs, the progression in clustering coefficients exemplifies an increase in structural organization, highlighting the potential impacts on weather. This suggests higher IVT thresholds correspond to more concentrated and impactful AR formations. The present results offer potential lead indicators for forecasting intense rainfall, thereby highlighting the usefulness of the network-based approach for examining connections in ARs.

Stable water isotopes record the history of the phase change of water from vapor source region to sink region. Data assimilation using the stable water isotopes has the potential to more appropriately forecast moisture transportation. Thanks to recent satellite techniques, we can observe the vapor isotope ratio at a quasi-global scale daily. Here, we developed an isotope data assimilation system using the local ensemble transform Kalman filter and the isotope-incorporated non-hydrostatic icosahedral atmospheric model (NICAM-WISO-LETKF). We conducted an observation system simulation experiment (OSSE) at a low resolution (about 224 km) to test the assimilation system. The OSSE was conducted with 64 ensemble members using a synthetic observation dataset, including conventional observations. We first conducted a one-point assimilation experiment by inputting vapor isotope information at Tokyo, Japan. The experiment demonstrated that the vapor isotope information modified not only the water vapor field, but also the wind field around Tokyo. Then, we conducted OSSE for two months and calculated the global averaged root mean square error (RMSE) for vertical levels. A reduction rate of RMSE of meteorological variables, assimilating conventional observation, in addition to with and without vapor isotope information, was evaluated. We found that assimilating vapor isotope information improves the analysis in the troposphere globally. The reduction rate was large in the mid-troposphere, especially in the middle and high latitudes. Isotope data assimilation can provide a more precise forecast of moisture transportation via the constraint of a unique characteristic of vapor isotope information.

Moisture is a prerequisite for precipitation, and understanding the sources of moisture can provide insights into precipitation changes in a warming climate. The offline Lagrangian approach is a widely used method to track moisture sources for both climatological studies and individual events, especially for extreme events. This approach conceptualises the dynamic and thermodynamic processes involved in the movement of independent air parcels in the atmosphere, necessitating certain assumptions, such as mixing of evapotranspiration from surface, precipitation formation heights, vertical movement of air parcels, and convection.Although several Lagrangian models have been used in past studies, these models often have fixed assumptions and lack sensitivity tests to evaluate the impacts of these assumptions on the final results and their applicability across different regions and types of precipitation events. In this study, we utilise the Lagrangian model BTrIMS (Back-trajectory Identification of Moisture Sources) to conduct sensitivity tests on various assumptions, including the vertical movement of air parcels, the method of identifying moisture along trajectories, the release height of air parcels, the well-mixed assumption within the entire air column or the planetary boundary layer, the time step for backtracking, and the subgrid interpolation method for atmospheric fields.Our findings indicate that the vertical movement of air parcels, the release height of air parcels for backtracking, and the identification method significantly influence the results of moisture source tracking. In contrast, the time step and the interpolation method within the grid are less critical. This study highlights fundamental assumptions that require testing and suggests optimum choices to improve the accuracy of Lagrangian methods. This enhances the Lagrangian approach, broadening its applicability to more regions and diverse precipitation events.

This study investigated the motivation of the South Asian jet wave train (SAJW) and its underlying mechanism, by distinguishing different evolutions of disturbance sources over the North Atlantic. The disturbance sources propagate along the SAJW toward the western North Pacific in the propagating cases, while they are constrained over the North Atlantic and Europe in the interrupted cases. The possible motivating mechanisms are disclosed in the perspectives of energy dispersion and background potential vorticity gradient (PVG) affecting energy attraction. In the propagating cases, wave energy propagates eastward over the North Atlantic and converges into the Mediterranean Sea and Middle East, leading to the amplification and development of the wave train. The enhancement of PVG over south Europe–Mediterranean Sea is a key factor attracting such eastward propagation, which is associated with the positive PV anomaly over north Europe driven by the diabatic heating of precipitation. In the interrupted cases, energy dispersion is more divergent over the North Atlantic because of the weaker PVG, resulting in the failure of the SAJW excitement. The impacts of the SAJW on the temperature and precipitation along its path are evident, with the modulated snow depth agreeing more with temperature rather than precipitation except over the Tibetan Plateau. Hence, understanding the motivating mechanism of the SAJW is essential for improving weather forecasts over subtropical Eurasia by using the disturbances over the North Atlantic.

The asymmetry between the two phases of Atlantic Multi-decadal Variability (AMV) is examined using preindustrial control experiments from 46 Coupled Model Intercomparison Project 6 (CMIP6) models. The relative strength of tropical Atlantic sea surface temperature (SST) anomalies compared to subpolar Atlantic SST anomalies varies significantly across models. Some models exhibit a stronger tropical response during the positive AMV phase, while others show a more pronounced response during the negative phase.In models favoring the positive AMV phase, the Wind-Upwelling-SST feedback, driven by thermocline adjustments to trade wind anomalies, plays a key role in amplifying SST asymmetry in the Tropical North Atlantic. This effect is further reinforced by the Wind-Mixing-SST feedback. While previous studies suggested a model preference for negative AMV with minimal dynamic ocean involvement (Beak, 2021), the newly identified diversity in AMV asymmetry simulations underscores the need for a more comprehensive investigation into the fully coupled mechanisms governing AMV variability and the underlying model biases.

Under the background of global warming, the impact of Arctic sea-ice loss on mid-latitude weather and climate in the Northern Hemisphere has attracted widespread attention. However, the underlying physical mechanisms remain debated. Combining observations and model simulations, we demonstrate that early-winter sea-ice loss in the Barents-Kara Seas (BKS) enhances atmospheric baroclinicity and frontogenesis, leading to increased daily cold extremes over East Asia. This study investigates the influence of BKS sea-ice loss on winter surface air temperatures in China, focusing on both tropospheric and stratosphere-troposphere coupling processes. Furthermore, the respective roles of tropospheric processes and stratosphere-troposphere coupling processes are investigated. For the tropospheric processes, an eastward propagating wave train stimulated by sea-ice loss induces negative geopotential height anomalies over the western Pacific, favorable for the transport of cold airmass into China. In terms of the stratosphere-troposphere coupling processes, sea-ice loss leads to the extension of stratospheric polar vortex edge toward North China by modulating upward propagating planetary waves. These results could improve our understanding of the potential linkage between Arctic sea-ice loss and winter weather extremes in East Asia.

In recent decades, the Arctic has witnessed dramatic changes, including rapid warming, pronounced Atlantification, and a swift retreat of sea ice. However, the intricate coupling relationships among them remain poorly understood and lack systematic conclusions. This study employs a multivariate EOF method to objectively unravel the coupled modes of sea surface temperature (SST), sea ice concentration (SIC), and surface air temperature (SAT) in the winter Arctic Atlantic sector. The second mode, explaining 17.05% of the variance, emerges as a coherent pattern characterized by anomalously high SST, diminished SIC, and elevated SAT, with a dominant decadal signal. Take the positive phase as an example, its spatial structure reveals a quasi-barotropic anticyclonic anomaly over the Barents-Kara Seas, resembling Ural blocking. Southerly winds along the blocking’s flank channel warmth into the high Arctic, confined to the middle and lower troposphere. This warmth reinforces the blocking above, creating a feedback loop that amplifies poleward heat and moisture transport. Under this atmospheric circulation, regions with high SST exhibit reduced upward turbulent heat flux, indicating that atmosphere drives the underlying surface. Meanwhile, localized heat transport in the lower atmosphere over the Barents Sea partly explains the co-variability of SAT and SIC, while increased Atlantic heat influx enhances northward and eastward oceanic heat transport in the Barents Sea’s upper 300m, further amplifying sea ice melt at the margins. These findings unveil a tightly coupled system where atmosphere, ocean, and sea ice interact in a coherent, decadal rhythm.

This study reveals that the dominant mode of the Tibetan Plateau atmospheric heat source in boreal summer, characterised as the southeastern heating mode, exerts an inverse influence on the tropical easterly jet (TEJ) entrance region by strengthening the northern branch and weakening the southern branch. On the one hand, the heating effect in the southeastern Tibetan Plateau induces positive temperature anomalies in the middle and upper troposphere, which directly accelerate the TEJ over South Asia via thermal wind principle. On the other hand, the heating effect causes local air ascent over the plateau, driving upper-level divergent flow towards the western North Pacific (WNP). This results in the subsidence and downward propagation of negative vorticity over the WNP, leading to the development of an anomalous anticyclone. The anticyclone suppresses local precipitation, and the associated diabatic cooling generates a heat sink over the WNP. Through the Gill-Matsuno response, this heat sink weakens the TEJ intensity over the Maritime Continent (MC), while strengthening the TEJ over the WNP. Consequently, the southeastern heating mode induces an inverse influence on the TEJ entrance region, highlighting the critical role of the WNP in modulating the large-scale atmospheric circulation.

Freezing rain is supercooled liquid precipitation, and its key feature is that there is a significant temperature inversion layer from 900hPa (low layer) to 750hPa (middle layer). After melting through the warm layer above 0℃ in the middle layer, snow becomes supercooled droplets in the cold lower layer and condenses on the ground to form freezing rain. The freezing rain in China is concentrated in the central and western parts of Jiangnan region in January-February, with significant diurnal variation and interdecadal variation in frequency. The duration of large-scale freezing rain events is as short as 1-2 days and as long as about a week. Freezing rain events are closely related to frontal systems, which provide warm advection at the middle level and cold advection at the lower level to maintain the inversion structure. In particular, the quasi-stationary front usually leads to prolonged freezing rain. In the longer duration freezing rain event, the cold air in front of the Ural Mountain high pressure ridge is more inclined to move south from the east side of the Qinghai-Tibet Plateau, and the upper altitude cyclone is abnormal from Mongolia to the central and western China, and the cold advection is continuous. At the same time, the Northwest Pacific anticyclone anomaly and the Indo-Burma trough is longer lasting and stronger, which further provide a continuous supply of southwest warm and wet air in the middle layer. This atmospheric circulation may be related to the Arctic climate system and tropical SST to varying degrees. For example, the AO and PDO indices are statistically related to the frequency of freezing rain on an interdecadal scale.

Extreme winds diasters occurred in Nanchang, Jiangxi Province in early morning on March 31, 2024, which responsible for falling of 3 residents from high building. Based on dual-polarization Doppler weather radar observations and surface observations, this study revealed that a bookend vortex was embedded in the northern end of the bow echo within a quasi-linear convective system, with its horizontal scale and vertical vorticity comparable to that of a strong mesocyclone. Tornadic vortex signatures (TVS) with tornadic debris signatures (TDS) were identified in the center of the bookend vortex which passed over the the building of the incident. The rotational velocity of the TVS exceeded 40 m·s^-1. Damage surveys conducted along the path identified by TVS and TDS revealed tornadic damages, which included strong cyclonic rotational winds, localized convergent winds, debarked trees, and the complete collapse of stone bridges. This indcated that an EF2 tornado is most likely responsible for the fatal fall. The simultaneous emergence and intensification of a bookend vortex along with two to three leading-edge misocyclones embedded within it is a new phenomenon. An observational study on the formation mechanisms of tornado-related vortices indicates that these vortices result from the tilting of horizontal vorticity. The horizontal vorticity is associated with the strong low-level vertical shear of the elevated rear-inflow jet and the baroclinic vorticity related to the gust front.

Tornado-induced railway accidents have been recorded in Japan, including incidents on the Touzai Line (1978), the JR Uetsu Line (2005), and the JR Nippo Line (2006). The latter two were formally investigated by the Aircraft and Railway Accident Investigation Commission, which assessed the running safety of railway vehicles in tornadic winds using methodologies originally developed for synoptic wind conditions. Our previous study refined this evaluation approach by incorporating realistic tornadic wind distributions and multibody simulations, revealing that the quasi-static analysis employed in these investigations systematically underestimates the dynamic response of railway vehicles. To address this limitation, we introduced a dynamic amplification factor (DAF) to quantify the underestimation. In this study, we extend the investigation by comparing evaluation methodologies adopted in different countries, including Japan, China, and Europe, to determine the most appropriate approach for assessing railway vehicle stability under tornadic wind conditions. By correlating our findings with documented accident sites, we identify a suitable evaluation framework. Furthermore, given that these tornado-induced accidents predominantly occurred on narrow-gauge railways, we systematically compare vehicle responses on narrow and standard gauges to elucidate the influence of track width on running safety in extreme wind events.

The risk of tornadoes depends on their parent convective systems. Therefore, we have to establish tornado climatology with respect to parent convective systems. The present study aims to classify the parent convective systems and to examine their characteristics, e.g. size, occurrence location, the intensity of vortices in the parent systems. The tornado events are picked up from the Japan Meteorological Agency (JMA) tornado database and are those located within the observation range of JMA radars and XRAIN radars. Analysis periods are 12 years from 2013 to 2024. The figures of reflectivity and Doppler velocity were drawn around the tornado events for more than 2 hours. We classified the parent convective systems based on their shape and size of strong echo region of more than 40 dBZ in reflectivity and temporal change in shape. Doppler velocity images were used to extract vortices in the parent convection system and to measure the diameter, tangential velocity, vorticity of the vortices. The moving velocity of the parent convective systems were obtained from the movement of the vortices. From more than 170 systems, we classified into 6 kinds; “Isolated cumulonimbus”, “Supercell”, “Cloud cluster”, “Quasi-linear rain band”, “Local front” and “Inner rainband of typhoon”. The most convective system is cloud cluster, and second measure system is isolated cumulonimbus. The cloud clusters tend to occur in the Pacific Ocean side and inland in morning in summer season. On the other hands, the isolated cumulonimbi occur in coastal area and offshore in rate afternoon in autumn. We also found the characteristics of tornadoes with respect to the moving direction, the changes of velocity difference and vortex diameter after landfall, and the relationship between maximum wind speed and gust damage.

Tornadoes are a significant weather hazard in Northeastern China, but their characteristics and environmental conditions differ from those in other regions. This study presents a comprehensive analysis of 132 tornadic events spanning a 20-year period from 2004 to 2023, utilizing radar and ERA5 reanalysis data to investigate the convective modes and environmental parameters associated with tornadoes in this region. The research particularly focuses on the relationship between tornadoes and Northeast China cold vortices (NCCVs). The analysis reveals that discrete storms are the most common type (70%), with the CC (clustered cell) mode being the most frequent. Significant tornadoes (EF2 and above) are primarily associated with discrete storms, particularly the IC (isolated cell) and BL (broken line) modes. Environmental parameters such as storm relative helicity (SRH) and significant tornado parameter (STP) are identified as the most effective indicators for distinguishing tornadic from non-tornadic events and for assessing tornado intensity. However, no single parameter clearly differentiates between all tornado modes. In comparative terms, the discrete mode is characterized by its low shear and high convective available potential energy (CAPE), while the linear modes exhibit the opposite characteristics. A significant association (83%) is found between tornadoes and NCCVs, with a preferential distribution in the southeast quadrant. The study categorizes tornadoes associated with NCCVs into three synoptic patterns: T1, with strong and deep cold vortices and a high proportion of BL and QL (quasi-linear) modes; T2, associated with weaker NCCVs and a predominance of CC mode tornadoes; and T3, characterized by strong NCCVs with a transverse trough, leading to fewer tornadoes due to spatial separation of CAPE and SRH. This study enhances our understanding of tornadoes in Northeastern China, informing future research on formation mechanisms, prediction methods, and disaster prevention strategies.

Japan experiences an average of about 20 tornadoes annually. The number of tornadoes per 10,000 km² (hereafter referred to as tornado density) is 0.54, which is approximately 40 percent of that in the United States. Furthermore, tornado-related fatalities in Japan is on average only 0.5 per year, significantly fewer compared to the 72 fatalities per year in the United States. Due to these relatively low numbers, modern tornado research in Japan did not begin until the early 1990s, when a Doppler radar for research purposes first detected supercells in the country.However, Japan had a relatively advanced computational environment, which facilitated several notable numerical studies. The first numerical simulation of a tornado in a realistic environment was conducted in 2009, focusing on a tornado associated with Typhoon Shanshan in 2004, and investigating the mechanism of tornadogenesis. In 2018, an ensemble simulation study examined a supercell tornado that occurred in 2012. This study highlighted the importance of a strong low-level mesocyclone, which generates intense near-surface updrafts capable of tilting and stretching horizontal vorticity of various origins. More recently, numerical studies have also been conducted on tornadoes spawned by a newly observed meso-beta-scale convective vortex and a quasi-linear convective system.Regarding observational research, three phased-array Doppler radars have been deployed in the Kanto Plain, where tornado density is relatively high. These radars have occasionally captured tornadogenesis events, providing detailed observations of the development and behavior of incipient vortices.Additionally, the structure and environment of extratropical and tropical cyclones that generate tornadoes were studied. The results emphasized the importance of the hierarchical structure of atmospheric disturbances, ranging from large-scale cyclones to supercells and tornadoes. It has also been noted that E-CAPE, which accounts for the effects of entrainment on CAPE, is a valuable environmental parameter for assessing tornado potential.

The Ozone Monitoring Suite-Nadir (OMS-N), a state-of-the-art hyperspectralultraviolet-visible (UV-VIS) sensor onboard China’s FengYun-3F (FY-3F) satel-lite, was launched in August 2023. Designed for a morning orbit, OMS-Nrepresents a significant advancement in global atmospheric composition moni-toring, offering an unprecedented spatial resolution of 7 km × 7 km. The totalozone column (TOC) product derived from OMS-N is critical for understand-ing UV radiation exposure, climate change, and environmental hazards. Thisstudy presents the first TOC retrievals from OMS-N, utilizing an adapted Differential Optical Absorption Spectroscopy (DOAS) algorithm. The retrievalalgorithm overcomes traditional DOAS limitations by incorporating key inno-vations, including optimized radiative transfer calculations and refined a prioriinformation on surface properties and ozone profiles, which are derived directlyfrom OMS-N spectra rather than relying on external datasets or climatologies.Validation against ground-based measurements from Brewer, Dobson, and SAOZinstruments at 33 sites demonstrated strong agreement, with correlation coeffi-cients mostly greater than 0.9. Comparisons with other well-established satelliteinstruments, including TROPOMI and GOME-2B, showed that OMS-N can con-sistently capture global seasonal ozone patterns, with biases typically within 2 %across hemispheres and seasons. These results establish OMS-N as a reliable toolfor high-resolution dynamic ozone monitoring, significantly enhancing our abilityto address climate and environmental challenges.

Eddy covariance (EC) measurements of air-sea CO₂ exchange (Net Ecosystem Exchange) over coral reefs in the Gulf of Eilat (GoE), Israel and at Heron Reef on the southern Great Barrier Reef, Australia show these ecosystems are net sinks of atmospheric CO₂. This result contrasts with marine productivity models and bulk formula calculations that for more than three decades have often concluded that coral reefs are net sources of CO₂ to the atmosphere with only rare cases finding otherwise. Here we present results from our direct measurements of air – sea CO₂ exchange over coral reefs using EC systems. These show coral reefs in the GoE sequester 3 to 10 times more CO₂ than other marine and terrestrial ecosystems including tropical rainforests. Observations from Heron Reef highlight the heterogeneous nature of CO₂ air-sea exchanges that may occur over coral reefs in response to different geomorphic, hydrodynamic, and ecological zones found on coral reefs. This emphasizes the need for further direct measurements of air-sea CO₂ exchanges over coral reefs. These should be conducted in different environmental settings and climates using EC so that the role of coral reefs in the global carbon cycle can be accurately quantified.

Per- and polyfluoroalkyl substances (PFASs) have gained increasing global attention due to their environmental persistence and potential health impacts. In almost two decades, studies of airborne PFASs increased from only one in 2005 to 65 until 2024, reflecting increased awareness and importance of monitoring airborne PFASs globally. We overview studies of airborne PFASs, focusing on geographical distribution, PFASs classes, sampling characteristics and temporal trends.  Of total 66 published studies, >90% focuses on PFASs in particulate matter (PM) with much fewer work investigating gaseous PFASs. More than 60% of studies sampling airborne PFASs were in urban environments, followed by rural and semi-urban locations. Fewer than 5% of the reviewed research works examined airborne PFASs in remote locations such as forest or sea. Measurements of airborne PFASs indoors accounted for about 20% of the reviewed studies, indicating the needs of more data related to personal exposure indoors. The geographical distribution among a total of 20 locations highlights China with a significant portion of studies (21 published work), reporting airborne PFASs levels as high as 6.8x10⁵ pg/m³. This is followed by eight studies each in the United States (0−7x10⁴ pg/m³) and Germany (0−3x10⁵ pg/m³). Around 40 studies in other 16 countries (including as Norway, Canada, Australia, European nations, Asian nations etc.) reported concentrations of airborne PFASs of 0−1.4x10⁵ pg/m³. Amongst different PFASs classes, airborne PFCAs (perfluorocarboxylic acids) are most studied (70% of total studies) with concentrations ranging from 0−5x10⁵ pg/m³, followed by PFASs precursors (67%, 0−3x10⁵ pg/m³, e.g., fluorotelomer alcohols (FTOH)), and PFSA (perfluorosulfonic acids, 56%, 0−6.8x10⁵ pg/m³). Taken together, our review demonstrates that more information of potential sources of airborne PFASs is needed to enhance our understandings of their emissions, fate, global distribution and impacts.

Light-absorbing aerosols, which are primarily composed of black carbon (BC), brown carbon (BrC), and mineral dust, have attracted significant research attention due to their substantial impacts on human health and climate effects. Among these, BC and BrC are the primary light-absorbing carbonaceous aerosols, with BC being recognized as the predominant warming particulate, underscoring the imperative for effectively managing its emissions to mitigate regional climate warming. However, atmospheric aging processes, including coagulation, condensation, and heterogeneous reactions, modify the morphology and coating of BC, enhancing its light absorption through the lensing effect and complicating the assessment of its optical properties. Given this, it is crucial to account for such aging effects when assessing source-specific optical properties.
This study employed high-time-resolution instruments to continuously monitor multi-wavelength aerosol light absorption, chemical species, and particle size distributions during an intensive observation period (IOP) in winter in urban Northern Taiwan. A dataset combining these monitoring data was analyzed using Positive Matrix Factorization (PMF) to determine optical source apportionment. To accurately quantify BC’s optical characteristics, the analysis focused on BC from the three non‐mineral-dust sources—thus isolating the primary light-absorbing component at 880 nm. The results show that BC accounts for over 86% of total light absorption, while BrC and mineral dust contribute less than 6% and 9%, respectively. Notably, BC associated with secondary aerosol formation exhibited the highest mass absorption cross section (MAC) of 12.08 m²/g; however, its direct radiative forcing is relatively low (0.01 W/m²) because its overall mass fraction is small. This contrast suggests that atmospheric aging, via mechanisms such as the lensing effect, may substantially enhance BC’s light absorption, which a critical for evaluating its climate impact.

Traffic-related pollutants pose a significant environmental and public health risk in urban areas. Ultrafine particles (UFPs) smaller than 100 nm can penetrate deep into alveolar tissue, posing serious health hazards. In October 2024, the EU introduced stricter UFP regulations to mitigate these risks. Additionally, volatile organic compounds (VOCs) and nitrogen oxides (NOₓ), which serve as precursors to ozone formation, further degrade air quality. Although motorcycles are a major mode of transportation in many Asian countries, research on their emissions remains limited compared to light-duty vehicles (LDVs) and heavy-duty vehicles (HDVs). Most studies rely on dynamometer tests under controlled conditions, which fail to capture real-world emissions influenced by driving behavior, traffic, and environmental factors. This study addresses this gap by quantifying VOCs and UFPs emission factors from different vehicle types through tunnel experiments.
The measurements were conducted at the Ziqiang Tunnel in Taipei City, Taiwan. This tunnel measurement campaign utilized high-temporal-resolution instruments for both stationary and mobile measurements to capture real-world emissions. Multiple linear regression was used to derive emission factors for different vehicle types. The results indicate that the UFPs emission factor of HDVs is 5.9 times higher than that of LDVs and 7.2 times higher than that of motorcycles, while LDVs exceed motorcycles by only 1.2 times. For aromatic VOCs, motorcycles contribute higher emission factors of benzene and trimethyl benzene than LDVs due to their lower combustion efficiency. This study provides real-world measurements and quantifies VOCs and UFPs emission factors from motor vehicles, offering critical insights for regulatory frameworks. Future emission regulations should account for motorcycle emissions to avoid underestimating their impact on air pollution.

The physical mechanisms by which the classical El Niño–Southern Oscillation (ENSO) affects Australian climate are well understood. However, the impacts on Australian rainfall vary significantly depending on different ENSO strengths, spatial patterns, and location of maximum sea surface temperature (SST) anomalies. Over recent decades, substantial scientific advancements have aimed to explain these differences, particularly between Eastern Pacific (EP) and Central Pacific (CP) El Niño, as well as multi-year La Niña events. This study reconciles previous research on ENSO diversity to better understand the mechanisms that influence Australian rainfall, particularly in austral spring when ENSO’s impact is strongest in the region. We find that the teleconnection mechanism in the tropics is direct via changes in the Walker Circulation and can be explained by the Gill-Matsuno response to diabatic heating at different equatorial locations. In the extratropics, ENSO affects rainfall through the Pacific–South America (PSA) pattern and an indirect pathway involving changes in the Indian Ocean SST anomalies. Our results indicate that La Niña events typically enhance rainfall via a PSA-driven high-pressure system over the Tasman Sea, combined with low-pressure systems over eastern Australia. This synoptic pattern increases inland moisture advection, leading to heavy rainfall during La Niña years. We also discuss the influence of multi-year La Niña on heavy rainfall in northern and eastern Australia, amplified by soil moisture-rainfall feedback. Additionally, we show that while EP El Niño has a weaker and less consistent influence on Australian rainfall than CP El Niño, its impact is indirect via Rossby wave trains originating from the Indian Ocean. The wave train produces a high-pressure system to the south of Australia that significantly dries southwestern Western Australia—an area traditionally considered less affected by ENSO. Our findings highlight the importance of considering ENSO diversity to better understand variations in Australian spring rainfall. 

The 28°C Isotherms warm pool changes and the associated air-sea heat budgetsNoraini Mohyeddin^1*, Wee Cheah^1, Mohd. Fadzil Firdzaus Mohd Nor¹ , Huang-Hsiung Hsu³,
Wan-Ling Tseng³ and Azizan Abu Samah^1,21 Institute of Ocean and Earth Sciences (IOES), Universiti Malaya, Malaysia² National Antarctic Research Centre (NARC), Universiti Malaya, Malaysia³ Research Center for Environmental Changes, Academia Sinica, Taiwan, R.O.C ^*Corresponding email: noraini0210@gmail.com28°C isotherm warm pool (OWP₂₈), usually known as the Indo-Pacific warm pool (IPWP), marked an area with lasting warm sea surface temperature on the Earth and has shown remarkable expansion. In this study, the characteristics and evolutions of surface and subsurface warm pools are examined using ECMWF Reanalysis products – ORAS5 (global ocean) and ERA5 (global climate and weather) with 0.25° x 0.25° horizontal and monthly temporal resolutions. From 1960 to 2020, decadal variations of SST revealed an increasing trend of warming spatially; however, decadal changes of OWP₂₈'s area (AOWP₂₈) revealed its unique increasing trend with 2.38 x 10¹¹ per decade of change. The ratio of AOWP₂₈ in the Pacific Ocean (AOWP₂₈-PO) to AOWP₂₈ in the Indian Ocean (AOWP₂₈-IO) has been steadily decreasing over time, indicating faster lateral changes in AOWP₂₈-IO as compared to AOWP₂₈-PO. Early analysis of key variables in the ocean heat budget revealed that net surface heat fluxes have decreased in the area; thus, ocean heat storage and ocean heat transport played key role for the changes of OWP₂₈, especially from 1980 to 2020. Quantification of the 28°C isotherms warm pool changes and the associated air-sea heat budgets is of great importance for regional sea level change, transient climate sensitivity and related feedback. Keywords: 28°C isotherm warm pool, Indo-Pacific warm pool, air-sea heat budgets

The Madden-Julian Oscillation (MJO) is a significant intraseasonal phenomenon that greatly impacts the global hydrological cycle and induces notable climate and weather anomalies. Previous research has frequently documented an increase in the amplitude and eastward propagation speed of MJO activity under global warming scenarios and over long-term records. As anthropogenic warming intensifies and internal climate variability persists, we identify a recent regime shift around 2011 in the MJO and the background mean state over the Maritime Continent that remains inadequately explained. Here, we focus on identifying the connection between changes in MJO variance and shifts in the associated mean state. Our findings reveal a weakening of MJO activity over the Indian Ocean and a strengthening east of the Maritime Continent after 2011. These changes in MJO activity are linked to the local enhancement of the MJO background flow that was favorable for the MJO development. The background flow change was found associated with an enhanced winter monsoon mean flow, and the anomalous extratropical circulations and ocean warming in the eastern Pacific that appeared in the early 2010s. In summary, our study emphasizes the importance of the mean state in driving the decadal changes in MJO activity, with a notable influence from the extratropic. These findings highlight the need for further exploration and discussion of the complex regional characteristics associated with decadal changes in the MJO from a global perspective.

The Madden-Julian Oscillation (MJO) is the most dominant mode of tropical variability on subseasonal-2-seasonal (S2S) timescales, which affects global high-impact weather and rainfall variability through its large-scale convection as a forcing for global circulation and teleconnection. This study examines changes in the MJO precipitation from 1979 to 2023, using a novel method tracking the MJO large-scale precipitation explicitly, which is distinct from previous studies. We find a significant increasing trend in MJO precipitation event frequency and a poleward expansion into both hemispheres, particularly over the Maritime Continent (MC) and western Pacific. The intensity and duration of MJO precipitation events have increased by 4 mm/day and 7 additional days per event, respectively, in some regions. The MJO precipitation over the Indo-Pacific warm pool has a growing influence on global precipitation patterns, explaining 20–40% of interannual rainfall variability in regions such as Southeast Asia, tropical Africa, and the western United States. These findings offer insights for understanding the increasing trend of MJO precipitation and its influence on the MC and global rainfall variability in a changing climate.

Western Java experiences the most frequent hydrometeorological hazards in Indonesia. Such hazards are closely associated with extreme rainfalls which is known to be driven by multi-scales climate events such as Madden-Julian Oscillation (MJO) and El Nino-Southern Oscillation (ENSO). However, the impact of such events to western Java and how it interact with the local factors (e.g., topography) remain unclear, mainly due to the limitation of in-situ weather observation used in previous studies. We filled-in the gap by analyzing daily rainfall data from 372 rain gauges across the region. We revealed the role of topography in modulating spatio-temporal variability of rainfall associated with the MJO, Monsoon, Inter-tropical Convergence Zone (ITCZ) and ENSO. We have also highlighted the impact of large to global-scale climate change to the extreme rainfall over the region by analyzing its short and long-term trends. Our analyses had comprehend the general understanding of the variability and trend of rainfall over western Java.

Dry spells, defined as a period of consecutive days with daily rainfall less than 1 mm, are the building blocks of meteorological droughts. Prolonged periods of little to no rainfall can have significant impacts on water resources, ecosystems and agriculture. Over Southeast Asia, extended dry spells during the dry season can also increase the risk of transboundary haze from open biomass burning. However, there is no universal definition on the length of the period of dry days required to constitute a dry spell or a drought. Rather, it depends on the regions and their climate zones. For instance, in drier regions dry spells are commonly defined as at least 5 consecutive days of no rainfall, whereas in wetter regions the threshold can be as short as one day. The definition can also depend on the environmental and societal impacts. These nuances in the region-dependant definition of dry spells also apply to drought definition. In this work, we explore the present-day characteristics of dry spells longer than 5 days for the Southeast Asian wet and dry seasons using gridded rainfall datasets. Metrics analysed include the frequency and duration of dry spells, as well as their areal coverage. The observed characteristics are compared to simulated dry spells over the historical 1981-2024 period, which we examine in downscaled high-resolution (8-km) regional climate simulations vis-à-vis their parent ERA5 reanalysis or CMIP6 models. The results provide a baseline to understand future changes at different warming levels as explored in a companion study.

To better understand the concentrations and impacts of gas- and particle-phase contaminants in a rural site of Tibet, a thermal desorption comprehensive two-dimensional gas chromatography-mass spectrometer (TD-GC×GC-MS) was utilized to qualify and quantify air pollutants. Almost all chemicals qualified were quantified or semi-quantified. Alkenes were predominant components in mass concentration, ozone formation, and secondary organic aerosol (SOA) formation. The key alkene compounds were α-pinene, cedrene, 3-carene. Intermediate volatility organic compounds (IVOCs) accounted for 26.4%, 92.3%, 15.5%, and 25.2% of the total gas-phase concentration, particle-phase concentration, ozone formation potential (OFP), and SOA estimation, respectively. Typical oxidation products of terpenes, i.e., pinediol and cedrol were detected, yet 2-caren-10-al and 5-isopropenyl-2-methyl-7-oxabicyclo[4.1.0]heptan-2-ol formation pathways were not reported in previous studies. By combining the pixel-based estimation of octanol–air partition coefficient (K_o–a), air–water partition coefficient (K_a–w), octanol–water partition coefficient (K_o–w), and OH rate constant (kOH), the long-range transport potential (LRTP) of chemicals were assessed. D6, D7, pinediol, bornyl acetate, cedrol, diisobutyl phthalic acid ester, nonanoic acid, and C22-C32 n-alkanes were found to be chemicals of high LRTP concern. The results of concentration weighted trajectory (CWT) analysis demonstrated that anthropogenic and biogenic sources from south and southeast Asia played a role in the local air pollution of Lulang, Tibet. This work gives more insight into the occurrence, formation pathways, and sources of air pollutants in a typical rural site.

Ozonolysis reactions predominate the removal of certain volatile organic compounds (VOCs) containing a C=C bond in the atmospheric troposphere, given the average atmospheric concentration of O₃. These reactions are facilitated by the formation of primary ozonides (POZs) through the electrophilic 1, 3-cycloaddition of O₃ to the C=C bond. Subsequently, these POZs rapidly decompose into carbonyl compounds and carbonyl oxides, namely Criegee intermediates (CIs). Part of the initially thermalized CIs may promptly dissociate to form OH radicals, serving as a significant non-photolytic source of OH radicals in the atmosphere. The remaining CIs are collisionally stabilized, termed as stabilized CIs (SCIs), which can engage in bimolecular reactions with various trace small molecule gases such as H₂O, HCOOH, SO₂ and CH₃OH in the atmosphere, thereby influencing the formation of secondary organic aerosols (SOA). The reaction rate of CIs with HCOOH is about 10^-10 cm³molcule^-1s^-1, and the reaction rate of SO₂ with CIs is much faster than with OH radical. We have constructed matrix isolation technology combined with vacuum Fourier transform infrared spectrometer (MIFT-IR) to capture the different structures of CIs. Furthermore, additional smog chamber experiments with alkene ozonolysis were conducted to investigate the impact on SOA formation mechanisms. A series novel CIs spectra have been obtained, and formation of oligomers containing CIs have been confirmed in SOA.

Chlorine radicals have emerged as significant yet underexplored drivers of atmospheric oxidation, with important implications for secondary organic aerosol (SOA) formation, air quality, and human health. Unlike traditional oxidants, Cl-initiated oxidation follows distinct reaction pathways that contribute to the chemical complexity and physical properties of atmospheric organic matter. These reactions can proceed via both addition to unsaturated bonds and hydrogen abstraction, generating a diverse mixture of chlorinated (Cl- RO₂) and non-chlorinated (plain-RO₂) peroxy radicals. However, the interactions between these radical species remain poorly understood, despite their critical role in shaping SOA composition, volatility, and potential toxicity. In this work, we investigate the influence of Cl chemistry on volatile organic compound (VOC) oxidation, with a focus on the formation and transformation of oxidation products. Preliminary findings suggest that chlorine-driven processes may alter product distribution, oxygenation levels, and molecular interactions, potentially impacting SOA mass and reactivity. These observations underscore the need for a deeper understanding of chlorine’s role in atmospheric oxidation, particularly in regions with elevated chlorine emissions. Further research is essential to elucidate the underlying mechanisms and assess their broader implications for air pollution, oxidative stress, and environmental sustainability.

Isoprene emissions from tropical plants under moderate conditions are more temperature-sensitive than temperate plants and current model predictions. However, it remains unclear how extreme heatwaves might alter this behavior. Here, we present controlled measurements of isoprene temperature responses for a subtropical eucalyptus species, demonstrating a surprising shift in emission behavior under heatwave conditions. During non-heatwave periods, isoprene emissions followed established temperature sensitivity patterns with a well-defined optimum temperature (T_opt); In contrast, heatwaves significantly restricted plant physiological processes, resulting in an unexpected decrease in T_opt. Current models, which account for acclimation effect using long-term temperature averages, failed to reproduce this shift, instead predicted higher T_opt values during heatwaves. Remarkably, simulations using the default model curve, which assumes no acclimation, were able to replicate observed isoprene emissions under both non-heatwave and heatwave conditions. Our findings highlight the potential for extreme heat to suppress biogenic isoprene emissions in tropical regions and the need to reconsider isoprene-temperature relationships under future climate.

Organic aerosols (OA) account for a large portion of atmospheric aerosols, which play a significant role in climate change and human health. Accurate characterization of OA concentrations and composition is essential as it directly affects its environmental, climate and health effects. OA concentrations and composition are influenced by meteorological conditions, particularly weather extremes such as heatwaves. With the warming climate, frequency and intensity of heatwaves are on the rise, making it imperative to understand heatwaves impacts on urban OA concentrations and composition. Effects of heatwaves on the molecular composition and concentrations of urban organic aerosols are still unclear due to limited observations. In this study, we measured chemical composition of OA at molecular level using extractive electrospray ionization time-of-flight mass spectrometer (EESI-TOF-MS) at Shanghai in summer of 2024 when a number of heatwaves occurred Shanghai. We investigated effects of heatwaves on the molecular composition of urban organic aerosols. We found that heatwaves can significantly affect molecular composition and concentrations of OA. We also analyzed the underlying mechanisms of this influence by combining changes in gas precursors and formation process of secondary OA during heatwaves compared to non-heatwave periods. This information enables better understanding of OA and more effective air pollution control strategy during heatwaves.