Biomass burning processes emit substantial amounts of compounds into the atmosphere, such as phenolic compounds and furans. These compounds have been recognized as potential precursors for secondary organic aerosols (SOA). Here, the kinetics of phenolic compounds and furans with hydroxyl radical (OH) were characterized in the gas phase and aqueous phase, and corresponding SOA formation mechanisms were determined. Phenolic compound with high water solubility can be involved in aqueous-phase reactions. Bimolecular aqueous-phase reaction rate constants of phenolic compounds with OH range from (3.8 ± 0.7) × 10⁹ to (1.1 ± 0.3) × 10¹⁰ M⁻¹ s⁻¹. Aqueous-phase OH oxidation of phenolic compounds can form high-molecular-weight compounds, which are important SOA components. Hydroxylation of benzene ring occurs through OH addition, which could increase SOA oxidation degree. Furans are mainly oxidized by gas-phase OH to form SOA, and oxidation processes are influenced by different factors. Specifically, we observed a positive relationship between the furan SOA yields and NO_x concentration, which is attributed to the production of low-volatility hydroxyl nitrates and dihydroxyl dinitrates. Increasing RH results in high OH concentration and enhances carbonyl-rich products condensation, thereby promoting furan SOA formation. In addition, changing SO₂ level in the range of 0–80 ppb cause furan SOA yield to increase by 3.8%, while the presence of ∼50 ppb NH₃ increased furan SOA yield by 3.3%. SO₂ can lead to three C5-organosulfate formation, which may be precursors for the atmospheric C5-organosulfate. NH₃ has a substantial promotion effect on the formation of light-absorbing nitrogen-containing products, which may be important brown carbon constituents. Our work provided mechanism insights about SOA formation from biomass burning-derived compounds, which is helpful for refining the representation of biomass burning-related SOA formation in models.

Photochemical air pollution is a pressing environmental problem in the world’s urban and industrialized regions, including Hong Kong. Ozone and fine particulate are major air pollutants. In 2018, the Hong Kong Research Grants Council funded a comprehensive research project on the sources, atmospheric processes, and control strategy of photochemical pollution. This talk will give an overview of the project’s scope and key findings in the past 5 years. The major scientific achievements made thus far include the discoveries of the important roles of halogen atoms in air quality of polluted regions, the new source/production pathway of organic particulate matter from urban and biomass-burning emissions, and the complex responses of secondary air pollutants to emission controls. Based on the findings, the project has recommended additional measures to mitigate winter haze in north China and photochemical smog in south China.

Biomass burning (BB) and biogenic emissions of saccharides were investigated in 3 rural sites at Lincang, in the southwest border of China. PM_2.5 samples were simultaneously collected on three mountaintop _sites in Lincang: Datian, Dashu, and Yakoutian, which are located ∼ 300 km west of Kunming (the capital of Yunnan Province in China) and ∼ 120 km east from the Burma border. Five saccharide alcohols (glycerol, erythritol, inositol, arabitol, and mannitol) and five primary saccharides (fructose, glucose, mannose, sucrose, and trehalose), together with three anhydrosugars (levoglucosan, mannosan, and galactosan), were quantified by an improved high-performance anion-exchange chromatograph coupled with a pulsed amperometric detector. The total saccharides accounted for 1.6 ± 0.6 % in PM2.5. The anhydrosugars accounted for 48.5 % of total saccharides, among which levoglucosan was the most dominant species. The high level of levoglucosan was both attributed to the local BB activities and biomass combustion smoke transported from the neighboring regions of Southeast Asia (Myanmar) and the northern Indian subcontinent. The mono- or di-saccharides and sugar alcohols accounted for 24.9 ± 8.3 % and 26.6 ± 9.9 % of the total saccharides, respectively, and both proved to be mostly emitted by direct biogenic volatilization from plant material or surface soils rather than byproducts of polysaccharide breakdown during BB processes. Five sources of saccharides were resolved by non-negative matrix factorization (NMF) analysis, including BB, soil microbiota, plant senescence, airborne pollen, and plant detritus with contributions of 34.0 %, 16.0 %, 21.0 %, 23.7 %, and 5.3 %, respectively. The results provide information on the magnitude of levoglucosan and contributions of BB, as well as the characteristic of biogenic saccharides, at the remote sites of southwest China, which can be further applied to regional source apportionment models and global climate models.

Global ammonia is increasingly recognized as a significant contributor to fine particulate matter with a diameter smaller than 2.5 microns (PM2.5) due to the lack of mitigation policies for this species. Apart from agriculture fields, feedlots, and dairy farms, the more frequent wildfires during the dry season are contributing to rising emissions. Airborne and satellite instruments, such as Hyperspectral Thermal Emission Spectrometer (HyTES) and Cross-track Infrared Sounder (CrIS), have both detected and observed the ammonia plumes and enhancement from wildfires. Quantitative retrievals of HyTES observations at Gulch Fire in Arizona on July 3rd, 2014, suggest that the enhancement level is about 5 ppb, which is lower than both the emission from power plants (~10 ppb) and feedlots (~50 ppb) but ineligible. A further study with CrIS observations showed an unusual high level of ammonia at the Los Angeles area during the Bobcat fire in September 2020. A similar abnormal peak during the same month over Imperial Valley, a remote region southeast of Los Angeles basin. This may suggest the remote impact of wildfire emissions. Further modeling studies will help us to understand the process and contribution of wildfires in addition to local anthropogenic sources.

Volatile organic compounds (VOCs) are important species in the atmosphere which affect air quality and climate. Fire burning, including wildfires, agricultural fires, prescribed burning, and residential wood combustion, is the largest combustion-related source of VOCs to the atmosphere ^1–3. In addition to CO, CO₂ and CH₄ hundreds or even thousands of VOCs are emitted from fire burning ⁴. Such emissions are likely to increase in the future, with the spatial extent, number, and severity of wildfires having increased markedly at the global scale in recent decades ^5,6, this is predicted to continue as the climate warms ^7,8. In general, VOCs from fire burning include non-methane hydrocarbons (NMHC, C₂-C₁₂), aromatics, oxygenated hydrocarbons, nitrogen containing compounds. In the atmosphere, the emitted VOCs could be oxidized by OH radical, NO₃ radical and O₃ to form photochemical oxidants (e.g. O₃) and fine particles, which can cause severe air pollutions and damage human health. In this presentation, we review the knowledge about VOCs emissions from fires and discuss their atmospheric impacts on air quality and climate using the recent data from laboratory and fields studies. Reference:(1) Reid, J. S.; et al., Atmospheric Chem. Phys. 2005, 5 (3), 799–825. (2) Chen, J.; et al., Sci. Total Environ. 2017, 579, 1000–1034. (3) Granier, C.; et al., Clim. Change 2011, 109 (1–2), 163–190. (4) Koppmann, R.; von Czapiewski, K.; Reid, J. S. Atmospheric Chem. Phys. preprint; 2005. (5) Jolly, W. M.; Cochrane, M. A.; Freeborn, P. H.; et al., Nat. Commun. 2015, 6 (1), 7537. (6) Harvey, B. J. Proc. Natl. Acad. Sci. 2016, 113 (42), 11649–11650. (7) Krikken, F.; Lehner, F.; Haustein, K.; et al., Nat. Hazards Earth Syst. Sci. 2021, 21 (7), 2169–2179. (8) Lohmander, P. Cent. Asian J. Environ. Sci. Technol. Innov. 2020, 1 (5).

Nitro-phenolic compounds (NPs) have attracted increasing attention because of their health risks and impacts on visibility, climate and atmospheric chemistry. Both primary emission and secondary formation of NPs have been reported in recent studies, despite increasing research, the knowledge of the abundances, sources, atmospheric fates, and impacts remains incomplete, especially the gaseous species. In the present work, we conducted continuous measurement of eighteen gaseous NPs with a chemical ionization mass spectrometer at a background site in South China. Abundant NPs were observed in the continental outflows, with a total concentration of up to 122.1 pptv. A tracer method estimation showed that besides the direct emission from biomass burning, the secondary formation from the transported precursors contributed significantly to the observed NPs, with mono-NPs exhibiting photochemical daytime peaks and nighttime enrichments of di-NPs and Cl-substitute NPs. The budget analysis with a photochemical box model indicates that besides the ŸOH oxidations of aromatics, the NO₃Ÿ oxidation also contributed significantly to the daytime mono-NPs, and was more dominant for the nocturnal formation of di-NPs. Photolysis was the main daytime sink of NPs, producing substantial HONO, and thus would largely influence atmospheric oxidation capacity in downwind and background regions. This study provides quantitative insights on the characteristics, formation and impacts of gaseous NPs in the continental outflow, and highlights the role of NO₃Ÿ chemistry in the nitro-aromatics production that may facilitate regional pollution.

Urban fires in densely populated areas can burn large amounts of manmade structures and synthetic materials, which potentially have different chemical characteristics to wildfire smoke. Realworld chemical composition data of urban fire plumes are vital for evaluating health impacts; however, they are rarely available. Online monitoring of trace elements in ambient PM_2.5 at five sites throughout Hong Kong in 2021 provided unprecedented opportunities to observe the highly transient urban fire plumes. We observed sharp spikes for bromine (up to 0.65 µg m^-3), chlorine (7.0 µg m^-3), and certain metals such as lead (0.97 µg m^-3), zinc (8.3 µg m^-3), and copper (0.32 µg m^-3) during four urban fires, which showed ~10–6,000 times higher concentrations than pre-fire periods. These elements behaved decreased concentrations with increased distance from the fire location due to transport and dilution, but their concentration ratios remained relatively constant. During fire-influenced hours, these elements exhibited 3–101 times enrichment relative to iron, an element that had no discernable increase in concentration. It was distinct to one monitored hill fire that was characterized of remarkably high enrichment of tracers for biomass burning, i.e., Rb (11.6) and K (6.5). Our results represent the first observation of the co-emissions of Br and certain metals from urban fires. Reactive bromine released from burning of the wide-spread brominated synthetic chemicals was suggested to facilitate the mobilization of toxic metals into the atmosphere. This mechanistic knowledge will help us understand the emissions and impacts of fires occurring in wildland urban interface zones. The work also calls attention to previously overlooked influence of urban fires-mobilized metals on human health and of bromine on atmospheric oxidizing capacity in megacities.

Methoxyphenols represent one of the most abundant biomarker tracers for atmospheric wood smoke pollution. The reactions of atmospheric oxidants (ozone, OH, and NO₂) with methoxyphenols can contribute to the formation of secondary organic aerosols (SOA). The ionic strength effect can affect the kinetics and products distribution within the aerosol deliquescent particles. However, the ionic strength effects on aqueous phase reactions of atmospheric relevance have been barely studied in the past. In this work, we have leveraged our knowledge on ionic strength (I) effects on the reactive uptakes of gas-phase oxidants ozone (O₃) and nitrogen dioxide (NO₂) on methoxyphenols by the well-known wetted wall flow tube technique. The obtained results highlight the importance of the ionic strength effect and suggest a much faster oxidation kinetics of methoxyphenols in aerosol deliquescent particles, compared to that in the dilute aqueous phase of cloud droplets; hence, the oxidative power of O₃ and NO₂ could be enhanced by one order of magnitude at elevated ionic strength, compared to a dilute aqueous phase. Considering the high concentrations of inorganic salts occurring in aerosol deliquescent particles during haze events, the oxidative power of different atmospheric oxidants might become considerably higher than that predicted for oxidation reactions in cloud droplets. The addition of SO₄^2- typical for the liquid core of aerosol deliquescent particles increase the formation of condensed aromatics but also organosulfates (-OSO₃H) arose upon ozone reactions with methoxyphemols. The addition of NO₃^- ions, substantially favors the formation of multi-core aromatic compounds through heterogeneous ozone processing of methoxyphenols. More importantly, the addition of NO₃^- ions favors the formation of nitrooxy (-ONO₂) or oxygenated nitrooxy group of organonitrates, which are potential components of poorly characterized light absorbing organic matter (“brown carbon”) which in turn can affect the climate and air quality.

Severe turbulence encounters can result in serious injuries and millions of dollars of operational costs to airlines. In addition to the turbulence generated within the storm, thunderstorms can induce turbulence in the surrounding clear air, which can extend large distances from the storm boundary due to gravity waves and other circulations. Using relatively limited datasets, past studies have indicated that severe turbulence can extend beyond the current federal guidelines for avoidance (20 mi/~32 km), but many questions remain. For example, 1) What is the spatial distribution of convection induced turbulence relative to individual storms? 2) What dynamical factors determine the spatial distribution? Now, extensive archives of operational radar data and automated eddy dissipation rate (EDR) reports from commercial aircraft present an opportunity for more detailed analysis. To address the above questions, we objectively identify thunderstorms in the GridRad hourly radar archive over the contiguous USA and compare them to automated turbulence reports over a period of 9 years (2009-2017). This analysis, based on millions of turbulence data points, is used to diagnose the risk of turbulence relative to storms.

Clear-air turbulence (CAT) has a large impact on the aviation sector. Our current understanding of how CAT may increase with climate change in future is largely based on simulations from CMIP3 and CMIP5 global climate models (GCMs). However, these models have now been superseded by high-resolution CMIP6 GCMs, which have grid lengths at which median individual turbulence patches may start to be resolved. Using a CMIP6 multi-model approach, projected CAT changes over the North Atlantic have been quantified. Twenty-one CAT diagnostics are used, to represent the uncertainties in CAT production mechanisms. Each diagnostic responds differently in time, but the majority display an increase in CAT between 1950 and 2050. These results refer to Northern Hemisphere (NH) CAT changes, particularly over the North Atlantic basin. Although NH winter is historically the most turbulent season, there is strong multi-model agreement that autumn and summer will have the greatest overall percentage increase in CAT frequency. Future NH summers are projected to become as turbulent as 1950 winters. Using the global-mean seasonal near-surface temperature as a comparative metric, with every of global near-surface warming, NH autumn, winter, spring, and summer are projected to have an average of 14%, 9%, 9%, and 14% more moderate CAT, respectively.

The one-third power of the energy dissipation rate (EDR) is calculated based on the Thorpe method using high vertical-resolution radiosonde data (HVRRD) at 68 operational stations in the United States for six years (2012–2017) observed at 00 and 12 UTC. The spatiotemporal distributions of HVRRD-derived EDR (HVRRD-EDR) are compared with those of in-situ flight EDR (flight-EDR) observed along main flight routes of commercial airlines over z=20–45 kft and within ±1 h centered at 00 and 12 UTC. Both datasets show similar seasonal variations, with the largest total occurrences in June–August (JJA) and the smallest in December–February (DJF). Vertically, both ratios of light-or-greater (LOG) and moderate-or-greater (MOG) intensity turbulence events to the total turbulence events of HVRRD-EDR reveal two local peaks at z=20–23 kft and 41–44 kft, while those of flight-EDR at z=20–23 kft and 35–38 kft. The LOG and MOG ratios of HVRRD-EDR show clear seasonal variations with large values in JJA and small values in DJF, while those of flight-EDR show somewhat different variations: large values in DJF and small values in JJA at z=20–30 kft, large values in March–May and small values in September–November at z=30–45 kft. Consequently, correlation coefficients between HVRRD-EDR and flight-EDR are negative at z=20–30 kft and positive at z=30–45 kft. Horizontally, the MOG ratio of HVRRD-EDR is large primarily over the Rocky Mountains which is consistent with the flight-EDR. The discrepancies in the spatiotemporal distributions between the two datasets likely stem from, at least, two reasons: (i) turbulence observed from the two datasets cannot be the same events, and (ii) HVRRD-EDR associated exclusively with static instability cannot include shear instability under stable atmospheric condition that is related to the Kelvin-Helmholtz instability of which most of flight-EDR is involved.

Turbulence is one of the most significant physical processes in aviation meteorology. Rapid changes in a headwind in the vicinity of an aircraft can lead to aviation accidents. Although many efforts have been made to detect such turbulence in advance, difficulties have been encountered due to its small spatial (1 mm ~ 1 km) and temporal (1 s ~ 1 hr) scales, and unpredictable behavior. By utilizing Doppler lidar to observe turbulence with a high spatiotemporal resolution continuously, it is possible to overcome the limitations of in situ observation and detect turbulence in advance. Energy dissipation rate (EDR) is a valuable measure that can quantitatively represent the intensity of turbulence. Various techniques have been proposed to estimate EDR using Doppler lidar. In this study, we will compare the performance of three inertial range EDR estimation techniques (power spectrum, second-order structure function, variance). They are applied to Doppler lidar measurements operating in the vertically fixed-pointing mode. The accuracy of the estimated EDR was evaluated by the one from in-situ sonic anemometer. The results show that the data sampling rate greatly affected the EDR estimation using the power spectrum. The variance technique produces the most accurate estimation when compared to the value calculated from the sonic anemometer. This study will examine the benefits and weaknesses of each technique, taking into account the instrumental characteristics and scanning strategy of Doppler lidar.

ACKNOWLEDGEMENT
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2021R1A4A1032646). This work was funded by the Korea Meteorological Administration Research and Development Program "Developing Technology for High-Resolution Urban Weather Information Services” under Grant (KMA2018-00627).

The current aviation turbulence forecast systems worldwide, including in South Korea, are mostly based on deterministic forecasts that provide the turbulence potential fields predicted by calculating turbulence diagnostics using the single numerical weather prediction (NWP) model output. Given that both the turbulence diagnostics and NWP model have inherent uncertainties, the need for probabilistic forecasts that can take such uncertainties into account and provide confidence in the forecasting is increasing. In this study, we perform probabilistic turbulence forecasts using the two operational Global Data Assimilation and Prediction System (GDAPS) outputs based on the Unified Model (UM) and the Korean Integrated Model (KIM) of the Korea Meteorological Administration (KMA), and evaluate the performance of the two GDAPS-based probabilistic turbulence forecasts against available turbulence observation data for one year (2022.01–2022.12). Although the UM- and KIM-based GDAPS have a similar horizontal resolution (~10 km for UM; ~12km for KIM), the KIM-based GDAPS has a higher vertical resolution than the UM-based GDAPS (70 levels below 80 km for UM; 91 levels below 80 km for KIM). To perform a single NWP model-based probabilistic turbulence forecast, two probabilistic forecasting approaches are considered, which are a forecast time-lagged ensemble and a multi-diagnostic ensemble. The UM-based deterministic turbulence forecast with a higher horizontal resolution predicts wider and stronger turbulence regions than the KIM-based deterministic turbulence forecast, regardless of the higher vertical resolution of KIM-GDAPS, although the predicted turbulence regions are overall similar. Detailed comparison results of performance validation of two GDAPS-based probabilistic turbulence forecasts will be presented in the conference. Acknowledgment: This work was funded by the Korea Meteorological Administration Research and Development Program under Grant KMI2022-00410.

As global air traffic has steadily increased over the past few decades, the need for more accurate flight time compliance and safe and efficient air operations has arisen. To this end, the ICAO (International Civil Aviation Organization) established a global navigation plan and urged each country to implement it. Accordingly, Korea also launched the National Aviation Plan (NARAE) to realize a safe and efficient air traffic system. To successfully carry out the NARAE, it is necessary to provide detailed and three-dimensional aircraft weather information to support optimal operation for each flight stage in preparation for environmental changes in future air operations. In particular, at airport points, it is essential to provide probabilistic forecasts for decision-making of aircraft operations for factors such as visibility, ceiling, and wind shear, which significantly affect the takeoff and landing of aircraft. In this study, we configured the ensemble system to produce probabilistic forecasts of aircraft meteorological factors. An ensemble system is configured based on the Korea Meteorological Administration's KLAPS (Korea Local Analysis and Prediction System) and employs the stochastic ensemble methods; SPPT (stochastically perturbed physics tendencies) and SKEBS (stochastic kinetic-energy backscatter scheme). We conducted experiments on low visibility and wind shear cases at Incheon Airport. We confirmed the ensemble spread of each experiment by comparing the results of experiments with the different numbers of ensemble members. This work was funded by the Korea Meteorological Administration Research and Development Program under Grant KMI2022-00410. And the main calculations in this study were performed using supercomputing resources provided by the Korea Meteorological Administration (National Center for Meteorological Supercomputing).

The sea-effect generally refers to the process of air mass denaturation after the cold air flows above the warm sea surface. This type of fog is often observed in the western coastal zone of Bohai Sea and has an important impact on the local aviation. There is little research on the heavy fog in this area due to the influence of the "sea effect". This study investigates the mechanism of a typical sea-effect fog case in the morning of October 17, 2007 using WRF model over western coastal zone of Bohai Sea. Results indicated that: contrary to the most analyses of sea-effect snowstorm and precipitation studies that generally emphasize the role of oceanic heating, evaporation of Bohai Sea and accompanying advective vapor transport are shown to play more important roles in the formation of fog. The moisture budget reveals that the evaporation of the Bohai Sea at night is ten times greater than in the western coast. The water vapor transport guided by the back-flow in Northeast China contributes the most to the fog water. The vapor flux in this direction fluctuates and increases from 16:00, Oct 16 to 03:00, Oct 17 (LST), and vertical integral corresponding to the tendency of liquid water reaches a maximum at 03:00, Oct 17. In contrast, the temperature advection is weak. Sensitivity tests further show that the changes of the SST field only cause a shift in the location of the fog, while the humidity change triggered by the Bohai Sea effect plays a decisive role on fog formation in western coastal zone. The sensitivity test of evaporation shows that: with the decrease of evaporation in Bohai Sea, the advective vapor transport will be significantly weakened, and the simulated fog on the western coast of the Bohai Sea will disappear completely.

A modified algorithm which represents additional top-down turbulent mixing driven by cloud/fog-top radiative cooling was provided by Wilson and Fovell in 2018. The algorithm is appended to the Yonsei University (YSU) planetary boundary layer (PBL) parameterization scheme to improve turbulent mixing within the cloud/fog. Considering the modified YSU (YSU_mod) simulation may be affected by different combined parameterizations, three different combinations of parameterization schemes, including the schemes by formula provider recommended (WF), the schemes using by operational numerical prediction of North China Plain (NCP), and the land surface process parameterization in WF schemes being substituted by that in NCP (WF_1), are selected to evaluated fog cases over North China Plain using the Weather Research and Forecasting (WRF) model in order to promote the application of YSU_mod in NCP operational forecast system. The results show that the influence of the additional top-down turbulent mixing in fog is mainly shown as, (1) it significantly increases PBL temperature while the slight and uncertain effect on specific humidity, which lead to less fog layer liquid water and fog cover, and later formation and early dissipation of fog. (2) Under different combinations of parameterization schemes, the influence is that the warmer bottom of the PBL in WF schemes which promotes fog layer lifting, while the warmer upper of the PBL in NCP schemes which lowers fog layer. (3) The warmer fog layer don’t always correspond to the truth, the negative temperature deviation is modified successfully in WF schemes, while the temperature and humidity deviations are larger than observations in NCP schemes. In addition, an appropriate land surface parameterization greatly affects fog simulation which can be concluded from the significant fog simulation improvement in WF_1 comparing to the WF.

Vanuatu is one of the world's most vulnerable countries to natural disasters. The main climate hazards for Vanuatu include tropical cyclones, heavy rainfall resulting in flooding, extended periods without rain causing drought, rising sea levels threatening coastal environments and property, as well as sea temperature increase and ocean acidification impacting highly valuable coastal ecosystems and resources (including coral reefs, seagrass and fisheries). Recognising urgency to assist Vanuatu with climate change adaptation, the Green Climate Fund provided support for "Climate Information Services for Resilient Development Planning in Vanuatu (Van-CIS-RDP)" project, which is named Van-KIRAP (Vanuatu Klaemet Infomesen blong Redy, Adapt mo Protekt) in local language. Van-KIRAP project helps to address the needs to strengthen Climate Information Services (CIS) in Vanuatu delivering climate science to support decision makers and communities in Vanuatu to prepare for and adapt to climate variability and change. World Meteorological Organization's Global Producing Centre for Long-Range Forecasts (WMO GPC LRFs) Melbourne hosted by the Australian Bureau of Meteorology disseminates sub-seasonal to seasonal (S2S) climate information from the Australian Community Climate Earth-System Simulator – Seasonal (ACCESS-S) on global, regional and national scales, including climate forecast products for Vanuatu. This presentation provides an overview of products for climate monitoring and ACCESS-S S2S products for climate prediction disseminated via WMO GPC LRFs, with emphasis on utilising them under Van-KIRAP project for production of Climate Outlook Bulletins for tourism, agriculture, and fisheries sectors.

Persistent heavy rainfall produced by western North Pacific (WNP) tropical cyclones (TCs) can lead to widespread flooding and landslides in Asian countries. On July 2021, unprecedent rainfall amount occurred when Typhoon Infa passed through the highly populated eastern China. While the associated synoptic features have been analyzed, the extreme characteristics and return periods of rainfall induced by In-fa remain unexplored. Analyses of rainfall data from a WNP TC database of the China Meteorological Administration (CMA) show that Typhoon In-fa not only produces record-breaking rainfall accumulations at individual surface stations, but generates unprecedent rainfall amounts for the whole area of eastern China. Quantitatively, 2, 4, 11, 24 and 55 stations are exposed to once in 200-, 100-, 50-, 20- and 10-year extreme TC rainfall accumulations, respectively, and total rainfall at 75 stations reaches a record high since 1980. Overall, the return period is up to 481 years for the total rainfall amount accumulated in eastern China during the 1980–2019 baseline. The extremely long rainfall duration is identified as key to the torrential rains in the Yangtze River Delta before In-fa changes its direction of movement from northwestward to northeastward, while the extreme rain rate plays a dominant role in the northern areas afterwards. Probabilities of occurrence of such an unprecedented TC rainfall event have increased in most (75%) of the eastern China during the period of 2000–2019 compared with those during 1980–1999. Our study highlights the likely increase in risk of extreme TC-induced rainfall accumulations which should be considered in disaster risk mitigation.

Previous studies focused on the intense TCs in the central-southeastern western North Pacific (WNP), whose variability is intimately linked to El Niño-Southern Oscillation and extratropical sea surface temperature anomaly (SSTA) in the Pacific. Compared with them, weak TCs (WTCs) are more numerous and form further northwestward. The great number of WTCs and thereby the landfall cases may also cause huge damage to countries in Southeast Asia. However, their modulators are far from fully understood. Our research emphasizes the delayed impact of the early spring North Atlantic tripole SSTA (NAT) on the WTC formation frequency through the “capacitor” effect of sea ice (SIC) and SST in the Barents Sea. Detailed analysis indicates that a positive NAT may modulate an anomalous high in the Barents Sea-North Europe and decrease the local low cloud cover. Thus, more downward solar radiation tends to heat the local SST and decrease the SIC. This warmer Barents Sea could maintain through the typhoon season and excite a significant southeastward wave train, with several centers in the Arctic, Central Asia, and East Asia. The abnormal easterly to the south of the anticyclone in East Asia facilitates the cyclonic anomaly in the South China Sea, the Philippines, and the subtropical WNP, which reinforces the monsoon trough and favors the WTC formation there. A physical-based empirical model is developed for the WTC frequency, and hindcast is performed from 1979 to 2018. It shows the early spring NAT effectively improves the prediction skill for the WTC frequency, which can be considered as a crucial source of predictability for WTCs.

Tropical cyclones (TCs) over the western North Pacific (WNP) mainly occur in boreal summer and autumn. Using three best track datasets for 1979-2021, this study finds a significant inverse relationship between the relative TC genesis frequency (TCGF) during summer and that during autumn. This strong relationship is characterized with TC activity transitions from active (inactive) season in summer to inactive (active) season in autumn. That is, positive (negative) anomaly of TCGF in summer is followed negative (positive) anomaly in autumn. Such a reversal of TCGF anomaly is generally supported by the changes in large-scale environmental fields over the WNP from summer to autumn. Based on diagnostic analyses of TC genesis potential index, it is suggested that the mid-level relative humidity and vertical motions play major roles for the phase-transition of TCGF anomalies, followed by the secondary roles of low-level vorticity and mid-level meridional shear of zonal wind. Moist static energy budget analyses demonstrate that the phase reversal of climatological moist enthalpy advection by anomalous winds contributes to the phase transitions of anomalous vertical motion and relative humidity over the WNP. Seasonal evolution in sea surface temperature over the Indian Ocean is hypothesized to be associated with the phase transition of TCGF anomaly and corresponding large-scale environmental factors over the WNP.

Seasonal prediction of tropical cyclone (TC) activity has been a hot research theme in the past decades. Usually, the tropical sea surface temperatures (SSTs) provide considerable predictability sources for the western North Pacific (WNP) TC activity. Here, we emphasized that the Chukchi-Beaufort (C-B) and Greenland (GL) sea ice variability is closely linked to the year-to-year variations of the early autumn WNP TC formation frequency (TCF). Observational and numerical evidence proved that the excessive C-B and GL sea ice sustains from August to the following early autumn and triggers the southeastward propagation of the Rossby wave trains originating from the Arctic across Western Eurasia (Okhotsk Sea) to the WNP. The resultant anomalous low pressure over WNP provides suitable environmental conditions for TC formation—the enhancement of the lower-level relative vorticity and water moisture, and the decrease of vertical wind shear. For the reduced sea ice, an opposite situation tends to emerge. The persistent combined sea ice signal makes it a physically meaningful precursor for TCF prediction. The cross-validated hindcast and independent forecast based on both the tropical SST and the Arctic sea ice precursors present that the TCF index is predicted with much higher correlation coefficients than those of the empirical models with only the tropical SST predictors. The results demonstrate that the Arctic sea ice truly promotes the seasonal prediction capability of the WNP TCF.

Understanding how global warming affects tropical cyclone (TC) intensity and precipitation for target regions is essential to preparing for associated damages but detailed processes remain uncertain. This study provides a first quantification of anthropogenic influences on TC characteristics affecting South Korea using convection-permitting model (CPM) simulations (3 km resolution). For the observed four recent TCs that strongly affected South Korea, CPM simulations were performed under current (ALL) and pre-industrial conditions (NAT). The observed sea surface temperature and lateral boundary conditions were used for ALL while changes attributable to human influences (estimated using CMIP6 multimodel simulations) were removed from observed boundary conditions for NAT runs. ALL experiments captured the observed TC intensity and precipitation reasonably. After removing human influences, TC intensity and precipitation were reduced in NAT experiments. Importantly, areas with extreme precipitation (i.e., having precipitation larger than 150 mm) were found to expand by 16~37% in ALL compared to NAT, which was induced by an enhanced upward motion near the TC core and an increase of background water vapor in line with warming. Further, the role of increased moisture was found to become important as TC moves to mid-latitudes. This study provides new insights into how greenhouse warming can intensify TC-induced extreme precipitation over East Asia.

Tropical cyclone (TC) track characteristics in a changing climate remain uncertain. Here, we use downscaled simulations from nine Coupled Model Intercomparison Project Phase 6 (CMIP6) models to investigate the genesis, tracks, and termination of >64,000 synthetic TCs traveling within the Southeast Asia region from the historical era (1881-1900 CE) to the modern era (1991-2000 CE) to the future (2081-2100 CE). Under both a moderate emissions scenario (SSP2-4.5) and a high-emissions scenario (SSP5-8.5), TCs are more likely to form closer to major land masses, terminate farther inland, and move most slowly while over mainland Southeast Asia or in the Bay of Bengal from the historical era to future. Furthermore, TCs become more likely to intensify most quickly in locations near mainland Southeast Asia in the future compared to the historical era, and the rate at which TC maximum winds increase is amplified in the future under both of our modeled scenarios. Additionally, results suggest a northern shift in TC track locations in the northwest Pacific basin, as well as an increased tendency for TC tracks to traverse mainland Southeast Asia in the future compared to the historical era.

Convectively coupled equatorial waves and Madden and Jullian Oscillations (MJO) are known to significantly impact cyclogenesis throughout the tropical ocean basins. Studies show that these waves can modulate cyclogenesis potential index, thus affecting the genesis frequency. Recent studies show that the frequency and intensity of cyclones forming over Arabian Sea and Bay of Bengal have undergone significant trends over the recent decades. In this study we explore the role of tropical intraseasonal oscillations and their long term variabiliity on the observed trends of cyclogenesis over Bay of Bengal and Arabian Sea for a period of 40 years from 1981 till 2020. Long term trends in the factors that determine genesis potential index are analyzed along with their relation to equatorial waves and MJO. The study shows that the trends in the wave modulated of the genesis parameters, namely, relative humidity at mid troposphere, vorticity, vertical wind shear and potential intensity, each play a different role in the observed trends in cyclogenesis for the two basins during pre and post monsoon seasons. The results give an insight into the importance of these subseasonal processes in modulating the long term changes in extreme events in addition to background climate change.

It has been observed that some tropical cyclones (TCs) rapidly intensify within 24-48 hours of landfall along the South China coast, which poses a great challenge in terms of intensity forecasts and subsequent disaster preparedness. Some case studies have shown that an increase in sea-surface temperature (SST) might be responsible for such rapid intensification. In this paper, we show that the SST increase is likely due to solar heating of the ocean over a shallow continental shelf, which can occur for a TC with a small cloud cover. Numerical simulations of TCs with small and extensive cloud cover demonstrate that this is indeed the case. Further studies based on a large number of cases illustrate further that a combination of a high TC translation speed and a small cloud cover has the highest chance of causing a TC to intensify rapidly just before landfall.

Current theories of tropical cyclone (TC) intensification give little direct indication of the role of the TC size in intensity changes, although there are observations showing a relationship. We develop a new model of TC central pressure tendency where the pressure change can be expressed as exponential with a time constant determined by the ratio of radius maximum wind (R_max) and the column inflow or outflow speed. An analysis of observations confirms the relationship which becomes more important for a larger pressure tendency and suggests an upper bound on pressure tendency for a given R_max. The dependence of the pressure tendency on size poses a challenging constraint on the accurate forecasting of TCs in numerical weather prediction and climate models.

Understanding the responses of landfalling tropical cyclones to a changing climate has been a topic of great interest and research. Among them, the recently reported slowdown of tropical cyclone landfall decay in a warming climate engenders controversy. Here, the global climatology of landfall decay, based on the tropical cyclone best-track data available, reveals that the reported trends are uncertain and not universal, but spatial, temporal, data, and methodology dependent such that any claim of a climate trend could be misleading at present. The effective area of moisture supply from the ocean, most likely determined by the landfalling track modes, is demonstrated to be an important factor for the decay. This study provides timely essential clarifications of the current contentious understanding.

Accurately prediction of tropical cyclone (TC) intensity is quite challenging due to multiple competing processes among the TC internal dynamics and the environment. Most previous studies have evaluated the environmental effects on TC intensity change based on TC best-track data which results from both internal dynamics and external influence. This study quantifies the environmental effects on TC intensity change using a simple dynamically based dynamical system (DBDS) model recently developed. In this simple model, the environmental effects are uniquely represented by a ventilation parameter B, which can be expressed as multiplicative of individual ventilation parameters of the corresponding environmental effects. Their individual ventilation parameters imply their relative importance to the bulk environmental ventilation effect and thus to the TC intensity change. Six environmental factors known to affect TC intensity change are evaluated in the DBDS model using machine learning approaches with the best-track data for TCs over the North Atlantic, central, eastern and western North Pacific and the statistical hurricane intensity prediction scheme (SHIPS) dataset during 1982–2021. Results show that the deep-layer vertical wind shear is the dominant ventilation factor to reduce the intrinsic TC intensification rate or to drive the TC weakening, with its ventilation parameter ranging between 0.5–0.8. Other environmental factors are generally secondary, with their respective ventilation parameters over 0.8. An interesting result is the strong dependence of the environmental effects on the stage of TC development. Finally, applications of the DBDS model to real TC intensity prediction are briefly discussed.

Unlike most tropical cyclones (TCs) that begin to weaken when approaching the land, there are some TCs that intensified near-landfall on the contrary. Those TCs bring a great challenge to typhoon forecasts and disaster prevention. Based on three best track datasets JMA, CMA and JTWC from 1980 to 2021, the intensity variation characteristics of tropical cyclones in the Western Pacific within 48 hours before landfall were studied, and 11 common typical near-landfall intensified typhoon cases in the three datasets were selected for composite analysis. The environmental wind vertical shear, relative eddy flux convergence and local radiation flux during landings were calculated using ERA5 reanalysis. The intensity changes of 24hr, 12hr and 6hr before landfall were documented. According to statistical analysis, tropical cyclones whose intensity changes of 24hr before landfall exceeded 20kt and did not weaken within 48hr before landfall were defined as near-landfall intensification cases. Statistics show that those TCs have northwestward prevailing tracks. The high activated period is from July to September, and the probability during nighttime is significantly higher than that of daytime. Composite analysis of selected typical near-landfall intensified TCs shows that the vertical wind shear, which has no significant decreasing before landfall. Further analysis find out that the decreasing of local vertical wind shear and the increase of relative eddy flux convergence can favor the development of inner-core convection. The asymmetric distribution of latent heat, sensible heat flux and long-wave radiation on the underlying surface around the typhoon before landfall is further analyzed. It is found that the asymmetric distribution of local non-adiabatic typhoon circulation is conducive to the development of offshore intensity. Under favorable dynamic and thermal processes. Statistical analysis of such typhoons reveals common characteristics and possible influential factors.

Using validated Weather Research and Forecasting (WRF) model simulation data, kinetic energy (KE) budgets were analyzed to study the dynamic processes associated with the radial movement of maximum wind in the lower troposphere during the rapid intensification (RI) of tropical cyclone (TC) Khanun (2017). In this study, the radial movement of the maximum symmetric rotational KE is computed to measure the movement of the maximum wind. RI is divided into four stages: before RI, the radial distance decreases inward rapidly (stage 1); during RI, the radial distance undergoes a slowly inward decrease (stage 2), barely unchanged (stage 3) and a slowly inward decrease again (stage 4). The comparison between the radii of maximum symmetric rotational KE and its maximum tendency reveals that the radius of maximum symmetric rotational KE generally is larger than that of maximum tendency of symmetric rotational KE in all stages except in stage 3 when maximum symmetric rotational KE and its maximum tendency are nearly collocated. The analysis of symmetric rotational KE budgets shows that flux convergence of KE always is an important term determines maximum tendency in all stages. The conversion from asymmetric rotational KE to symmetric rotational KE, and the conversions from environmental KE to symmetric rotational KE and from symmetric divergent KE to symmetric rotational KE contribute to the maximum tendency, respectively, in stage1 and stages 3 and 4. The important terms in determining maximum tendency are further partitioned and the results will be reported in the conference. Keywords: tropical cyclone, rapid intensification, kinetic energy budget, symmetric and asymmetric circulations, divergent and rotational circulations.

Higher tropical cyclone (TC) intensification rates are affected by smaller radius of maximum wind (RMW) while continuing convection within the RMW can cause RMW contraction. Therefore, understanding the factors affecting convection distribution and RMW is crucial for characterizing TC intensification rates. Previous studies have shown that convection distribution is affected by TC genesis type, but subjective classification of TC genesis doesn’t rely on data distribution. Hence, this study develops a new objective method to classify TC genesis type based on a K-means cluster analysis of critical environmental parameters available in ECMWF Reanalysis v5 (ERA5) data. For a proportionate comparison between intensification rate and RMW, the lifetime maximum intensification rate (LMIR) in each case is estimated in this study. The K-means cluster analysis shows four TC genesis types: (i) monsoon confluence (MC), (ii) easterly wave or north of monsoon trough (EW), (iii) monsoon shear (MS), and (iv) monsoon depression (MD). The positive vorticity area in MC and MS indicates the monsoon trough. MC forms at the east of the monsoon trough while MS forms inside the monsoon trough. EW has a small RMW and a small high positive vorticity area. In contrast, MD has a larger RMW and a larger high positive vorticity area. Each genesis type has different convection distribution compared to previous objective method. Owing to larger circulation and scattered convection, MD has a significantly larger RMW and lower LMIR than EW based on Conover's test. In contrast, EW cases have high specific humidity only around the center. Although MC and MS have medium RMW sizes between those of EW and MD, their LMIR is as high as that in EW due to the aggregated convection.

The occurrence of rapid intensification (RI) of a tropical cyclone (TC) involves multi-scale processes and complex interactions, posing a major challenge in operational forecasting. RI is also a well-trodden path for TCs reaching high storm intensity. The most widely-applied RI threshold is 30 kt intensity change in 24 hours, which is the 95^th percentile of the overwater intensity changes based on National Hurricane Center best-track database in Atlantic and eastern North Pacific basins during 1995-2012. The statistics of the 95^th percentile of intensity change can vary for TCs in other basins or during a different time period. To conduct a process-based study, this work attempts to identify a better RI definition that can most representatively identify the storm structure difference between RI and non-RI periods for TCs in the western North Pacific. Results based on 40-year (1979-2020) TC best-track data from JMA and JTWC are presented. This study explores the differences in storm structure and TC-environment conditions between RI and non-RI periods. It is shown that an RI threshold of 50 (40) kt intensity change in 36 (24) h based on JMA (JTWC) yields the most distinct warm-core structure difference between RI and non-RI periods.

The September rainfall over Northern China (NC) in 2021 was the heaviest since 1961 and had unprecedented socioeconomic impacts. Holding the hypothesis that the drivers of extreme climate events usually contain extreme factors, we firstly propose the Ranking Attribution Method (RAM) to find the possible air–sea multi-factors responsible for this rainfall event. Via the atmospheric bridges of zonal-vertical circulation and Rossby wave energy propagation, the remote factors of warm sea surface temperature anomalies (SSTA) over the tropical Atlantic, cold SSTA over the tropical Pacific, Southern Annular Mode-like pattern in the Southern Hemisphere and North Pacific Oscillation-like pattern in the Northern Hemisphere jointly strengthened the Maritime Continent (MC) convection and Indian monsoon (IM). Through meridional-vertical circulation, the intensified MC convection enhanced the subtropical high over southern China and induced ascending motion over NC. The local factor of extreme air acceleration in the east Asian upper-level jet entrance region further anchored the location of the southwest-northeast rain belt. The strengthened IM and subtropical high over southern China induced considerable moisture transport to the rain belt via two moisture channels. The combined effect of these extreme dynamic and moisture conditions formed this unprecedented rainfall event. This study suggests that the RAM can effectively reveal the factors that contributed to this extreme rainfall event, which could provide a new pathway for a better understanding of extreme climate events.

Sudden heavy rainfalls occur more and more frequently in Japan during the summer. They are concentrated in restricted areas (typically of 5x5 km²) and can cause severe damage to infrastructure, often with casualties. The extrapolation of precipitation measured by radars is the conventional method for carrying out their nowcast in real time, i.e., short-term prediction on small spatial scales. However, the limit of predictability of such storms with current operational nowcasts is less than 10 minutes. This is due to the short lifetime (<15 minutes) of the individual convective cells causing the precipitation, as well as the limitations of nowcast models to properly account for the 3D nonlinear evolution of the cells. This presentation is about AI based real-time nowcasts of heavy precipitation on meso-γ-scale (2-20 km). The X-band Multi-Parameter Phased-Array Weather Radar (MP-PAWR) operating in Saitama prefecture (Japan) provides 3D observations of individual convective cells with a time resolution of 30 sec. A recurrent neural network is trained to detect the cells aloft and extrapolate them up to 10 minutes ahead. The neural network uses conventional Long Short-Term Memory (LSTM) units enhanced with 3D spatial convolutions and a training method developed for Generative Adversarial Network (GAN). It exploits high spatio-temporal resolution of the measurement up to an altitude of 8 km and the dual polarization. The model is trained with observations of 2020 and successfully predicts the onsets of sudden storms that occurred in 2018 and 2019. Its high prediction skills are illustrated with examples of typical storms where the operational system of Japan Meteorological Agency (JMA) or a conventional approach fails. A real-time system has been prepared to continuously analyze the observations of summer 2023 and we will present the first results.

Because of the progress of global warming, heavy rainfall intensity and the total amount are increasing in Japan. The heavy rainfall in Japan is frequently observed especially during summertime and occurred by various causes. One of the causes of the heavy rainfall is stagnated pressure pattern around Japan which guides the moisture transport from southwest toward Japan. With the pressure pattern as a background field, extratropical cyclone, front activity, and tropical cyclone induce the heavy rainfall. Since the air-sea interaction has influence on the formation of the pressure pattern, ocean prediction has a possibility to increase the prediction skill for heavy rainfall, and may extend the prediction lead time. Using regional atmospheric and coupled models, the impact of ocean prediction on summer heavy rainfall in Japan was examined for four events. It was found that the ocean prediction could extend the prediction lead time, when the heavy rainfall was caused by atmospheric rivers. Also, it increased prediction skill in the case of front-derived heavy rainfall. Further investigations showed that ocean prediction can provide better air-sea coupling regarding the Pacific high locating southeast of Japan, leading to provide better moisture transport from southwest of Japan. Analysis on reanalysis data indicates that the air-sea coupling around Japan is atmospheric feedback dominant region. Thus, the ocean prediction is necessary to consider the feedback from the atmosphere. Otherwise, the atmospheric feedback is neglected, and sea surface temperature prevails the atmospheric feedback as seen in tropics, then it causes unrealistic air-sea coupling and deteriorates the prediction. It is concluded that the ocean prediction contributes to increase prediction skill and can extend the lead time of heavy rainfall prediction.

To study the effects of different boundary layer schemes on the simulation of landing attenuation stage of typhoon Meranti (1614), a series of high-resolution (1.33 km) numerical tests were carried out using seven boundary layer parameterization schemes in the mesoscale numerical model WRF v3.8, namely, YSU, MYJ, QNSE, ACM2, UW, GBM, and Boulac, in terms of movement track, intensity, structure, rainfall, and near-surface physical variables. The results indicate the following. First, boundary layer schemes significantly influenced the simulation of typhoon Meranti’s track, intensity, and rainfall during its landing attenuation stage. Second, in the overall simulation of track, intensity, and precipitation of the typhoon, Boulac and MYJ schemes showed optimal results, in which the Boulac scheme was superior in simulating the typhoon track and precipitation and the MYJ scheme was superior in simulating typhoon intensity. The YSU and GBM schemes had the second best simulation results, whereas QNSE, UW, and ACM2 schemes had worse simulation performance. Moreover, the boundary layer schemes significantly differed in calculating the latent heat flux and sensible heat flux of near-surface layer, thereby affecting the simulation of typhoon track, intensity, and rainfall, leading to significantly different simulation results. The QNSE scheme resulted in an abnormally high latent heat flux, the MYJ and Boulac schemes resulted in the most modest values, and other schemes resulted in slightly smaller values. On the other hand, the QNSE scheme had a slightly higher sensible heat flux, the MYJ scheme showed the most modest one, and other schemes resulted in significantly smaller values. Finally, the boundary layer schemes significantly differed in the simulated thermal and dynamical structure of boundary layer, and Boulac scheme had the obvious advantages, particularly for the structure of boundary layer in daytime.

The Maritime Continent (MC) regularly experiences powerful convective storms that produce intense rainfall. Often the intense rainfall leads to flooding and landslides and thus to widespread destruction. At short lead times (0-12 hours), numerical weather prediction (NWP) models show low skill in the MC for forecasting convective activity. Nowcasting aims to solve this issue by using alternative methods to NWP models to make more accurate and reliable predictions of convective activity from observations over this key timescale. Optical flow algorithms are one of the most effective nowcasting methods as they are able to accurately track clouds across observed image series and predict forward trajectories. Optical flow is generally applied to weather radar observations, however, the radar network over the MC is sparse and cannot penetrate the high mountainous regions. In this research, we present the results of applying the Lukas-Kanade optical flow algorithm to infrared satellite imagery to generate 1 – 6 hour lead time nowcasts of heavy-precipitation-producing storms over the MC. For evaluation of the nowcasts, we present a novel adaption of the fractional skill score to quantify how nowcast skill varies spatially and temporally across the MC domain. Overall, the results show that the Lukas-Kanade algorithm has good skill, outperforming a persistence forecast for all lead times and showing skill up to 6 hours on a 50 km spatial scale. Low skill is observed over the mountainous regions in the early afternoon due to the algorithm’s inability to predict convection initiation. Overnight, however, high skill is seen over the sea as the model is able to accurately predict the offshore propagation of storms. These results show that the Lukas-Kanade algorithm can effectively nowcast mature storms and, when analysed alongside the skill maps, has potential to be used in an operational nowcasting system for the MC.

The major precursor phenomena were derived for detecting torrential rain clouds by intuitively analyzing satellite images. There were 17 cases of torrential rains in 2020-2021 in South Korea. The precursor phenomena were derived for satellite images such as water vapor, RGB composite, and secondary outputs from the Geo-KOMPSAT. We summarized major 12 possible check lists of key precursors that could become torrential rain clouds.
1. Updraft zone in front of the boundary of the upper dry area of North pacific High
2. Updraft zone in front of the dry area due to the Low pressure trough
3. Compressed wet zone between the North-South dry zones
4. Warm advection accompanied by Low pressure in warm conveyor belt
5. Lower cumulus clouds from strong southwesterly flow
6. Cirrus cloud as divergent in the upper strong wind zone
7. Upper layer cold core
8. Low-pressure rotating clouds of upper, middle, and lower layers on the stationary front
9. Periodic upper-level wave inflow on the stationary front
10. The cooling rate of the developing convective cloud lasts less than -3℃/10minutes
11. Clouds thickness of 10 km or more
12. Heavy rainfall critical index of 30 mm/hr or higher
If six or more phenomena among the above 12 conditions are present, the clouds would be developed to cause heavy rainfall. In addition, more than 60 mm/hr of possible precipitation, a strong instability index, the speed of movement of the developed clouds, and terrain factors should be considered. It is necessary to prepare for intense rainfall in summer by considering a combination of topographical factors. Several dominant cases with satellite images and checklists were introduced. This study was carried out with the support of the Korea Meteorological Administration R&D program [Meteorological satellite forecast support and convergence service technology development] (KMA2020-00120).

The Meteorological Service Singapore (MSS) maintains and runs a convection-permitting tropical configuration of the Met Office Unified Model (UM) known as SINGV for weather forecasts and climate projections over the Western Maritime Continent. A new Regional Atmosphere and Land (RAL) science configuration that includes, among other changes, a multi-moment cloud microphysics scheme has been proposed for kilometre-scale UM simulations in the mid-latitudes as well as in the tropics. Short-range forecast experiments with SINGV show that, compared to the operational configuration, the new science configuration produces more coherent storm structures with broader stratiform regions and reduces the tendency to over-forecast (under-forecast) precipitation over land (sea). An objective evaluation of the precipitation forecasts reveals significant improvements in Fractions Skill Scores (FSS) across forecast lead times and rainfall thresholds. The new RAL science configuration is a significant step towards better precipitation forecasts in this region.

The Cloud AeroSol Interacting Microphysics (“CASIM”) is the new double moment cloud microphysics scheme that will be introduced into the Met Office operational regional model next year. Results will be shown for performance in the Tropics and the UK. We show comparisons against the currently operational single moment cloud microphysics and investigate the effects of different aerosol activation protocols for activating droplets from the environmental aerosols. In the Tropics, CASIM outperforms the single-moment microphysics as evident from improved comparison with radar derived precipitation rates. We demonstrate that these improvements derive, in part of improved representation of rain evaporation and sedimentation when two prognostic moments are used. CASIM is also used for regional climate projections, and we will show results from dynamical downscaling over China with a particular emphasis on temperature and precipitation extremes. 

Our understanding of the vertical coupling between the lower atmosphere and the upper atmosphere has significantly advanced in recent years, spurred by the large amount of new space-borne and ground-based observations and the extension of atmospheric models into the mesosphere and thermosphere. Observational and model studies have revealed that the mesosphere-lower thermosphere (MLT) region is the nexus where the forcings by gravity waves (GWs), atmospheric tides and planetary waves (PWs) contribute to driving the mean meridional circulation (MMC). Sandwiched between the two summer-to-winter overturning circulations in the mesosphere and the upper thermosphere, the climatological lower thermosphere mean meridional circulation is a narrow gyre that is characterized by upwelling in the middle winter latitudes, equatorward flow near 120 km, and downwelling in the middle and high summer latitudes. Based on the hourly output from the 2000–2014 simulations of the National Center for Atmospheric Research’s vertically extended version of the Whole Atmosphere Community Climate Model (WACCM-X) in specified dynamics configuration, we examine the roles of PWs, GWs and atmospheric tides in driving the mean meridional circulation in the lower thermosphere and its response to the sudden stratospheric warming phenomenon with an elevated stratopause in the northern hemisphere. Following the onset of the sudden stratospheric warmings, this gyre reverses its climatological direction, resulting in a “chimney-like” feature of un-interrupted polar descent from the altitude of 150 km down to the upper mesosphere. This reversal is driven by the westward-propagating planetary waves, which exert a brief but significant westward forcing between 70 and 125 km, exceeding gravity wave and tidal forcings. We present evidence of this circulation in observational data and the observational needs to improve our understanding of its variability.

An international joint research project, entitled Interhemispheric Coupling Study by Observations and Modelling (ICSOM), is ongoing. In the late 2000s, an interesting form of interhemispheric coupling (IHC) was discovered: when warming occurs in the winter polar stratosphere, the upper mesosphere in the summer hemisphere also becomes warmer with a time lag of days. This IHC phenomenon is considered to be a coupling through processes in the middle atmosphere (i.e., stratosphere, mesosphere, and lower thermosphere). Several plausible mechanisms have been proposed so far, but they are still controversial. This is mainly because of the difficulty in observing and simulating gravity waves (GWs) at small scales, despite the important role they are known to play in middle atmosphere dynamics. In this project, by networking sparsely but globally distributed radars, mesospheric GWs have been simultaneously observed in seven boreal winters since 2015/16. We have succeeded in capturing five stratospheric sudden warming events and two polar vortex intensification events. This project also includes the development of a new data assimilation system to generate long-term reanalysis data for the whole middle atmosphere, and simulations by a state-of-art GW-permitting general circulation model using reanalysis data as initial values. By analyzing data from these observations, data assimilation, and model simulation, comprehensive studies to investigate the mechanism of IHC are planned. This paper provides an overview of ICSOM, but even initial results suggest that not only gravity waves but also large-scale waves are important for the mechanism of the IHC.

Radar observations of the troposphere and lower stratosphere made in northern Germany with two VHF MST radars separated by about 30 km in quite different topography over a period of one week are presented. One of the radars, the SOUSY MST radar, was located in the Harz Mountains; the other radar, the Mobile SOUSY MST radar, was operated in a limited form and provides an interesting and rare example of a small beam swinging Doppler radar operating in the lower VHF band (53.5 MHz). Wind results are consistent between the radars, but there are some differences in the measured upward fluxes of horizontal momentum. Wave activity associated with the passage the jet stream over the location of the larger radar in the Harz Mountains is quite evident in the measured winds at the two locations. Stratospheric and tropospheric measurements of aspect sensitivity, mean winds and wave fluxes measured at the two sites from this campaign are presented and discussed.

Bromine nitrate (BrONO₂) is a major reservoir of the active bromine radicals that contribute to the formation of ozone hole. Energetic particle precipitation (EPP) generated NO_x enhancement is known to lead to ozone loss in the mesosphere and upper stratosphere. In the lower stratosphere, however, this NO_x enhancement is believed to restrain ozone depletion by deactive chlorine and bromine species to the reservoir forms (HCl/HBr and ClONO₂/BrONO₂). In this study we report BrONO2 response to EPP for the first time, using the latest BrONO₂ measurement data from the Michelson Interferometer for Passive Atmospheric Sounding (MIPAS) on Envisat. We report a positive and significant correlation between BrONO2 and EPP Ap index above 30km in Arctic. In addition, the Whole Atmosphere Community Climate Model with D-region ion chemistry (WACCM-D) was used to characterize the EPP contribution to BrONO₂ formation.

The mesosphere and lower thermosphere (MLT) region is strongly modulated by gravity wave (GW) perturbations, and the momentum flux of GW is essential in quantitatively understanding and modeling atmospheric structure at all altitudes. In this study, we present the seasonal variations, long-trend of mesospheric density and GW momentum flux observed by the meteor radars at Davis Station (68.6°S, 77.9°E), in Antarctica, Svalbard (78.3°N, 16°E), Tromsø (69.6°N, 19.2°E) in the Arctic. In general, u’w’ decreases and v’w’ increases as altitude increasing, and westward maximum of u’w’ and northward maximum of v’w’ in spring and autumn at upper altitude. the momentum flux u’w’ and v’w’ is anti-correlated with the background winds (u/v). The seasonal variations in the Davis Station meteor radar relative densities in the southern polar mesopause are mainly dominated by an annual oscillation (AO). The mesopause relative densities observed by the Tromsø meteor radars at high latitudes in the Northern Hemisphere show mainly an AO and a relatively weak semiannual oscillation (SAO).

Thermospheric nitric oxide (NO) is a minor species that not only cools the thermosphere through 5.3 µm radiation but also changes significantly during geomagnetic storms. NO column densities derived from TIMED/GUVI data reveal a number of features: (1) NO enhancement is very sensitive to energy input in high latitudes; (2) NO column density and O/N2 column density ratio are anti-correlated on a global scale but show a noticeable shift between the boundaries of NO enhancement and O/N2 depletion regions during geomagnetic storms; (3) Flare time NO enhancement. The features (1) and (2) are likely caused by storm-time meridian circulation that bring high latitude NO enhancement and O/N2 depletion to mid and low latitudes. The feature (3) is caused by increased production of NO. TIMED/GUVI observations of NO serve as a way to monitor the storm-time dynamics and state of the thermosphere for space weather research and operation.

A new dual-frequency atmospheric radar system is built and installed in Langfang Observatory, north China. It utilizes novel two-frequency system design that allows interleaved operation of 53.8 MHz for stratosphere and troposphere wind observation, and 35.0 MHz for meteor observation, which optimizes performance for both ST wind retrieval and meteor trail detection. In dedicated meteor mode, the daily meteor count rate reaches over 40,000 and allow wind estimation of finer time resolutions, such as 15-min and 30-min interval, better than the 1-hour typical of most meteor radars. The uncertainty of the ST wind measurements is better than 2 m/s when estimating the line of best fit with radiosonde winds. Preliminary observation results of typical winter gravity waves (GWs) momentum fluxes in the mesosphere, lower stratosphere and troposphere are also presented.

Polycyclic aromatic hydrocarbons (PAHs), one class of most toxic compounds in atmosphere, showed much longer than predicted lifetimes in ambient air due to heterogeneous scavenging by ozone (O₃) according to previously studies. Here we show heterogeneous reactions with hydroxyl (OH) radical dominates far more over that with O₃ in the degradation of particle-bound PAHs based on chamber simulation on the evolution of plumes from biomass burning, the largest emitter of PAHs on a global scale. The evolution of crop straw burning fumes in a 30 m³ smog chamber revealed that for heterogeneous reactions of PAHs with OH, their second order degradation rate constants were estimated to be 4.4~8.6×10^-12 cm³ molecule^-1 s^-1 at the initial stage of photo-oxidation, and they dropped 4-5 times to to 1.1~1.7×10^-12 cm³ molecule^-1 s^-1 after 1~5×10¹¹ molecule cm^-3 s OH radical exposure, corresponding to a decrease of uptake coefficients of OH (γ_OH) from 2.2±0.9 to 0.5±0.1. The decrease of second order rate constants and γ_OH were mainly caused by the reduction in secondary chemistry and the burial effect of coatings by secondary organic/inorganic aerosols during the atmospheric oxidation. Moreover, the OH oxidation pathway (1.1~1.7×10^-6 s^-1) dominates over ozonolysis pathway (1.7~12.7×10^-7 s^-1) in heterogeneous degradation of PAHs. The lifetimes of particulate PAHs were estimated to be 93.5~219 h, approximately 120 times that obtained previously by exposing pure PAHs to oxidants, implying their stronger long-range transport and more persistent health risks.

A better understanding of spatial-temporal variations of open bio-mass burning in China is required to assess its impacts on the air quality and especially on the heavy haze pollution. The MODIS fire spots data and the calculated burned areas were used in this research, which shows the varying number of fire spots in China from 2013 to 2017, with the highest in 2014 and the lowest in 2016. Meanwhile, the fire spots were found mainly concentrated in three key periods (March-April, June and October-November) and two zones with inter-annual variations of burned areas. In addition, the contribution of major vegetation types burning was studied, the cropland occupied the largest proportion of burned area of more than 70 % in any period time, followed by forest. Finally, from the perspective of climate and human activities, the causes of inter-annual variations were discussed. By comparing the average temperature and precipitation in the two zones from 2013 to 2017, it was found that the burned forest area is positively correlated with the average temperature of the zones and negatively correlated with the average precipitation. Meanwhile, the relationship between the El Niño events and the bio-mass burning was discussed. Finally, by using the datasets of MODIS fire spot, land cover, vegetation cover, bio-mass loading and emission factors, a bio-mass emission model is developed, which is then embedded as an on-line module to an air quality model (WRF-CUACE) to quantitatively assess the impacts of bio-mass burning on surface PM_2.5 concentration in China.

Formaldehyde (HCHO) plays a vital role in atmospheric chemistry and O3 formation. Open biomass burning (OBB) is considered to be an important source of HCHO; however, its quantitative contribution to ambient HCHO remains poorly understood due to the lack of reliable high-resolution emission inventories. In this study, a satellite-based method coupled with local emission factors was developed to estimate the hourly primary emissions of HCHO and volatile organic compound (VOC) precursors from OBB in Guangdong (GD) Province of southern China. Furthermore, the contribution of OBB to ambient HCHO was quantified using the Community Multi-scale Air Quality model. The results suggested that in average OBB emissions contributed 5293 tons of primary HCHO per year, accounting for ~14% of the total anthropogenic HCHO emissions in GD. The ambient HCHO concentration ranged from 0.3 ppbv to 8.7 ppbv during normal days, and from 8 ppbv to 45 ppbv in downwind area during OBB impacted days. The monthly contribution of OBB to local HCHO levels reached up to 50% at locations with frequent fires and over 70% during a forest fire event. Ambient HCHO was heavily affected by primary OBB emissions near the source region and by the oxidation of OBB-emitted VOCs in the downwind area. Secondary HCHO formation from OBB emissions was enhanced during photochemical pollution episodes, especially under conditions of high O3 and low NOx. OBB-emitted ethene was identified as the most important VOC precursor of HCHO and contributed to the formation of ~50% of the secondary HCHO. The HCHO formation potential of cropland fires was 26% higher than that of forest fires. Our results suggest that OBB can elevate ambient HCHO levels significantly. Thus, strict control policies on OBB should be implemented, especially for open-burning agricultural residues in upwind areas on serious photochemical pollution days.

This study provides the first comprehensive inventory of all known pyrocumulonimbus (pyroCb) events observed worldwide (546 confirmed events) over the nine-year period 2013-2021. PyroCbs are a dangerous and severe type of fire weather, which present many hazards to firefighting efforts and communities along the wildland-urban interface. These unique storms also serve as a vertical transport pathway (large chimney) facilitating rapid injection of smoke into the upper troposphere and lower stratosphere (UTLS). This inventory provides insight into basic questions on inter-annual, seasonal, sub-daily, and regional variability of pyroCb, along with potential controlling factors. Development of this inventory has included detailed analysis of the distribution and variability of pyroCb smoke injection altitudes, quantitative estimation of the aerosol mass associated with each stratospheric plume, and examination of the impact of pyroCb activity on stratospheric aerosol loading worldwide. This pyroCb inventory provides the means to address a wide range of significant open questions about the nature, behavior, and impact of this phenomenon. Answers to these questions are critical for advancing pyroCb prediction capabilities to mitigate aviation hazards and aid firefighting efforts. This new multi-year inventory dataset also sets a foundation for an official Earth System Data Record that can be maintained into the future and extended back in time to identify longer-term trends in pyroCb activity and ensuing impacts on the climate system.

Pyrocumulonimbus (pyroCb) are fire-induced and smoke-infused thunderstorms that serve as the primary pathway for smoke to reach the upper troposphere and lower stratosphere (UTLS).The magnitude of smoke plumes observed in the UTLS has increased significantly in recent years, rivaling or exceeding the impact from all volcanic eruptions observed over the last decade, with the potential for significant climate feedbacks on seasonal and hemispheric scales. The Black Summer fire season of 2019-2020 in southeastern Australia contributed to an unprecedented pyroCb ‘super outbreak’ that took place over 51 non-consecutive hours. More than half of the 38 observed pyroCb updraft pulses injected smoke particles directly into the stratosphere, producing two of the three largest smoke plumes observed at such altitudes to date. Over the course of three months, these plumes encircled a large swath of the Southern Hemisphere while continuing to rise, in a manner consistent with existing nuclear winter theory. Fewer than three years earlier, a large pyroCb outbreak in Canada produced a persistent smoke plume that encircled a portion of the Northern Hemisphere. We summarize what the community has learned from these extreme events and identify science questions that remain unanswered. A recently-developed pyroCb inventory for 2013-2021 facilitates the first analysis of regional, seasonal, monthly, and inter-annual variability worldwide. Unique in-situ and remotely-sensed measurements of pyroCb activity observed during the 2019 FIREX-AQ field experiment identify the fire characteristics, cloud microphysical properties, and smoke plume chemistry associated with this extreme fire-weather phenomenon.

Wildfire events are increasing globally which may be partly associated with climate change, resulting in significant adverse impacts on local, regional air quality and global climate. In September 2020, a wildfire event occurred in Souesmes (Loir-et-Cher, Sologne, France), and its plume spread out over 200 km on the following day as observed by the MODIS satellite. Based on measurements at a suburban site (~50 km northwest of the fire location) in Orléans, young wildfire plumes were characterized. Significant increases in gaseous pollutants (CO, CH₄, N₂O, VOCs, etc.) and particles (including black carbon) were found within the wildfire plumes. Emission factors, defined as EF (X) = ∆X/∆CO (where X represents the target species), of various trace gases and black carbon within the young wildfire plumes were determined accordingly and compared with previous studies. Changes in the ambient ions (such as ammonium, sulfate, nitrate, chloride, and nitrite in the particle- and gas- phase) and aerosol properties (e.g., aerosol water content, aerosol pH) were also quantified and discussed. Moreover, we estimated the total carbon and other gas (e.g., CO) emissions that were compared with fire emission inventories. We found that the Global Fire Assimilation System (GFAS) may underestimate emissions (e.g., CO) of this small wildfire while other inventories (GFED, FINN) showed significant overestimation. Considering that few wildfires are recorded in this region, related atmospheric implications are presented and discussed. More details can be found in Xue, Chaoyang, et al. "A study on wildfire impacts on greenhouse gas emissions and regional air quality in South of Orléans, France." Journal of Environmental Sciences (2022).

The Global Satellite Mapping for Precipitation (GSMaP) produces high-resolution and high-frequent global rainfall map based on multi-satellite passive microwave radiometer observations with information from the Geostationary InfraRed (IR) instruments (Kubota et al. 2020). Outputproduct of GSMaP algorithm is 0.1-degree grid for horizontal resolution and 1-hour for temporal resolution. Images and data of the GSMaP are available at JAXA GSMaP website (http://sharaku.eorc.jaxa.jp/GSMaP/). The GSMaP has been developed by the Japan Aerospace Exploration Agency (JAXA). The JAXA has participated in the World Meteorological Organization (WMO) has been initiated the Space-based Weather and Climate Extremes Monitoring (SWCEM) (Kuleshov et al. 2020) and provided 22-yr GSMaP data. Currently, the gauge-adjusted near-real-time version 6 (GNRT6) of the GSMaP (Tashima et al. 2020) has been distributed to the SWCEM.In December 2021, the GSMaP algorithm was updated to version 8 by implementing various improvements to better estimate precipitation (Kubota et al. 2022). We plan the reprocessing of the GSMaP version 8 in a period during the past 24 years ”since Jan. 1998” using JAXA super computer system (JSS3). Past GSMaP products did not cover the first 2 years of TRMM era (1998-Mar. 2000) due to lack of NOAA CPC-4km Global IR dataset. There is a possibility that GridSat-B1 data (Knapp et al. 2011) enables to fill the lack of period. In order to extend the period of the GSMaP before 2000, the validity of GridSat-B1 data is investigated in this study.

Satellite rainfall estimates provide valuable spatial information of precipitation, although biases in the estimate need to be corrected, especially for regions near the equator with large spatial and temporal variability in precipitation. It was found that the large variability in the bias can significantly reduce the benefit of quantile-mapping methods for the calibration of the satellite rainfall estimates with gauge measurements. Further investigations found that the bias is possibly related to the type of weather system and this bias can be sub-divided into different categories accordingly. In this presentation, the results of the calibration of weekly and sub-weekly rainfall from the SWCEM-EAWP (Space-based Weather and Climate Extremes Monitoring - East Asia West Pacific) products (CMROPH-CRT and GSMaP_GNRTv6) using rainfall measurements from subset of gauges over Singapore will be presented. The categorization of the rainfall to the respective bias group is first trained using machine learning methods and followed by quantile-mapping for calibration. The improvement in the performance in the machine-learning aided quantile mapping is then evaluated with the full set of rain gauges in Singapore, and compared to the satellite rainfall estimates before calibration and with just quantile mapping calibration applied.

Quite a few real-time forecasts of the annual frequency of tropical cyclones (TCs) have been made in recent years, using statistical and dynamical approaches. However, such forecasts are not very useful for disaster preparedness. Since 2015, we have been making seasonal forecasts of the number of TCs making landfall in different regions along the East Asia coast using a regional climate model with boundary conditions from the Climate Forecast System (CFS). In this paper, how such forecasts are made will be presented together with verification statistics for the last six years. A possible extension of this approach to forecast the intensity of these landfalling TCs will also be discussed.

Marine Heatwaves (MHWs) have dire impacts on aquatic ecosystems and the communities that rely on them for livelihood, recreation, and cultural identity. In Vanuatu, a MHW event in 2016 led to record high sea surface temperatures with widespread reports of fish mortality. Given the severity of impacts, monitoring and predicting MHW events is critical in allowing communities and key sectors to effectively manage (and where possible) mitigate the impacts of MHW events. In this study, we use in situ data, satellite data and dynamical forecasts to derive MHW climatologies and investigate key MHW events. Results for the common period of overlap between all examined datasets (1981-2018) will be presented.

Droughts are a cyclical feature of Australia's climate and have compounding effects on agricultural productivity and wellbeing. Understanding future conditions in context of antecedent observations is critical to providing informed early warning of drought. In this study we pair probabilistic seasonal forecasts with monitoring datasets to provide early warning of drought. Hindcasts from the Bureau of Meteorology's sub seasonal to seasonal forecast model, ACCESS-S2, are paired with MSWEP satellite blended precipitation data and AWRA-L water balance modelled soil moisture and evapotranspiration. Principal Component Analysis (PCA) is used to derive objective weightings to combine precipitation, soil moisture and evapotranspiration percentiles in a multivariate manner similar to the U.S. Drought Monitor. The final DEWS maps overlay forecasting information with PCA-weighted antecedent conditions. We produce 1-, 3- and 6-month maps and analyse drought concern over the common period of overlap between our datasets (1981-2018) and conduct case studies for the 1982-1983 Ash Wednesday tinder drought and the 1997-2001 Millennium drought. We validate PCA-weighted maps with satellite vegetation data and find performance is strongest over the Murray Darling Basin region (R = 0.63, p= 0.009) and poorest over Central interior Australia (insignificant correlations). We also validated PCA-weighted maps using agricultural commodity data from ABARES. Significant negative correlations at 95%, 99% and 99.9% confidence intervals were found between %-Area in drought category and crop cultivation area; export volume/value; crop yield; and rural debt. Our findings indicate that early warning of drought can be categorised by concern – wherein dry antecedent conditions and dry forecasted conditions are of highest concern. Our proof-of-concept drought early warning system contributes to the growing body of proactive drought research. In a drought vulnerable future, operationalising and communicating drought early warnings will be critical to reducing the harmful impacts of drought on economies, environments, and people.

There is a near consensus among scientists that the average global temperature will increase in the next several decades. Many international organizations are taking steps to reduce this anticipated phenomenon by as much as possible. However, there are many forces, social and economic interests that make this goal almost impossible to achieve. Significant climate changes will occur and the world must be prepared. The talk will discuss the top ten reasons why the efforts taken by certain governments and motivated people are not going to be enough. The USA supreme court decision from June 2022; Increased world population; Rapid increase in GDP mainly in developing countries; Increased global demand for electric power; The coal oil and Gas Industry; Increased demand for air and ocean transportation and the tourism Industry; The increased demand for food, mainly but not only from live stock; Increased demand for raw materials mainly steel and cement; The Garment industry and Production of solar panels and storage devices and 

The zonal sea surface temperature (SST) gradient over tropical Pacific is important for global climate and known to be a pacemaker of global warming. During recent decades, observations show a “La Niña-like” strengthening of the zonal SST gradient, whereas most climate models produce an “El Niño-like” weakening of this gradient. There is also large model uncertainty in the future projection of this gradient change under global warming. Quantifying the sources and understanding the mechanisms for the uncertainty is needed to reconcile the discrepancy between observations and models and reduce model uncertainty of future projections. Here we use a total of 342 simulations, including six large ensembles, 36 CMIP5 and 36 CMIP6 models, to quantify the relative contribution of internal variability and model spread to the uncertainty in this gradient change on multidecadal timescales and future projections. On multidecadal timescales, the uncertainty of zonal SST gradient change over the tropical Pacific is dominated by internal variability. Although external forcing is increasing under global warming, the ratio of internal uncertainty is decreasing but still >80% at the end of the 21^st century. The Pacific Decadal Oscillation (PDO) is revealed as the key internal mode influencing the uncertainty of the multidecadal changes. The PDO-related SST pattern exhibits a pronounced zonal gradient over the tropical Pacific. For future projection, the uncertainty of zonal SST gradient change over the tropical Pacific is mainly from the model spread in response to external forcing. Based on future change between 2070-2099 and 1961-1990, model spread accounts for ~70% of the total uncertainty. The SST change in the western and eastern tropical Pacific dominates the total uncertainty of the zonal SST gradient for CMIP5 and CMIP6 models, respectively. That is mainly due to the stronger low cloud positive feedback over the southeastern Pacific among CMIP6 models.

Satellite altimetry is a matured technology that provides accurate measurements of ocean geophysical information of sea surface heights (SSHs), significant wave heights (SWHs), and wind speed. It was designed specifically for observing ocean’s geo-physicality and dynamics. Furthermore, these technologies also have the potential to be developed to study the ocean and atmosphere, simultaneously. It is well known that satellite altimetry is also equipped with sensors to measure atmospheric components to provide corrections for the range measurement. On-board microwave radiometer measures atmospheric water vapour to provide wet path delay correction for the main altimeter range measurement. This perspective develops a new potential for satellite altimetry to also be exploited to observe the atmosphere. It has been proven that the interaction of ocean and atmosphere have a significant role on influencing climate phenomena and their variations through its complex process. Several studies also found that the uncommon reactions between anomalies in atmosphere and ocean hydrological cycle occurred due to rapid climate changes resulting in hydro-meteorological catastrophe. These situations make a thorough study about the interaction between ocean and atmospheric component are necessary. Thus, this study is performed to leverage satellite altimetry sea level anomaly and water vapour measurement to analyse the complex interactions between the ocean and atmosphere to determine the nature and characteristics of these interactions. Using both ocean and atmosphere measurement from satellite altimetry not only present a new potential of the satellite itself, but also a way to overcome the drawback of previous ocean-atmosphere studies, which has temporal and spatial disparities as each parameter were observed using different platforms at different time and locations. Moreover, by multi-missions observations, spatial and temporal resolution are enhanced, making it reliable enough to monitor long periodic phenomena, such as El Niño–Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), Madden-Julian Oscillation (MJO), and monsoons.

Predicting tropical cyclone intensity changes, especially rapid intensification (RI), has been a more challenging topic than tropical cyclone tracking because of its multi-scale physical process with significant contributions from convective-scale phenomena. Before intensification, tropical cyclones experience precession process, in which tilted vortices rotate counterclockwise around the center of circulation, and develop an axisymmetric structure. The forecast uncertainty in precession process limits the predictability of early-stage intensification of TCs. In this study, we have explored the contribution of moist convective activity to the variability of TC precession process using a simple toy-model of the vortex precession process, together with fully dynamic ensemble-based sensitivity experiments using convection-permitting Weather Research and Forecasting model (WRF-ARW) simulations. The results indicate the potential existence of a threshold in vortex strength and structure that governs whether it completes precession and initiates RI. Given the strong nonlinearity of the onset process of RI, the advancement of our understanding of the uncertainty sources will provide an insight about the observation network that may effectively constrain the TC forecasting.

As an artificial intelligence method, machine learning (ML) has been widely used in prediction models of high-dimensional datasets. This study proposes an ML method, the Gradient Boosted Regression Tree (GBRT), to predict the intensity changes of tropical cyclones (TCs) in the Western North Pacific at 12-, 24-, 36-, 48-, 60-, and 72-h (hr) forecasting lead time and the model is optimized by the Bayesian Optimization algorithm. The model predictands are the TCs intensity changes at different forecasting lead times, obtained from the best track data of the Shanghai Typhoon Institute (STI) and the Joint Typhoon Warning Center (JTWC) from 2000 to 2019. The model predictors are the synoptic variables, climatological and persistent variables derived from the reanalysis data obtained from the National Centers for Environmental Prediction (NCEP), and the sea surface temperature (SST) data obtained from the National Oceanic and Atmospheric Administration (NOAA). The results show that the GBRT model can capture the TCs intensity changes well for the succeeding 12-h, 24-h, 36-h, and 72-h. Compared with the traditional multiple linear regression (MLR) model, the GBRT model has better performance in predicting TCs intensity changes. Compared with the MLR model, R² of the GBRT model for TCs intensity forecast increases by an average of 8.47% and 4.45% for STI data and JTWC data. MAE (RMSE) drops by 26.24% (25.14%) and 10.51% (4.68%) for the two datasets, respectively. The potential future intensity change (POT), the intensity changes during the previous 12 h (Dvmax), Initial storm maximum wind speed (Vmax), SST, and the Sea-Land ratio are the most significant predictors for the GBRT model in predicting TCs intensity change over the Western North Pacific.

The aim of this study is to enhance our understanding about the intensification phase of tropical cyclones that occurred in the southwest Indian Ocean. Tropical cyclone (TC) Batsirai (2022) which unexpectedly and rapidly intensified into a Saffir-Simpson Category 4 cyclone (lowest sea level pressure of 934 hPa) after undergoing an eyewall replacement cycle was selected in this study. During the passage of the intensified tropical cyclone (1−4 February 2022), Mascarenes Islands (i.e. Rodrigue, Mauritius, Réunion) were heavily impacted though the effects were relatively minor; the major devastating hazards were reported in Madagascar (121 deaths). The French cloud-resolving mesoscale model, Meso-NH, successfully reproduced the lifecycle of TC Batsirai. In this study, the intensification phase of TC Batsirai has been focused to understand the physical and thermodynamical processes responsible for the intensity change of the tropical cyclone. In addition, the impact of the orography of Mascarenes Islands on localized heavy rainfall and strong winds has been studied. The detailed analysis using the numerical simulation results with fine horizontal grid spacings (250 m) has been done. The key results will be presented at the conference.

In this study, kinetic and potential energy budgets of Typhoon Mujigae (2015) are compared in slow (SI) and rapid (RI) intensification periods. The high-resolution numerical simulation data of Mujigae are validated with observations and are used for comparison analysis. The analysis of kinetic energy budgets reveals that the conversion between asymmetric rotational kinetic energy and symmetric rotational kinetic energy play a key role in determining the differences in the tendencies of symmetric and asymmetric rotational kinetic energy between RI and SI in the lower troposphere. Asymmetric rotational kinetic energy is converted to symmetric kinetic energy in RI, whereas symmetric rotational kinetic energy is converted to asymmetric kinetic energy in SI. This can explain why the symmetric rotational kinetic energy tendency is larger in RI than in SI, but asymmetric rotational kinetic energy tendency is smaller in RI than in SI. In both RI and SI, both conversions from symmetric divergent and environmental kinetic energy to symmetric rotational kinetic energy contributed to the increase in symmetric rotational kinetic energy. The analysis of symmetric divergent kinetic energy budget shows that conversion from symmetric divergent kinetic energy to symmetric rotational kinetic energy is nearly balanced out by the work done by symmetric divergent winds against pressure gradient, which is further related to the conversion from potential energy mainly due to the latent heat release. The detailed results will be presented in the conference.

Tropical cyclones (TCs) moving poleward to mid-latitudes in the western North Pacific basin are complex to forecast their track since their tracks are largely influenced by the upper-level synoptic fields in subtropical and mid-latitude regions (i.e., western North Pacific subtropical high, mid-latitude trough, and jet stream). Thus, forecast performances are especially poor when TCs approach the East Asian region. It is known that soil moisture (SM) modulates the total available energy into sensible heat fluxes (SHF) and latent heat fluxes (LHF) at the land surface. It can affect the atmospheric temperature, clouds, and even precipitation through land-atmosphere interaction. Thus, those land surface processes are necessary for accurate weather forecasts or climate predictions. Therefore, we prescribed two types of SM data for the initial condition of the Weather Research and Forecasting (WRF) model to investigate the sensitivity of the SM initialization on the TC forecasts. The Global Land Data Assimilation System (GLDAS) Version 2.1 and the European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis v5 (ERA5) were selected in this study. The result showed that ERA5 SM data had more wet biases than GLDAS data over the whole simulations except for some areas located south of Mongolia. Also, TC intensity forecast performances were similar in both runs, while TC track forecast performances were improved in the experiments with GLDAS data. The differences in track forecasts between the two runs were more considerable during the landfall period. In addition, in the experiments with ERA5 data, simulated TCs tended to move westward compared to the experiments with GLDAS data due to the strengthened interaction between simulated TCs and the mid-latitude trough. This study showed that it is important to provide realistic SM data to improve the TC track forecast performances during the TC landfall period.

Predicting Tropical Cyclogenesis has long been a challenging research task. Machine learning methods recently have been shown to withhold a high prediction skill for forecasting tropical cyclogenesis. In this study, we first use Kalman filters applied on the Gaussian-smoothed 850-hPa vorticity field to label tropical disturbances. Dynamic and thermodynamic variables, such as wind shear and sea surface temperature, are extracted following the disturbance center in a Lagrangian sense. Three machine learning models with varying complexity: a random forest, a support vector machine and a neural network, are trained to classify developing and non-developing tropical disturbances at a forecast lead time of 24 hours. All of the three models have displayed consistently good performance, with the neural network having the highest f1-scores of 0.79. SHAPley values analysis shows that the mid-level vorticity at 500 hPa is the most essential feature in deciding if a tropical disturbance is developing or non-developing for all the models. Wind shear and tilting are found to possess a certain level of importance as well. Our results suggest that rich mid-level cyclonic vorticity in a mesoscale convective vortex is instrumental in the subsequent low-level spin-up of the disturbance and is an essential sign for the prediction of Tropical Cyclogenesis.

The prediction of tropical cyclones (TCs) has progressively improved in the past few decades. How long can we further extend skillful TC prediction? The answer to this question relies on in-depth studies about TC intrinsic predictability. In this study, the HWRF-based convection-resolving ensemble forecasts were adopted with perturbed initial conditions to study the error growth and intrinsic predictability of TCs. The new aspect of our study is the focus on the sensitivity of TC track and intensity predictability on region-dependent initial errors. The main conclusions include: (1) the most sensitive region for TC track forecasts is case-dependent, for TC intensity forecasts, however, consistently locates within the TC vortex. (2) The 0-2 wave number structures of the TC inherent vortex flow can be predicted for more than 4 days, which is primarily determined by the predictability of the synoptic-scale environment. In contrast, the remaining components (wave numbers >2) are only predictable for a few hours. These finer scales, despite much less variance than larger scales, saturate rapidly and intensively limit the predictability of TC intensity which is generally defined by the maximum surface wind at a certain point in time and space rather than the azimuthal mean wind.

With the nonhydrostatic axisymmetric cloud-resolving model CM1, two sets of ensemble runs are conducted to study the influence of initial vortex depth on the early development of an idealized tropical cyclone (TC). It is found that with the shallow initial vortex, the intensification rate (IR) is smaller and the TC takes a longer time to reach a state of slantwise moist neutrality. These behaviors are primarily ascribed to the less active convection in the shallow case. In contrast to the deep initial vortex, the shallow initial vortex is characterized by a more stable stratification in the lower troposphere and a smaller gradient of angular momentum (M) in the mid-to-upper troposphere. The former tends to hinder the occurrence of deep convection in the TC while the latter makes it difficult to intensify the mid-toupper level vortex and to establish a deep vortex. Thus, deep convection bursts intermittently in the early developing stage of the TC originating from the shallow vortex. During the periods of inactive convection, the TC weakens by the decline of M source associated with reduced horizontal advection, which is overwhelmed by sinks of M from moderately decreased vertical advection and nearly-constant dissipation term. Therefore, the IR of the TC is smaller in the early developing stage and the development of TC from the shallow vortex usually needs a longer time to build the steady deep secondary circulation favorable for more rapid intensification.

A record-breaking precipitation event caused 398 deaths and 20.06 billion RMB economic losses in Henan Province of China in July 2021. A maximum 24-h (1-h) precipitation of 624 mm (201.9mm) was observed at the Zhengzhou weather station. However, all global operational forecast models failed to predict the intensity and location of maximum precipitation for the event. This high social impact event has drawn much attention from the research community. This presentation provides a high-level review of the event and its research from the perspectives of observations, analysis, dynamics, predictability, and the connection with climate warming and urbanization. Global reanalysis revealed obvious abnormality in large-scale circulation patterns that resulted in abundant moisture supplies in the region of interest. Recent studies of this event also revealed, via high-resolution model simulation and data assimilation, that three mesoscale systems (a mesoscale low pressure system, a barrier jet and downslope gravity current) contributed to the local intensification of the rainstorm. Further, observational analysis suggested that an abrupt increase of graupel through microphysical processes contributed to the record-breaking precipitation. Although these findings aided in our understanding of the extreme rainfall event, preliminary analysis indicated that the practical predictability of the extreme rainfall for this event was rather low. The contrary influences of climate warming and urbanization on precipitation extremes as revealed by two studies could add further challenges to the predictability. We concluded by emphasizing that data sharing and collaboration between meteorological and hydrological researchers would be crucial in the future research on high-impact weather events.

Climate risks from compound extremes are much greater than the sum of their sub-components. In early summer 2022, record-breaking rainstorm and heatwave were respectively observed in the south and north of eastern China and they formed an unprecedented compound extreme (hereafter referred to as SRNH). However, most of seasonal predictions failed to capture the SRNH and the behind mechanisms are still unclear. Here we show that the wavier mid-latitude westerlies and the ridge of the western Pacific subtropical high stabilized at 20 °N were the main causes in the atmosphere and the ultra-strong 2022 SRNH event could be well reproduced. The February-March sea surface temperature anomalies in the North Atlantic concurrently promoted the 2022 SRNH extreme, which was successfully verified by numerical experiment. Main physical mechanism is the warmer sea surface excited a Rossby wave train to enhance the contrast between hot North China and wet South China.

An extremely heavy rainfall event lasting from 17 to 22 July 2021 occurred in Henan Province of China, with accumulated precipitation of more than 1000 mm over a 6-day period that exceeded its mean annual precipitation. The present study examines the roles of persistent low-level jets (LLJs) in maintaining the precipitation using surface station observations and reanalysis datasets. The LLJs triggered strong ascending motions and carried moisture mainly from the outflow of Typhoon In-fa (2021). The varying directions of the LLJs well corresponded to the meridional shifts of the rainfall. The precipitation rate reached a maximum during 20−21 July as the LLJs strengthened and expanded vertically into double LLJs, including synoptic-weather-system-related LLJs (SLLJs) at 850–700 hPa and boundary-layer jets (BLJs) at ~950 hPa. The coupling of the SLLJ and BLJ provided strong mid- and low-level convergence on 20 July, whereas the SLLJ produced mid-level divergence at its entrance that coupled with low-level convergence at the terminus of the BLJ on 21 July. The formation mechanisms of the two types of LLJs are further examined. The SLLJs and the low-pressure vortex (or inverted trough) varied synchronously as a whole and were affected by the southwestward movement of the WPSH in the rainiest period. The persistent large total pressure gradient force at low levels also maintained the strength of low-level geostrophic winds, thus sustaining the BLJs on the synoptic scale. The results based on a Du-Rotunno 1D model show that the Blackadar and Holton mechanisms jointly governed the BLJ dynamics on the diurnal scale.

This study investigates the influences of urban land cover on the extreme rainfall event over the Zhengzhou city in central China on 20 July 2021 using the Weather Research and Forecasting model at a convection-permitting scale [1-km resolution in the innermost domain (d3)]. Two ensembles of simulation (CTRL, NURB), each consisting of 11 members with a multi-layer urban canopy model and various combinations of physics schemes, were conducted using different land cover scenarios: (i) the real urban land cover, (ii) all cities in d3 being replaced with natural land cover. The results suggest that CTRL reasonably reproduces the spatiotemporal evolution of rainstorms and the 24-h rainfall accumulation over the key region, although the maximum hourly rainfall is underestimated and displaced to the west or southwest by most members. The ensemble mean 24-h rainfall accumulation over the key region of heavy rainfall is reduced by 13%, and the maximum hourly rainfall simulated by each member is reduced by 15–70 mm in CTRL relative to NURB. The reduction in the simulated rainfall by urbanization is closely associated with numerous cities/towns to the south, southeast, and east of Zhengzhou. Their heating effects jointly lead to formation of anomalous upward motions in and above the planetary boundary layer (PBL), which exaggerates the PBL drying effect due to reduced evapotranspiration and also enhances the wind stilling effect due to increased surface friction in urban areas. As a result, the lateral inflows of moisture and high-θe (equivalent potential temperature) air from south and east to Zhengzhou are reduced.

Moisture transport, associated with moisture sources and synoptic-scale weather conditions, is a key dynamic process of precipitation events. Using the K-means clustering method and the FLEXPART Lagrangian particle dispersion model, this paper investigates moisture contributions from different source regions to extreme precipitation in the first rainy season (hereafter FRS) over South China. In average, land regions contribute more to the FRS extreme precipitation over South China than the ocean regions. The main source regions are Southeast Asia (22.7%), South China (17.2%), the South China Sea (14.3%), and the Bay of Bengal (8.3%). Extreme precipitation events are classified into three types by the K-means clustering based on 850 hPa geopotential height, which are all characterized by an anomalous low-pressure system over South China with varying intensity and locations. The distribution of geopotential height anomaly for Type I (30.3%) is characterized the low trough extending from Japan to South China, while Type II (42.5%) and Type III (27.2%) are characterized by “west negative–east positive” and “north positive–south negative” patterns over East Asia with anomalous cyclone over South China, respectively. The much larger contribution of land sources than ocean regions are mainly concentrated in Type I and Type III, of which the contribution from each source region is similar. Ocean sources play a more important role in Type II and are mainly from the Indian Ocean (16.2%) associated with the onset of South China Sea summer monsoon.

During the pre-summer rainy season in southern China, MCS-associated heavy rainfall frequently occurs in the warm sector hundreds of kilometers to the south of a front or without any front, which is one of the major contributors for coastal flooding events during this period. The intensity and frequency of strong convective storms are expected to rise with warming climate, posing a greater threat of extreme rainfall in the future. This study aims at revealing the future changes of coastal warm-sector MCS in southern China on its related precipitation features and essential mesoscale process associated based on quasi-idealized WRF simulations and pseudo global warming (PGW) approach. Typical warm-sector heavy rainfall events are selected to produce composite environments that force the quasi-idealized simulation in current climate (CTRL). After that, the climate sensitivity experiment (PGW) is conducted with the same configurations except that it is forced by reanalysis data plus a climate thermodynamic perturbation derived from a 32-model CMIP5 ensemble climate change signal for SSP5-8.5 scenario. Comparisons between PGW and CTRL reveal that 12-h accumulated area-averaged rainfall increases 51% by the end of 21st century, with the maximum rising from 333.1 mm to 703.4 mm. As for convection population, the frequency of convection weaker (stronger) than 20dBZ increases (decreases). The rainfall distribution and rainfall maximum are characterized by a north shift, which is largely due to stronger southerly LLJ in warming climate. Moreover, the MCS is initiated ~1h earlier in PGW, with more unstable inflow and stronger back building process leading to larger rainfall accumulation. Notably, although dynamic climate change signal is not included in the current experiment, the LLJ is significantly enhanced by increasing latent heating. The LLJ, latent heating, and deep moist convection may therefore link as a positive feedback in warming climate and eventually lead to extreme rainfall.

Extreme precipitation events cause damage to the natural environment and human society, but their jointly evolution behaviors in both time and space dimensions have not been extensively studied. By introducing a spatiotemporally contiguous events tracking (SCET) method, here we examine the climatologies and trends of the 3D (latitude×longitude×time) characteristics of spatiotemporally contiguous extreme precipitation events across the globe under the past and future climate warming during 1980 and 2020. The results show that the SCET can well demonstrate the jointly dynamic evolution patterns of contiguous precipitation events in time and space dimensions. Spatially, large contiguous precipitation events tend to be concentrated southeastern Asia, Europe and central Africa. Temporally, the frequency of contiguous precipitation events shows an overall increasing trend, but their affected area, intensity, duration and movement showed different characteristics. The future changes of spatiotemporally contiguous precipitation are further investigated based on a set of CMIP6 model simulations. In the coming decades (e.g., 2020–2100), the affected area, intensity and moving speed of contiguous precipitation events will increase. These trends are especially more prominent under high greenhouse gasses emission scenarios (e.g., SSP585). It is expected that extreme precipitation events that are more intense, affect larger areas, and move faster will cause even more harm to human society. Our results provide new insight to understand the spatiotemporal evolutions and future changes of precipitation extremes.

This study explores the coupled variability of temperature and precipitation in eastern China during summer using multivariate EOF analysis. Two leading modes are identified and analyzed, with the first mode showing a robust increasing trend in temperature and a robust drying trend in precipitation in the region north of the Yangtze River; the second mode suggests a systematic decadal variability in temperature and precipitation. The underlying mechanisms for these leading modes are revealed through correlation and composite analysis. A negative Pacific-Japan teleconnection pattern in the lower troposphere and a stationary Rossby wave train across Eurasia in the upper troposphere are contributing factors to the trend mode. The trend component is closely tied to global warming, and also to increased sea surface temperature anomalies over the western Atlantic Ocean, which amplify the wave train teleconnection across Eurasia. The decadal variability is found to be associated with internal decadal change in interannual variability in the atmosphere, specifically, the decadal alternation of active oscillations above the North Atlantic and Pacific, which are likely related to the North Atlantic Oscillation (NAO) and Pacific Decadal Oscillation (PDO) modes, respectively. Further analysis illustrates that the NAO has a stronger impact during its active periods, while the PDO tends to have an opposite influence to the NAO when the NAO is inactive.

For realizing a weather-controlled society, we need to discuss the way to maximize the effect of manipulations to the atmosphere. For that purpose, this project aims at developing methods that quantify weather controllability and mitigatable flood damage based on ensemble weather forecasts. To quantify weather controllability, this project investigates meteorological landscapes that separate disaster and non-disaster regimes which may be controllable with small manipulations. We also estimate economic damages under non-controlled/controlled scenarios, in order to quantify avoidable damage by weather control. We have started illustrating directed graphs as the first step in understanding the meteorological landscape. Typhoon Prapiroon in 2018 was used for the case study. Singular value decomposition (SVD) is employed for Japan Meteorological Agency’s operational meso-scale ensemble prediction data to extract principle components of atmospheric states, followed by a clustering using density-based spatial clustering of applications with noise known as DBSCAN. The illustrated graph succeeded in detecting separated two clusters that correspond to faster and slower movements of predicted Parapiroon. The developed algorithm is currently applied to other disastrous events as well as further investigations on non-linear data compression methods beyond SVD. This presentation includes the most recent achievements up to the time of the conference.

Ensemble sensitivity has been proved to be a very useful sensitivity measure in practice. In this study, we show the relevance of ensemble sensitivity in another important problem. Instead of estimating how each state element will change a forecast response, we now examine which analysis perturbation among all possible ones with the same magnitudes will yield the largest change in the forecast response. We have proved that changes of the forecast response are maximized along the direction of the vector consisting of ensemble sensitivities which forms the most sensitive perturbation. This fact is found to be applicable in the field of weather controllability. In the simplest problem of weather controllability, we concern the optimal perturbation that minimizes a damage cost over a critical area. This optimization problem can be solved with variational data assimilation techniques by including the damage cost into the cost function. However, the problem becomes more complicated if we consider two important factors: (1) the power of optimal perturbation is limited by our technology so that it is impossible to realize the optimal perturbation in reality; and more importantly (2) we need to control the damage not just over an area but over several areas and it is possible that the damage only redirected from an area to another area under modification. We will demonstrate how the new understanding on ensemble sensitivities can qualitatively give potential solutions for such complicated issues.

Forecast sensitivity has been formulated to identify fast-growing perturbations in ensemble forecast. The ensemble singular vector (ESV), which doesn't require the tangent-linear and adjoint models, is derived in this study to obtain the critical initial condition that affects the ensemble forecast of high-impact events. While past researches mainly use this method in large-scale systems, this research examines the feasibility of ESV for long-lived heavy rainfall associated with synoptic-mesoscale fronts prediction. All of three cases show large rainfall forecast uncertainty in targeted area at validation time. The ESV is calculated using the metric of rainfall accumulation in the target area to evaluate the final forecast. To validate the evolution of ESV, the leading ESV is compared with the post forecast sensitivity defined by the difference between the members with high- and low-performance rainfall forecast in each case. In addition, the leading ESV is used to perturb the initial condition of the ensemble forecast to investigate the impact of ESV on rainfall forecasts. The preliminary result with a Mei-Yu front associated heavy rainfall event shows that the rainband can be effectively adjusted with the perturbed forecasts. Furthermore, , we confirm that the final ESV agrees with the evolution of the ESV-based initial perturbations under the non-linear model dynamic to a great extent. Finally, the effectiveness and challenges of applying ESV will be shown and discussed in this presentation. 

The national project in Japan, the Moonshot Research and Development Program sets ambitious goals, and Goal 8 is “Realization of a society safe from the threat of extreme winds and rains by controlling and modifying the weather by 2050”. It is expected to develop weather control technology that is technically, ethically, legally and socially feasible. Between 1962 and 1983, research in hurricane modification centered on an ambitious experimental program, Project STORMFURY in the USA. In Japan, research on weather modification became active under the influence of Project STORMFURY. In this study, we investigated the weather modification research in 1960s and 1970s in Japan and the USA by looking through all the references and interviewing people. The main purpose of the project STORMFURY was to conduct cloud-seeding experiments within hurricanes. Unfortunately, during the Project, the project team didn not take good opportunities and was only able to conduct cloud-seeding experiments three cases. To increase experimental opportunities, the USA government proposed to Japan and other countries over the northwestern Pacific Ocean. The possibility of experimenting in the Pacific was considered, but it did not materialize. In Japan, weather modification research had been active for 7 years, but rapidly declined in 1971. In this presentation, we will report the reasons why the experimental relocation did not materialize and why the weather modification research declined in Japan. The USA had discussed the experimental relocation with countries concerned, but the results were not satisfactory and the deadline for reporting to the Navy was not met. In Japan, declining of weather modification research can be attributed to the 1971 Kawasaki accident, 15 people died in landslide experiments on artificially slopes. The members leading weather modification research at the time were involved in that accident. This research was supported by JST Moonshot R&D Grant Number JPMJMS2282.

A project ”Moonshot Goal 8” supported by Japan Science and Technology Agency was established to study the possible weakening of typhoon intensity due to artificial interventions. One such measure is to increase the sea surface drag by using obstacles such as large ships. The maximum potential intensity theory suggests that the equilibrium intensity decreases as the surface drag coefficient increases if the surface enthalpy exchange is unaffected, but numerical studies to test it are limited. Previous fine-resolution simulations (e.g., with a sub-kilometer grid) tend to agree with the theoretical indication, but the number of cases is limited. Studies with coarse-resolution models exhibit mixed results. Also, no studies have been conducted to elucidate the effect of surface drag coefficient change in a limited oceanic region. Therefore, we aim to conduct a comprehensive study on how tropical cyclones would react to surface drag change over limited regions that can be set in various ways. As a first step, we conducted preliminary numerical simulations of Typhoon FAXAI in 2019 by changing the drag coefficient (C_D) over the whole simulation area. Here we ran the Scalable Computing for Advanced Library and Environment (SCALE) at a coarse resolution of 5 km. The resultant minimum central pressure was nearly insensitive to C_D, but the maximum winds were weakened by about 60 % of the control run (CTL) when C_D was set to 1.5 to 3.5 times that in CTL. Also, the radii of average winds of 15ms^-1 were decreased by about 20 %. We will conduct further studies until the meeting. This research was supported by JST Moonshot R&D Grant Number JPMJMS2282.

A project of Moonshot Goal 8 supported by Japan Science and Technology Agency is to study the possibility to reduce typhoon intensity by artificial interventions. Ocean thermal energy conversion (OTEC) which is planned to be established in the ocean near Okinawa islands is considered one of artificial interventions. The OTEC technology collects cold deep sea water at a depth of 800m and warm sea water at a depth of 20m, and discharges this mixed water to a depth of 20m. Namely an OTEC decreases the sea surface temperature (SST). When a typhoon passes over the ocean where the SSTs are decreased by OTEC, it is considered that the typhoon intensity is lower than without an OTEC. The purpose of this study is to conduct a numerical simulation of real typhoons using an atmospheric model with a SST distribution simulated by ocean model including the effects of OTECs (OTEC experiment), and to evaluate the impact of a SST forcing in the OTEC (sensitivity experiment). In the case of 2020, Typhoon HAISHEN passed through the OTEC influence zone, and the central pressures of HAISHEN were 930hPa by the Best Track data created by the Regional Specialized Meteorological Center Tokyo-Typhoon Center, 924.4hPa by our experiment without OTECs effect, and 924.6hPa by the OTEC experiment with the cooled SST (from -0.05 to -0.02℃) spread from each point of OTECs to the north for 50 to 70 km (OTEC influence zone). In the sensitivity experiment, the central pressures of Haishen increased by 0.57hPa and 1.00hPa at SST forcing with -0.5℃ and -1.0 in the OTEC influence zone, respectively. The sensitivity experiment of the OTEC or SST forcing that effectively affects the other cases of typhoons will be investigated. This research was supported by JST Moonshot R&D Grant Number JPMJMS2282.

The Airborne and Satellite Investigation of Asian Air Quality is a multi-perspective field study (aircraft, ground, satellite) planned to take place in early 2024. It is jointly sponsored by NIER (Korea) and NASA (United States) with flights over South Korea as a definite part of the plan and flights over a second partner country still under negotiation at the time of this abstract. The project will make an important contribution to science-based validation of the Geostationary Environment Monitoring Spectrometer (GEMS) which serves to provide an important catalyst for increased dialogue and cooperation among Asian countries to address air quality. In combination with satellite and ground observations, data will support analyses for assessment of emissions, model evaluation, process-level understanding of secondary pollutants (i.e., fine particles and ozone), and satellite validation and interpretation. Aircraft observations can provide invaluable context to the satellite and ground-based perspectives that are used more routinely to inform air quality models used for both forecasting and attribution. Important information from aircraft includes measuring detailed composition for source fingerprinting, vertical profiling of composition for satellite validation and model assessment, observing chemical and dynamical processes affecting secondary pollution (i.e., fine particles and ozone), relating specific VOC mixtures to satellite HCHO, providing fine scale pollution mapping with remote sensors, etc. Such information is critical for understanding the local factors influencing air quality for a specific location, quantifying emission sources, and assessing potential mitigation strategies for decision makers. This presentation will provide an important update on final decisions for deployment locations and planned observing strategies.

Hourly observations of air quality (AQ) over Asia have been available by the Geostationary Environment Monitoring Spectrometer (GEMS) for the first time from a geostationary Earth orbit (GEO) since its launch in February 2020. After 8-month in-orbit tests, GEMS has observed column amounts of atmospheric pollutants (O₃, NO₂, SO₂, HCHO, CHOCHO, and aerosols) to capture their diurnal variations with the UV–visible spectrometer at 0.6 nm spectral resolution and sophisticated retrieval algorithms. Details of the GEMS mission are presented, including calibrations, results, validations, and case studies including volcanic eruption, dusts, and urban pollution. L2 algorithms have been updated for version 2 and the products were released on November 30, 2022. In version 2, there are noticeable improvements in trace gases from updated AMF and the separation of stratospheric/tropospheric components. Ongoing calibration/validation activities including the 2022 GMAP/SIJAQ campaign and international CAL/VAL team works are critical to diagnose and improve the overall data quality. The GEMS retrievals indicate good agreements from the validation campaign, but still require further improvement in L1 processing. We start testing improvements for L1 processing including BTDF correction. Faster sampling rates at higher spatial resolution increase the probability of finding cloud-free pixels, leading to more observations of aerosols and trace gases than has been possible from LEO. GEMS will be joined by NASA’s Tropospheric Emissions: Monitoring of Pollution (TEMPO) this year and ESA’s Sentinel-4 to form a GEO AQ satellite constellation in late 2024, respectively, as recognized by the Committee on Earth Observation Satellites (CEOS).

Although polar-orbit satellite NO₂ Vertical Column Density (VCD) observations have been widely used for inverse modeling of the NO_x emission inventory, the relative low temporal resolution of polar-orbit satellite data restricts the improvement of the temporal resolution and accuracy of posterior emission inventory. Geostationary Environment Monitoring Spectrometer (GEMS) will provide NO₂ VCD observations with unprecedented spatial and temporal resolution, and it is critical to investigate their impacts on NO_x emission estimates. In this study, Observing System Simulation Experiments (OSSEs) are performed to quantify the improvements of using GEMS high spatiotemporal NO₂ VCD observations to constrain anthropogenic NO_x emission over China in relation to the widely used polar-orbit TROPOspheric Monitoring Instrument (TROPOMI) measurements. Our OSSEs use NO₂ data from WRF-Chem simulations as the “true” atmosphere and sample it with a configuration designed to represent TROPOMI and GEMS. These TROPOMI and GEMS NO₂ “observations” are assimilated through GEOS-Chem adjoint model with the perturbed prior anthropogenic NO_x emissions to obtain posterior monthly and daily emission inventories the corresponding diurnal NO_x emission profiles. The posterior results based on GEMS “observations” are in better agreement with the “true” anthropogenic NO_x emissions used in the WRF-Chem simulations than TROPOMI “observations”.

A novel method developed at the University of Iowa for retrieving aerosol optical depth at high resolution over coastal areas has been implemented using data from the NASA MODIS instrument. This algorithm leverages the spatial autocorrelation of atmospheric properties and the temporal autocorrelation of near-coastal areas to obtain high-resolution retrieved AOD even in areas with significant water-leaving radiance. This code uses MODIS top-of-atmosphere reflectance data and numerous ancillary data sources, and produces retrievals of 550nm aerosol optical depth at the native pixel resolution of MODIS (1km at nadir). This retrieval is applied specifically in coastal areas that are rejected for processing by other MODIS-based algorithms including the Dark Target retrieval. The retrieval code has been ported for near-real-time production by the Naval Research Laboratory. The data have been post-processed to provide the cleanest possible set of observations for data assimilation and incorporate estimated uncertainties required for use in data assimilation. This paper briefly summarizes the novel algorithm, and the skill evaluations that establish its usefulness for characterizing coastal aerosol. The steps taken for near-real-time production are described, including code optimization and linking of near-real-time ancillary data. Post-processing is described, including discussion of the tradeoffs between data coverage and data quality mediated by filtering and aggregation decisions. Estimation of per-observation uncertainties for data assimilation is described. Finally, results from including these observations in the initialization of aerosol forecasts using the Navy Aerosol Analysis and Prediction System are shown, including discussion of the regional, seasonal, and diurnal impacts of this new dataset. The potential for operational use of this product, and more broadly the potential of multi-sensor and hybrid products using both remote sensing and modeling inputs is discussed. 

This study developed an aerosol data assimilation system with the Weather Research and Forecasting model coupled with Chemistry (WRF-Chem) model and Ensemble Kalman Filter (EnKF) approach to improve PM2.5 forecast skill over East Asia. Unlike the variational assimilation method, the advantages of EnKF are flow-dependent background error covariance which is important in a fast-developing air quality system. In spite of considering flow-dependent BEC, the baseline run analysis exhibits poor performance, primarily due to small ensemble spread. This study conducted new two effective methods for increasing ensemble spread: one considering the uncertainty of model physics and the other considering the uncertainty in the prognostic variables. Both methods improved the quality of surface PM analysis substantially, compared with baseline run. And the DA_all experiment which incorporates both uncertainty in model physics and prognostic variables, demonstrates the best performance. The physical perturbation and multiplicative perturbation have a non-linear relationship. The forecast skill is also improved over South Korea and Jing-Jin-Ji area. With the substantial increase of BEC, the revised EnKF system has significantly improved the PM2.5 forecast skills.

Fengyun-3(FY-3) is the second generation of polar orbit series meteorological satellite built by China, with the capability of global observation. The Medium-Resolution Spectral Imager (MERSI) is one of the main loads of the FY-3 series satellites, belonging to the visible infrared band scanning imager. MERSI and MODIS belong to the same type of sensors. Most of the channel designs are similar and have the ability of aerosol retrieval. Based on the principle of MODIS dark target (DT) algorithm, we build a globally applicable land aerosol retrieval algorithm for the new generation of MERSI-II carried by the newly launched FY-3D to test its quantitative capability. Compared with the DT algorithm of MODIS, the algorithm has been improved mainly in two aspects: MERSI's own surface estimation model and inland-water mask method for haze extraction. We have carried out aerosol retrieval tests on global data in 2019 and 2020, and the validation results using the ground-based observation data of AERONET have reached a high overall accuracy close to MODIS aerosol product. We also used the DT algorithm to conduct aerosol retrieval for MERSI over ocean. The global retrieval and validation test in 2019 shows that the validation results have reached an acceptable accuracy. Recently, we plan to develop a new aerosol retrieval algorithm for the bright surface for MERSI, extending the DT algorithm to the bright target. The algorithm has been preliminarily tested in Northwest China as a research area. The results and validation of two years show that this algorithm can realize the retrieval of aerosol optical thickness over the bright surface by MERSI, and the validation accuracy of the retrieval results for the research area is slightly better than MODIS DB and MAIAC aerosol products. 

Atmospheric boundary layer structure plays a critical role in controlling the occurrence and evolution of extreme air pollution episodes. Our knowledge on the boundary layer structure of Nitrogen dioxides (NO2) is still limited. Satellite remote sensing technique measures vertical column density of NO₂ with an extensive spatial and temporal coverage. Recent developments in geostationary satellites (e.g., GEMS) offer new opportunities to observe air pollution with a high resolution of 1 hour. In this study, we applied the GEMS measurements together with various ground measurements to evaluate the variations of the scale height of NO2, which is an important indicator of the vertical structure of NO2. The monthly and diurnal variations of the scale height of NO2 were explored. In particular, the association between the scale height of NO2 and various meteorological factors (e.g., temperature, wind, humidity, air pressure) were evaluated. Our results show that the variations of the vertical structure of NO2 were highly related to the meteorological conditions. For instance, with the subsidence effect of tropical cyclone, the scale height of NO2 were extremely low because most NO2 were accumulated near the ground. We then used random forest regression to estimate the scale height of NO2 based on the meteorological values. Good results with a R2 value of >0.9 were obtained. These analyses enhance our understanding of the vertical structure of NO2 and the relationships between column and surface NO2.

This study examines the long-lived concentric eyewall structure of Typhoon Lekima (2019) from an axisymmetric perspective. Possible maintenance mechanisms for the concentric eyewalls are investigated using a high-resolution WRF simulation (nested down to 1-km horizontal grid size). The secondary-circulation responses to the latent heating in the inner eyewall, moat and outer eyewall are diagnosed by solving the Sawyer-Eliassen equation individually to examine the corresponding contribution to the moat downdraft. By calculating the dynamic efficiency factor (DEF), the conversion of latent heating to kinetic energy is evaluated in the moisture-restricted inner eyewall. The Sawyer-Eliassen diagnoses show that the moat downdraft was contributed mainly by latent heating in the inner and outer eyewall, with a secondary contribution from latent cooling in the moat after concentric eyewall formation. DEF diagnoses show that the conversion of latent heating to kinetic energy in the inner eyewall was more efficient than in the outer eyewall. Although tangential wind within the boundary layer was weakened by friction, the compensative tangential wind in the inner eyewall was larger than in the outer eyewall. The compensative tangential wind indirectly accumulated moisture from the sea surface in the moat, aiding the moisture supply to the inner eyewall and enhancing the amount of kinetic energy converted from latent heating. Although the inner eyewall of Typhoon Lekima eventually weakens due to the moisture cut off from the outer eyewall, the inner eyewall can still be maintained for tens of hours by the high DEF from latent heating.

Idealized model simulations have shown that cloud radiative effect (CRE) can accelerate early development of tropical cyclones (TCs). Deep convective region near TC inner cores receives more energy than surrounding environment, and this spatial radiative heating difference can further drive a secondary circulation, transporting angular momentum and water vapor into TC center, thus favoring the TC intensification. However, the quantitative contribution of radiation to TC genesis in observation has not been well documented and need further investigation. In this study, tropical cloud clusters are identified according to their brightness temperature, then classified into developing and non-developing systems based on IBTrACS data. Satellite observations from CloudSat are used to composite vertical structure of radiative heating and atmospheric cloud radiative effect (ACRE), and to verify the differences between these two groups. Preliminary results show that, for developing systems, the upper troposphere absorbs more shortwave but lose more longwave from clouds. While at mid-level, more longwave energy is trapped but less shortwave energy can be transmitted through deep clouds to these layers. Overall, shortwave ACRE drives anomalous upper-level upward motion, and longwave ACRE promotes mid-level upward motion. Both of these effects enhance the secondary circulation between the convective region and their environment. As for non-developing systems, the general characteristics are similar to those in developing systems, except for smaller deep cloud extent and less radiative heating in surrounding area, implying that the deep convection structure, the cloud canopy extent, the associated cloud radiative effect, and the responded vertical motion could be related to the cyclogenesis process. More quantitative analyses, such as Sawyer-Eliassen equation, are still undergoing to further evaluate how secondary circulation and TC genesis process would respond to the radiative heating difference between developing and non-developing systems.

Super Typhoon Hinnamnor (2022) was rare and unique in the record over the western North Pacific as throughout its lifespan it featured all of the major frontier issues in typhoon research currently. Specifically, in different stages of its lifespan, it had a sudden change of track, underwent rapid intensification, interacted and merged with another vortex, expanded in size, underwent rapid weakening and had a strong cold wake, eyewall replacement, and extratropical transition. Therefore, a timely identification and review of these typical features of Hinnamnor (2022), as reported in this article, will help update and enrich the case sets for each of these scientific issues and provide a background for more in-depth mechanistic studies of typhoon track, intensity, and structural changes in the future. We believe that Hinnamnor (2022) can serve as an excellent benchmark to quickly evaluate the overall performance of different typhoon numerical models in predicting its track, intensity, and structural changes.

One of the most prominent diurnal cycle features in tropical cyclones (TCs) is the radially outward propagation of a cooling signal in the upper-level clouds, the so-called diurnal pulse (DP). Previous studies suggested that some of the DPs occur in the deeper convective layer and therefore may impact the TC structure and intensity. This study investigates how the internal structures and intensity will change associated with DPs in global TCs using 18 years of multi-source satellite observations. Over 3000 DP events are identified based on objective method using satellite infrared data. Satellite microwave observations and spaceborne radar measurements are used to examine the changes of the TC convective structures related to DP events. Results show that the precipitation, microwave ice scattering, radar-echo top height, and lightning in the TC inner core are all markedly enhanced on DP days compared to non-DP days. Spaceborne radar observations further indicate that convection becomes deeper in the upshear quadrants with the presence of DPs and convective depths remain similar in the downshear quadrants. These changes in the TC internal structures are consistent to changes in rapidly intensifying TCs reported in the literature. In fact, this study shows that rapidly intensifying TCs have a higher frequency of the long-duration DP event and significantly longer DP duration than steady-state and gradually intensifying TCs. In short, DP implies a coherent pattern of the convective structure in the TC’s evolution and provides important implications for TC intensity change.

The prediction of tropical cyclone (TC) intensity remains a major scientific challenge. Recent studies indicate that cloud-radiation feedback (CRF) plays a positive role in the intensification of TCs during their genesis. However, little attention has been given to how CRF affects TC intensity after genesis. This study shows that CRF may prevents TCs from attaining higher maximum intensities. The ascending motion induced by the anomalous radiative heating of TC promotes more latent heating on the outer side of the upper eyewall, resulting in a more tilted eyewall. A more tilted eyewall leads to a larger inner-core size and less inward flux of absolute vertical vorticity within the inner core, thus preventing the TC from reaching higher intensity. This work highlights that CRF may affect TC intensity by modulating the structure of the inner-core convection, and further advances our understanding of the interaction between radiation effect and TC dynamics

Toward the understanding of rapid intensification (RI) of tropical cyclones (TCs) in the western North Pacific, the TC's deep convective cloud (DCC), precipitation, and cloud properties in terms of cloud effective radius, optical thickness, and top height from satellite observations are investigated. Mean and radial distributions of the variables at different intensity stages and intensification categories are examined. The relationship indicates that the DCC percentage and temperature, especially their radial distributions, could be used to identify an impending RI regardless of TC intensity. Meanwhile, the mean and radial distribution of precipitation may discriminate RI from non-RI in tropical depression (TD) and tropical storm (TS). The radial distribution of the cloud properties in rapidly intensifying TD and TS also suggest that most of the clouds near the center of the storm has deepened already while those that are far from the center are generally in developing or dissipating stage. Moreover, rapidly intensifying TCs, regardless of their intensities, manifest common DCC, precipitation, and cloud properties characteristics near the TC center. It is to be noted that the different mean and radial distribution characteristics of the variables between initial and continuing stages of RI are inferred to be artifacts of their intensities and RI rates (or radius of maximum wind sizes) rather than whether the TCs are at the onset or 24 hr of RI.

The Western North Pacific basin has large uncertainty in future tropical cyclone (TC) frequency projection across climate models, but the underlying reason has been unclear. We show that uncertainty in TC frequency projection is highly correlated with uncertainty in the response of clouds to sea surface warming. Based on the previously developed seed propensity index and gross moist stability theory, we hypothesize that the pattern of cloud response influences TC frequency through its radiative effect on the atmospheric column energy budget. Cloud radiative heating over the Western North Pacific drives anomalous vertical ascent in the large-scale circulation, generating more precursory vortices and seeding more TCs. This hypothesis is supported by numerical experiments with perturbed radiative heating rates and sea surface temperature using global atmospheric models developed at the Geophysical Fluid Dynamics Laboratory. We show that uncertainty in Western North Pacific TC response is reduced when controlling for the radiative heating rate. We further show that cloud and TC response in the Western North Pacific has global impact on the top-of-atmosphere energy budget, using simulations with localized sea surface temperature perturbations. The results suggest that Western North Pacific TC frequency tends to decrease in a scenario with more positive cloud feedback; while the frequency tends to increase in a scenario with more negative cloud feedback.

Emissions from biomass burnings, cause large impacts on our society and environment and cause local to global scale issues. The properties of smoke particles from these emissions rapidly evolve at seconds to minutes scales. Being able to accurately measure the smoke absorption properties is essential when quantifying the aerosol impacts on radiation budgets and mesoscale dynamics and help improving the forecast as well as facilitate risk management. In-situ measurements can provide accurate measurements of aerosol properties but have limited sample size. From satellites, which have much wider temporal and spatial coverage, the fast-changing absorbing properties of smoke plumes are very challenging to retrieve. Utilizing the very fine temporal resolution observations from geostationary satellites and the powerful critical reflectance (CR) methods, we now have a pathway to retrieve wavelength dependent absorption at 10 minutes intervals. CR is a surface reflectance at which increasing or decreasing aerosol loading does not change the TOA reflectance. This reflectance value is highly associated with the absorption of the aerosol layer and not related to lower boundary conditions, which makes it suitable to apply over heterogeneous land surfaces. Using Advanced Base Imager onboard GOES-17, we demonstrated the power of CR technique by retrieving aerosol single scattering albedo on two fire events, following the track of freshly emitted smoke plumes, and compared with nearby ground-based observations. The uncertainty of the CR method is also tested using sensitivity studies. This method will be applied to all suitable fire events and the temporal and spatial variation of retrieved SSA for each case will be analyzed and compared to other aerosol absorption products. 

In the context of climate change, fire danger conditions are expected to increase in many regions of the world due to the projected changes in climate as the occurrences of biomass burning are events highly dependent on the meteorological driver. The threat from biomass haze in Malaysia and Indonesia will not only be about frequency but also intensity. There is a possibility that biomass burning will become more severe in the future. However, while the study of the near-real-time predictions of burning hotspots in Malaysia and Indonesia has been developed widely, the study of its projection has not been assessed. This study aims to project burning hotspots that are represented by FWI (Fire Weather Index), a fire danger system that relies upon the daily weather readings during the previous day, including temperature, relative humidity, wind speed, and rain. The projection is performed based on the future climate condition (2041 – 2070) under RCP4.5 and RCP8.5 scenarios using the downscaled simulations of the Southeast Asia Regional Climate Downscaling/Coordinated Regional Climate Downscaling Experiment—Southeast Asia (SEACLID/CORDEX-SEA). The result of the projection of trends of the fire hotspots by applying the FWI shows an overall significant increase in fire activities under the RCP4.5 and RCP8.5 scenarios in the future. The average FWI reached 9.6 and 11.1 under RCP4.5 and RCP8.5 scenarios, respectively. The average percentage increase of FWI in the future under the RCP8.5 scenario is 43.9%, higher than under the RCP4.5 scenario, which is 39.7%. In short, the result of this study provides an understanding of how fire events condition in the future as the efforts to control the forest and land fire disasters which affect the biomass burning haze.

Research has shown that short term events, such as forest-fires, agricultural-burning, dust-storms, fireworks, etc., significantly contribute to air-pollution, and impact the climate, air-quality, visibility, human-health, etc. Festivals that are marked by firework displays and wood-log fires are among these transient events that severely alter the local air-quality. In India, two major festivals, Diwali and Holi, are celebrated throughout the country, and are dominated by fireworks and wood-log/ biomass burning, respectively. As per the religious belief, Holi is preceded by Holikadahan, when wood-log/ biomass is burned in the last evening. The burning results in the emission of huge amounts of gaseous and particulate pollutants. Innumerable point sources Holikadahan in the vicinity make Holikadahan a huge polluter on a regional basis.The study, spanning six years (2017-2022), was conducted using data collected from 12 cities, extending from west to east in the Indo-Gangetic plains (IGP) to examine the impact of Holikadahan by comparing the period of background levels, pre-Holikadahan, and Holikadahan of PM2.5, PM10 and aerosol optical depth (AOD). As a general observation, the highest spike in pollutant concentrations, including PM2.5, PM10 were observed during the Holikadahan and immediately after the Holikadahan, followed by a decreasing trend, back to background levels, typically after 24-hours.In the western IGP, the maximum impact of Holikadahan on air-quality was observed in 2021, showing 71%, 63% and 82% increase in PM2.5, Pm10, and AOD values, respectively, between pre-Holika and Holika. Similar increases in central IGP, observed in 2021, were 80%, 25% and 55%, respectively. In the eastern IGP, however, the maximum impact was observed in 2020, with 74% and 75% increase in PM2.5 and PM10 concentration, respectively. This increase in primary data is expected to change the localized air-quality index (AQI), aerosol radiative forcing efficiency, and atmospheric heating-rate.

Underpinned by large-scale air-sea coupled dynamics in the equatorial Pacific, the El Niño-Southern Oscillation (ENSO) makes it the strongest natural interannual variability on Earth with far-reaching socio-economical influences across the globe. Compared to its extensively studied changes in warming scenarios, ENSO responses to anthropogenic forcing removal or reduction, however, remain less explored so far. Based on the 1% per year CO₂ changing for ramp-up and ramp-down experimental scenarios and climate models that capture key ENSO dynamics, here we show two major ENSO sea surface temperature (SST) statistics of variance and skewness exhibit prominent hysteresis responses in a CO₂ changing pathway. In contrast to largely uncertain responses in the ramp-up period, eastern Pacific SST variance and central Pacific SST skewness during the ramp-down period display exaggerated increases and decreases changes, respectively. Such ENSO hysteresis owing to corresponding SST feedback changes is closely associated with a continuously strengthening and eastward-propagating of the El Niño action center until the middle of the ramp-down period. The El Niño action center changes are further linked with similar hysteresis background changes manifested in the intertropical convergence zone, which increase the eastern Pacific atmosphere sensitivity to local SST anomalies during the El Niño initial stage. We also suggest that the presence of El Niño hysteresis would possibly exacerbate its amplified consequential global impacts in a warming world.

 In this study, a comprehensive analysis is conducted to explore the response of global precipitation extremes to CO₂ in terms of hysteresis and reversibility effect and associated population exposure. In this regard, climate outputs under two idealized CO₂ scenarios such as ramp-up (RU; about +1% annually until quadrupling of present level) and ramp-down (RD; around -1% annually set back to present level) from Community Earth System Model version 1.2, and the projected population data from the five shared Socioeconomic Pathways (SSPs) are used. Precipitation extremes are evaluated using the number of heavy precipitation days, maximum consecutive 5-day precipitation, and the precipitation of very wet days indices. Results show that the magnitude of extreme precipitation change and associated population exposure is higher in the CO₂ reduction period (RD) than in RU. All the indices show substantial irreversible and hysteresis effects, ~69% of the global land is expected to experience irreversible changes in precipitation extreme. Further, the hotspots of irreversibility (the region with irreversible change and a large hysteresis) will emerge in >20% of the global area. Spatially, strong hysteresis and irreversibility are particularly concentrated over global land monsoon regions. The leading exposure is estimated under SSP3 combined with both RU and RD periods. Under the SSP3-RD combination, the highest population exposure is estimated at ~67.1% (globally averaged), and ~72% (averaged over hotspots) higher than that of the present day. The exposed population is prominent in South Africa and Asia. Notably, the population change effect is the principal factor in global exposure change, while it is the climate change effect over the hotspots of irreversibility. These findings provide new insight into policymaking that only CO₂ mitigation effort is not enough to cope with precipitation extremes, rather advanced adaptation planning is a must to have more socio-economic benefits.

For investigation in the change of global temperature after the cessation of CO₂ emissions, Zero Emissions Commitment (ZEC) scenario has been proposed. In our four ensemble members of CESM2 ZEC experiment, CO₂ emissions are linearly increased and symmetrically decreased till zero, then this net zero condition is maintained over longer than 100 years. During this net zero emission period, authors found that the ensemble spread of global mean surface temperature (GMST) increases drastically and the ensemble mean also increases again, without increasing CO₂ emissions. The ensemble spread, especially over the northern or subpolar Atlantic Ocean, is associated with differences in the reconstruction of Atlantic Meridional Overturning Circulation (AMOC) within ensembles. This could be explained by AMOC–Salt–Advection positive feedback that a stronger surface salinity derives a weaker stratification of the northern Atlantic Ocean and a stronger AMOC. This raises the stronger meridional warm and salt advection in faster warming ensemble members than slower warming members during the net zero emission period. On the other hand, the positive trend of surface temperature over Southern Ocean (SO) persists during the entire ZEC scenario, thus this plays an important role in the increasing ensemble mean GMST during the net zero period. In contrast, it is noteworthy that the ensemble spread is not as large as the northern Atlantic Ocean or Arctic regions. Authors further found that this positive trend over SO is associated with changes of surface heat budget by the intensified downwelling radiation. From these results, present study argues that the roles of both Northern and Southern hemisphere in ZEC scenario is distinguished in the ensemble spread and mean of GMST, respectively.

When projecting future monsoon changes by carbon dioxide (CO₂) pathway, most studies have analyzed precipitation responses without considering monsoon area (MA) variations. Further, how MA responds to CO₂ removal remains uncertain. This study evaluates MA variations and impacts in idealized CO₂ ramp-up (toward CO₂ quadrupling), ramp-down, and stabilized simulations using the Community Earth System Model version 1. Global MA negatively overshoots (i.e., recovery with decreasing tendency beyond the original MA) during the ramp-down period due to reduced or rapidly recovered MA in several regional monsoons, including Northern and Southern Africa, South and East Asia, and South America, showing hysteresis when comparing ramp-up and -down periods despite similar global warming levels. These non-linear regional MA variations come from distinct regional summer and winter precipitation variations, which are found to be associated with Intertropical Convergence Zone movements and El Niño-like response. Further, regional monsoon precipitation characteristics also vary through ramp-up and ramp-down periods consistently with overall hysteresis. Changes in total monsoon precipitation resemble the distinct responses of MA. Our results suggest that regions characterized by a monsoonal climate may experience reduced seasonal rainfall variations under net-negative CO₂ emissions.

Understanding the regional hydrological response to varying CO₂ concentration is critical for cost-benefit analysis of mitigation and adaptation polices in the near future. To characterize summer monsoon rainfall change in East Asia due to a change in the CO₂ pathway, we used the Community Earth System Model (CESM) with 28 ensemble members in which the CO₂ concentration increases at a rate of 1% per year until its quadrupling peak, i.e., 1,468 ppm (ramp-up period), followed by a decrease of 1% per year until the present-day climate conditions, i.e., 367 ppm (ramp-down period). Although the CO₂ concentration change is symmetric in time, the rainfall response is not symmetric. The amount of summer rainfall anomaly in East Asia is increased 42% during a ramp-down period than that during a ramp-up period when the two periods of the same CO₂ concentration are compared. This asymmetrical rainfall response is mainly due to an enhanced El Niño-like warming pattern as well as its associated increase in the sea surface temperature in the western North Pacific during a ramp-down period. These sea surface temperature patterns enhance the atmospheric teleconnections to East Asia and the local meridional circulations around East Asia, resulting in more rainfall over East Asia during the ramp-down period. This result implies that the removal of CO₂ does not guarantee the return of regional rainfall to the previous climate state with the same CO₂ concentration.

A poleward shift of the Hadley cell (HC) edge in a warming climate has been widely documented. However, its possible change to CO₂ removal has not been explored. By conducting large ensemble experiments where CO₂concentrations are systematically increased and then decreased to the present-day level, we show that the poleward-shifted HC edges in a warming climate do not return to the present-day state when CO₂ concentrations are reduced to the present-day climate. While the Southern-Hemisphere HC edge remains poleward of its present-day state, the Northern-Hemisphere HC edge returns and remains farther equatorward. Such hemispherically asymmetric HC-edge changes, which contribute to drought-prone subtropical regions, are closely associated with the changes in the vertical wind shear. We attribute the vertical wind shear change to the hysteresis in the oceanic response to a changing CO₂ pathway.

The reversibility of South Asian summer monsoon (SASM) precipitation under the CO₂ removal scenario is critical for climate mitigation and adaptation. In the idealized CO₂ ramp-up (from 284.7 to 1138.8 ppm) and symmetric ramp-down experiments, SASM precipitation is largely reversible while exhibiting strong asymmetry: it may overshoot the unperturbed level when CO₂ recovers. Such asymmetric response is mainly due to the enhanced El Niño-like and Indian Ocean dipole-like warming during the ramp-down period. The uneven sea surface warming weakens Walker circulation, with anomalous sinking over the SASM region; meanwhile, the warming also affects the rainfall over the Maritime Continent and tropical western Indian Ocean. The suppressed rainfall over the Maritime Continent triggers the equatorial Rossby wave, which weakens the ascent over the SASM region. The increased rainfall over the tropical western Indian Ocean excites the equatorial Kelvin wave, which reduces moisture transport. Additionally, tropic-wide warming reduces the land-sea thermal contrast and weakens monsoonal circulation. Consequently, the combined effects of the weakened ascent and moisture transport lead to the overshooting of SASM rainfall. Our results suggest that symmetric CO₂ removal, although unlikely in the foreseeable future, may result in a risk of local drought over the SASM region.

This study aims to clarify the spatiotemporal variation, chemical fingerprint, transport routes, and source apportionment of marine fine particles (PM_2.5) at a remote island in the South China Sea (SCS). Field sampling and receptor modeling of PM_2.5 were conducted from September, 2020 to July, 2023. Twenty-four hour marine PM_2.5 was collected at the Dongsha Islands for continuous seven days in each season. Chemical composition of PM_2.5 was analyzed for water-soluble ions (WSIs), metallic content, carbonaceous content, anhydrosugars, and organic acids. Moreover, the potential sources of PM_2.5 and their contribution were further resolved by backward trajectory simulation and chemical mass balance (CMB) receptor model. Three-year field sampling results indicated that high concentrations of PM_2.5 were commonly observed in winter and spring. In terms of chemical composition of PM_2.5, secondary inorganic aerosols (SIAs; SO₄^2-, NO₃^-, and NH₄⁺) accounted for 53.0-64.1% of WSIs which dominated PM_2.5. Crustal elements (Ca, K, Mg, Fe, and Al) dominated metallic content of PM_2.5, while trace elements (V, Ni, and Cu) were originated from anthropogenic sources. The concentrations of organic carbon (OC) in PM_2.5 were generally higher than those of elemental carbon (EC). The highest concentrations of levoglucosan were 20.94 ng/m³ at the Dongsha Islands in winter, while galactosan and mannosan were non-detectable (MDL). Major sources of PM_2.5 resolved by CMB receptor modeling were sea salts (16.6-24.7%), fugitive dust (15.8-23.0%), industrial boilers (6.6-12.0%), secondary sulfate (7-10.9%), mobile sources (4.7-9.6%), waste incinerators (2.4-5.2%), and biomass burning (2.8-5.3%). Keywords: marine fine particles, chemical fingerprints, spatiotemporal variation, transport routes, source apportionment.

Multiphase chemistry of chlorine is coupled into a 3D regional air quality model (CMAQv5.0.1) to investigate the impacts on the atmospheric oxidation capacity, ozone (O3), as well as fine particulate matter (PM2.5) and its major components over the Yangtze River Delta (YRD) region. By including the chlorine chemistry, the model performances in simulating hydrochloric acid (HCl), particulate chloride (PCl), and hydroxyl (OH) and hydroperoxyl (HO2) radicals are significantly improved. O3 is enhanced in the high chlorine emission regions by up to 4% and depleted in the rest of the region. PM2.5 is enhanced by 2~6%, mostly due to the increases in PCl, ammonium, organic aerosols, and sulfate. Nitrate exhibits inhomogeneous variations, by up to 8% increase in Shanghai and 2~5% decrease in most of the domain. Radicals show different responses to the chlorine chemistry during the daytime and nighttime. Both OH and HO2 are increased throughout the day, while nitrate radicals (NO3) and organic peroxy radicals (RO2) show an opposite trend during the daytime and nighttime. Higher HCl and PCl emissions can further enhance the atmospheric oxidation capacity, O3, and PM2.5 so that the anthropogenic chlorine emission inventory must be carefully evaluated and constrained.

Tropospheric ozone is an important greenhouse gas and is central to atmospheric oxidation capacity. We examine observed tropospheric ozone trends, their attributions, and radiative impacts from 1995–2017 using aircraft observations from the In-Service Aircraft for a Global Observing System database (IAGOS), ozonesondes, and a multi-decadal GEOS-Chem chemical model simulation. IAGOS observations above 11 regions in the Northern Hemisphere and 19 of 27 global ozonesonde sites have measured increases in tropospheric ozone (950-250hPa) by 2.7 ± 1.7 and 1.9 ± 1.7 ppbv decade^-1 on average, respectively, with particularly large increases in the lower troposphere (950-800 hPa) above East Asia, Persian Gulf, India, northern South America, Gulf of Guinea, and Malaysia/Indonesia by 2.8 to 10.6 ppbv decade^-1. The GEOS-Chem simulation, driven by reanalysis meteorological fields and the most up-to-date year-specific anthropogenic emission inventory, reproduces the overall pattern of observed tropospheric ozone trends with an increasing trend of 0.4 Tg year^-1 of the tropospheric ozone burden in 1995–2017. Sensitivity simulations show that changes in global anthropogenic emission patterns, including the equatorward redistribution of surface emissions and the rapid increases in aircraft emissions, are the dominant factors contributing to tropospheric ozone trends by 0.5 Tg year^-1. In particular, we highlight the disproportionately large contribution of aircraft which emitting NO_x in the mid- and upper troposphere where ozone production efficiency is high. Regionally, the Chinese nationwide ozone monitoring network has observed rapid increases in warm-season surface ozone in cities during 2013−2019 by about 2.4 ppbv per year, among the fastest urban ozone trends in the recent decade reported in the Tropospheric Ozone Assessment Report (TOAR). Our model simulations reveal that the presence of soil NOx emissions in the North China Plain, mainly driven by agricultural fertilizer applications, present as an underappreciated challenge for ozone mitigation there.

Surface ozone (O₃) is a major air pollutant and draw increasing attention in the Pearl River Delta (PRD), China. Here, we characterize the spatial-temporal variability of ozone based on a dataset obtained from 57 national monitoring sites during 2013-2019. Our results show that: (1) the seasonal difference of ozone distribution in the inland and coastal areas was significant, which was largely affected by the wind pattern reversals related to the East Asian monsoon, and local ozone production and destruction; (2) the daily maximum 8hr average (MDA8 O₃) showed an overall upward trend by 1.11 ppbv/year. While the trends in the nine cities varied differently by ranging from -0.12 to 2.51 ppbv/year. The hot spots of ozone were spreading to southwestern areas from the central areas since 2016. And ozone is becoming a year-round air pollution problem with the pollution season extending to winter and spring in PRD region. (3) At the central and southwestern PRD cities, the percentage of exceedance days from the continuous type (defined as ≥ 3 days) was increasing. Furthermore, the ozone concentration of continuous type was much higher than that of scattered exceedance type (< 3 days). In addition, although the occurrence of continuous type starts to decline since 2017, the total number of exceedance days during the continuous type is increasing. These results indicate that it is more difficult to eliminate the continuous exceedance than the scatter pollution days and highlight the great challenge in mitigation of O₃ pollution in these cities.

A Source- and Age-Resolved Algorithm (SARA) was developed in the CMAQ model and applied to study the sources and ages of primary fine particulate matter (PPM) and secondary inorganic aerosols (SNA) in China during January 2013. Residential and industry are the major contributors to PPM and sulfate, especially for fresh particles. The contributions of power increase with age, accounting for 25% and 51% of aged sulfate and nitrate (> 48 h), respectively. Local emissions and intra-regional transport with younger age contribute to over 70% of PPM and SNA in Beijing, while long-range transport from northern China with older age becomes more significant in Shanghai. On pollution days, the ages of PPM and SNA consistently increase, suggesting enhanced contributions from regional transport. This study highlights that the age-specific particle source and region information produced by our SARA algorithm can help design cost-effective emission control strategies to reduce extreme haze pollution.

Fine plastic particles (particle diameter < 2.5 μm, hereafter FPP) have made considerable contributions to human health, aerosol pollution, climate effects, and many aspects of the ecosystem. However, the direct characteristics and high-time resolution (hourly) quantitative detection of atmospheric FPP were limited. We investigated the composition and probable sources of FPP in PM_2.5 in this study. Dual-hour-resolution was obtained for the quantitative study of FPP in PM_2.5 using the Versatile Aerosol Concentration Enrichment System (VACES) and Thermal Desorption/Pyrolysis-Gas Chromatography-Mass Spectrometry (TD/Py-GC-MS). The FPP in PM_2.5 accounted for an expected value of 5.57 μg/m³ (from 0 to 24.73 μg/m³) while a ratio of FPP to PM_2.5 was 13.2 % (from 4.1 to 42.6 %) in most PM_2.5 samples during the campaign. FPP are not significantly correlated with PM_2.5 yet accounted for in PM_2.5 concentrations, and PAEs are not part of atmospheric particulate matter yet are highly correlated with PM_2.5. Additionally, a comprehensive chemical analysis of the FPP in PM_2.5 has revealed the abundant presence of plastic particles, as opposed to other particles such as soot and tar balls. Heavy pollution of FPP was closely related to local human activities. This work implied that FPP in the air are a growing threat and atmospheric stocks of FPP were strongly related to human activity.

Increasing air pollution levels and associated detrimental health and environmental impacts is a rising concern for citizens of India. National policies and action plans have been sporadically implemented by the government along with various mitigation pathways suggested for further control of air pollution. To effectively combat air pollution, a detailed emission inventory is vital taking into account every sector. The Indian food service industry is one sector that is neglected. The food industry in India is a growing sector, and the unexpected rise due to the high demand in the service delivery market indicates the expected increase in cooking oil fumes. The cooking oil fumes produced as a result of high temperature processes are a major health hazards to human. However, there are limited studies addressing the issue and lack of data related to the sector for construction of a national emission inventory for the food service industry. To develop an inventory for India, firstly the structure of food/restaurant industries in the capital region of the country, megacity Delhi was studied by conducting a survey campaign. Based on the data collected, the emissions from the sector were calculated for Delhi. This study helps understand the pollutant emissions from the food service industry for megacity Delhi which is the hub for the largest number of restaurants in the country. This information can further be extrapolated across all states of India to help construct national emission inventory in future studies.

In the summer (July and August) of 2022, unprecedented heat wave occurred along the Yangtze River Valley (YRV) over East Asia while unprecedented flood occurred over western South Asia (WSA), which are located on the eastern and western sides of Tibetan Plateau (TP). Here, by analyzing the interannual variability based on observational and reanalysis data, we show evidences that the anomalous zonal flow over subtropical Tibetan Plateau (TP) explains a major fraction the extreme events occurred in 2022. As isentropic surfaces incline eastward (westward) with altitude on the eastern (western) side of the warm center over TP in summer, anomalous easterly (westerly) flow in upper troposphere generates anomalous descent (ascent) on the eastern side of TP and anomalous ascent (descent) on the western side of TP via isentropic gliding. The anomalous easterly flow is extremely strong to reverse the climatological westerly flow over subtropical TP in 1994, 2006, 2013 and 2022. The easterly flow in 2022 is the strongest since 1979, and it generates unprecedented descent (ascent) anomaly on the eastern (western) side of TP, leading to extreme heat wave over YRV and extreme flood over WSA in 2022. The anomalously strong easterly flow over subtropical TP in 2022 is dominated by atmospheric internal variability related to mid-latitude wave train, while the cold sea surface temperature anomaly over the tropical Indian Ocean increases the probability of a reversed zonal flow over TP by reducing the meridional gradient of tropospheric temperature.

Characteristics of extreme precipitation over Yellow River Loop Valley (YRLV) and links to European blocking are investigated in this study. Spatial and temporal analysis of extreme precipitation shows that it contributes more than 30% of the total summer precipitation in the YRLV and is characterized by a strong and short period of local rainfall. Most of the extreme rains in the YRLV occur in July and August. Two typical circulation patterns were identified using a k-means clustering method. The extreme precipitation results from the combined actions of intensified high pressure over northeast China (NECH) and the westward extension of the western Pacific subtropical high (WPSH). The intensified southerly flow of the amplified NECH strengthens the water vapor transport induced by the westward extension of the WPSH from the northwest Pacific or Bay of Bengal into the YRLV. The NECH is amplified by the wave energy propagating from European blocking via the Silk Road pattern (SRP). This is the subseasonal cause of extreme precipitation over the YRLV. The composited July and August mean 500 hPa geopotential anomaly pattern for extreme precipitation years shows a high-pressure anomaly over the European continent and a negative phase of the SRP. The former provides a background for the occurrence of European blocking, and the latter explains the preexistence of the NECH and provides a linkage between the activity of European blocking and the subseasonal evolution of the NECH. Thus, the interannual variation in the extreme precipitation over the YRLV is mainly reflected by the phase of the SRP and the stationary waves over Europe.

Southern China spring rainfall (SCSR) is significant for agricultural sowing and soil moisture accumulation before the rainy summer. A better prediction of the rainfall improves our ability to risk response to natural disasters. It is found that the SCSR can be promoted by multi-year El Niño events through the high-latitude pathway (HP) and low-latitude pathway (LP). The long-lasting El Niño warming heats the tropical troposphere persistently until the decaying spring, which strengthens the Arctic polar vortex and the mid-latitude blockings. This HP is in favor of more southward transport of Rossby wave energy and cold air, resulting in strong ascending motions over southern China in spring. The multi-year El Niño also induces an enhanced western North Pacific anticyclone and a secondary circulation transporting moisture to southern China through the LP. The HP is more important in the early spring, while the LP dominates the heavy SCSR in the late spring.

Investigation into the interannual variation of the Middle East jet stream (MEJS) and its influence on the Asian monsoon indicates that the eastward extension of MEJS is closely related with a wetter and colder winter in southern China and a later onset of the subsequent Asian summer monsoon, compared with normal conditions. When the MEJS extends eastward, a significant barotropic anomalous anticyclone is located over the Arabian Sea (AS), associated with the southeastward propagating wave train from Europe. Intense divergence in the southwest of the AS anomalous anticyclone favors more convection over the western tropical Indian Ocean, which excites an anomalous upper-level anticyclone to the north as a Rossby wave response, further intensifying the AS anticyclonic anomaly. This positive feedback loop maintains the AS anomalous anticyclone and results in the eastward extension of the MEJS. Accordingly, intense northeasterly anomalies over the Mediterranean Sea and the subtropical westerly anomalies bring abundant cold air from the middle-higher latitudes to subtropical regions, resulting in a widespread cooling in subtropical Eurasia including southern China. Barotropic anomalous westerlies occur around the Tibetan Plateau in the south and deepen the India-Burma trough, favoring more water vapor transport from the Bay of Bengal to southern China. This wetter and colder conditions in subtropical Eurasia can persist from winter to spring, leading to the much later onset of the Asian summer monsoon. Therefore, the winter MEJS variability can be considered as an important indicator for the Asian monsoon.

The towns of Cherrapunji and Mawsynram, located on the southern slopes of the Meghalaya Plateau in the northeastern part of the Indian subcontinent, are known to have the heaviest rainfall in the world. The Indian Meteorological Department announced that 972 mm of rain was recorded in the 24 hours ending 03 UTC on 17 June 2022. This was the third rain in the last 122 years. Severe flash floods occurred downstream in Sylhet, Bangladesh. Such case studies of heavy rain events must be useful in understanding the mechanism of heavy rain and in considering countermeasures against flash floods. An automatic weather station installed at Cherrapunji and an optical disdrometer installed at Sylhet recorded rain for the heavy rain event. The rainfall for the three days from the 14th to the 16th was 2412 mm, and it was 43% of the monthly rainfall in June 2022. Precipitation showed a distinct diurnal variation and increases from midnight to early morning. On the other hand, in Sylhet, which is located in the plains 40 km southeast of Cherrapunji, the precipitation variability was completely different and the diurnal variation of precipitation was unclear. The precipitation system was characterized by the images of Bangladesh Air Force (BAF) C-band radar at Jessore, Bangladesh, and the Global Satellite Mapping of Precipitation (GSMaP) product. The precipitation system was a topographical system localized in southern Meghalaya. The precipitation system on the 15th was not sufficiently detected by the BAF radar, suggesting that the echo top height was relatively low. While the GSMaP showed some agreement with the BAF radar, on the 14th it was not possible to properly detect the onset of the nocturnal rain around 12UTC. This was because the data from the satellite-borne microwave radiometer observations used to create the GSMaP were not available within 3 hours.

Heatwaves often cause immense stress on human society and the natural environment. While heatwaves can be classified into daytime, nighttime, and compound daytime-nighttime types, the specific processes associated with different heatwave types remain poorly understood. In this paper, we identify different mechanisms operating in compound (i.e., extreme heat during both day and night) and independent daytime and nighttime heatwaves in southern China. Compound heatwaves generally exhibit stronger temperature increases than either daytime or nighttime types. Daytime heatwaves are accompanied by increased downward shortwave radiation under a clear sky with reduced cloud cover and moisture. Nighttime heatwaves are characterized by more cloudy and moist conditions, and increased downward longwave radiation at the surface at night. A combination of these conditions for daytime and nighttime heatwaves prevail during compound heatwaves. All heatwaves are associated with strengthening and eastward extension of the South Asian high (SAH) in the upper troposphere, and strengthening and westward extension of the western North Pacific subtropical high (WNPSH) in the lower and middle troposphere. Further examinations suggest that compound heatwaves are accompanied by the strongest intensification of SAH and WNPSH. Compared with daytime heatwaves, nighttime events are associated with a stronger amplitude of SAH and WNPSH, and both highs tend to extend more southward when nighttime heatwaves occur. This southward extension induces an anomalous lower-level anticyclone that drives a southwesterly wind anomaly over southeastern China. This circulation feature transports warmer and more humid air towards southern China. The enhanced concentration of water vapor leads to increased absorption of outgoing longwave radiation, and increased re-emission of longwave radiation to the surface, thus resulting in surface warming at night.

Extremely heavy rainfall occurred over both Northwest India and North China in September 2021. The precipitation anomalies were 4.1 and 6.2 times interannual standard deviation over the two regions, respectively, and broke the record since the observational data were available, i.e., 1901 for India and 1951 for China. In this month, the Asian upper-tropospheric westerly jet was greatly displaced poleward over West Asia, and correspondingly, an anomalous cyclone appeared over India. The anomalous cyclone transported abundant water vapor into Northwest India, leading to the heavy rainfall there. In addition, the Silk Road pattern, a teleconnection pattern of upper-level meridional wind over the Eurasian continent and fueled by the heavy rainfall in Northwest India, contributed to the heavy rainfall in North China. Our study emphasizes the roles of atmospheric teleconnection patterns in concurrent rainfall extremes in the two regions far away from each other, and the occurrence of rainfall extremes during the post- or pre-monsoon period in the northern margins of monsoon regions.

This study investigates the impact of assimilating densely distributed GNSS zenith total delay (ZTD) and surface station data on very short-term heavy rainfall prediction associated with afternoon thunderstorms. A series of sensitivity experiments are applied to multiple cases characterized by intense rainfall rates within 2 hours in the Taipei Basin to identify key assimilation strategies for initializing afternoon thunderstorms. Data assimilation experiments are conducted with a convective-scale WRF-LETKF system, which assimilates the ZTD data and surface station data every 30 min. The results suggest that ZTD assimilation provides effective low-level moisture adjustments. The model generates strong convections due to the large amount of moisture flux over northern Taiwan, and thus heavy precipitation takes place in a short time. When station data is additionally assimilated, the location of the strongest convection is better predicted. Adopting variable- and scale-dependent covariance localization is crucial for optimizing the impact of ZTD assimilation. While the large-scale moisture correction from ZTD assimilation provides the essential thermodynamic precondition of convection development, the small-scale wind correction from surface station assimilation gives the ability to capture the wind direction over the complex terrain, which is decisive for the heavy rainfall location. Increasing the assimilation frequency gives the strongest near-surface convergence in the Taipei Basin. The well-coupled dynamic and thermodynamic conditions trigger extreme convection development in Taipei Basin and lead to a rainfall intensity that best agrees with the observations.

Water vapor (WV) is generally recognized as an essential climate variable and one of the most active components in the atmosphere. Being a typical type of greenhouse gas, the content and migration of WV are greatly associated with the intensity, time and extent of various types of severe weather events, e.g., rainstorm, typhoon and drought. Therefore, it is significant to further refine the existing methods to conduct continuous, timely and accurate monitoring of atmospheric humidity fields. Over the past few decades, on account of the benefits including high spatiotemporal resolution, long-term stability and all-weather capability, the emerging ground-based Global Navigation Satellite Systems (GNSS) atmospheric sounding technique has been widely used to sense WV content in the atmosphere. The GNSS-derived products of zenith total delay and precipitable water vapor have also advanced their usages in improving the accuracy of atmospheric WV fields simulated by numerical weather prediction models. In this paper, a comprehensive investigation on the optimal spatial resolution for assimilating GNSS-derived tropospheric products in the Weather Research and Forecasting (WRF) model to improve atmospheric humidity field is conducted in the context of Australia. By using the strategy of data thinning, the optimal spatial resolution for data assimilation was proved to be 46.4 km. By using the reanalysis data over the study period as the reference, it was found that, with the variational assimilation of GNSS-derived products, both the accuracy of initial field and that of final predictions were greatly improved. Therefore, the research findings not only further corroborate the effectiveness of using ground-based GNSS tropospheric products to improve model performance, but could provide some clues/insights in the development of more robust forecasting models.

The impact of assimilation of Moderate Resolution Imaging Spectroradiometer (MODIS) near infrared (NIR) water vapor data on the weather forecasting performance of mesoscale Numerical Weather Prediction (NWP) model, i.e., the Weather Research and Forecasting (WRF) model, is investigated in this study. Three WRF schemes including a no data assimilation scheme and two data assimilation schemes are performed over the South China area for February 2020 and July 2020, respectively. The no data assimilation scheme dose not assimilate any data and is treated as the background run. For the first data assimilation scheme, Precipitable Water Vapor (PWV) data from the MODIS onboard Terra satellite are assimilated into the WRF model. For the second data assimilation scheme, a Back Propagation Neural Network (BPNN) model with assistance of Global Navigation Satellite System (GNSS) PWV is adopted to calibrate the MODIS PWV data. Then, the MODIS calibrated PWV are assimilated into the WRF model. The performance of three WRF schemes are comprehensively assessed by actual observations from the GNSS, radiosonde, and meteorological stations. The results show that: (1) Within the first 12 h after data assimilation, assimilation of MODIS calibrated PWV gains an average PWV forecasting accuracy improvement of 7.6% for February period and 3.3% for July period, while the corresponding improvements of assimilation of MODIS raw (uncalibrated) PWV are 3.6% and 4.1% for February period and July period, respectively. (2) Both assimilation of MODIS raw PWV and calibrated PWV improve the humidity and temperature profile forecasting accuracy, particularly for July period. (3) For July period, after assimilating MODIS raw PWV and calibrated PWV, the accumulated rainfall forecasting success rate for the first 12 h after data assimilation increases from 62.7% to 63.1% and 62.9%, respectively.

Global positioning system (GPS) radio occultation (RO) data has been playing an important role in the new Korea Meteorological Administration (KMA) operational weather prediction system. The Korea Institute of Atmospheric Prediction Systems (KIAPS) delivered the global atmosphere NWP model - named the Korean Integrated Model (KIM) - to the KMA. GPS-RO provides a promising data set with a combination of global coverage, high vertical resolution, and all-weather capability. However, the accuracy of the observation operator in the lower atmosphere (troposphere) - particularly in the subtropical moist marine area - limits the effectiveness of assimilating RO data in these regions. We are currently using the standard Abel transform, which gives the minimum refractivity solution of the continuum consistent with a given RO bending angle (BA) profile, but this is known to give rise to large negative biases in the lower troposphere. As for this issue, we guessed two strategies could relieve the enormous gap of the refractivity (or bending angle) between the observation and the background field over the moist region. The first is to modify the observation operator directly to get more reliable simulation results based on a better understanding of the GNSS-RO features according to the atmospheric structure. And the other is enhancing the quality control (QC) process allows it could more sophisticatedly eliminate the observations that seem highly affected by the moisture.
In this study, we tried to strengthen the impacts of the GNSS-RO, especially over the lower atmosphere, by enhancing the QC process using the humidity information of the background field. And the new QC process was tested in conjunction with applying dynamical observation errors varying with the observation point. Preliminary results from the latest KIAPS DA cycle experiments with these GNSS-RO data processing modifications will be presented.

Precipitable water vapor (PWV), as the medium of land surface-atmosphere energy exchange, plays a crucial role in Earth radiative balance and global climate. However, the applicability of mainstream satellite PWV products is limited by the tradeoff between spatial and temporal resolution and spatiotemporal discontinuity caused by cloud contamination. In this study, we fused the reanalysis-based ERA5 and satellite-derived MODIS PWV to generate accurate and seamless PWV of high spatio-temporal resolution (0.01°, daily) across the continental Europe from 2015 to 2018 based on the generative adversarial network (GAN) model. Firstly, the optimal clear-sky ERA5- MOD05 image pairs are selected based on the image similarity as inputs to raise the fusion performance. Secondly, a flexible multiscale feature extraction architecture is used to capture spatio-temporal variations which primely addresses the non-stationarity of PWV. Finally, the attention-based GAN architecture is used to recover the high-resolution and seamless PWV based on the extracted multiscale features. The results indicate that the fused PWV has great consistency with the GPS retrievals (r = 0.88–0.95 and Bias = -0.89–3.13 mm), showing an improvement in the accuracy and continuity compared to the original MODIS PWV. The fused accurate and spatiotemporal continuous PWV has great potential for meteorological and hydrological analyses.

High-quality satellite remotely sensed water vapor observations present a vital role in Earth’s weather and climate. The Ocean and Land Color Instrument (OLCI) sensor, on-board Sentinel-3A and Sentinel-3B satellite platforms, can observe atmospheric water vapor data records at near-infrared (NIR) measurements. Yet the measurement performance of satellite-sensed OLCI water vapor estimates is generally inferior to that of ground-based and reanalysis-based water vapor observations. In our research, we develop a novel Back Propagation Neural Network (BPNN) based calibration approach to adjust the observation performance of official Sentinel-3 OLCI NIR water vapor products, which considers multiple influence factors that are correlated with the performance of satellite-based NIR water vapor measurements. The in-situ water vapor data records, measured from Global Positioning System (GPS) stations, are used as the desired resulting water vapor estimates. The model is validated based on ground-based water vapor data from GPS and radiosonde instruments. The results show that the BPNN-based adjustment approach can significantly improve the measurement quality of official Sentinel-3 OLCI NIR water vapor data records. Our model can reduce the root-mean-square error (RMSE) of official OLCI NIR water vapor by 24.52% from 3.10 mm to 2.34 mm for Sentinel-3A and 25.00% from 3.44 mm to 2.58 mm when compared to GPS-measured reference water vapor. When compared to radiosonde-observed reference water vapor, the RMSE reduces 21.01% from 4.57 to 3.61 mm and 20.81% from 5.19 to 4.11 mm for Sentinel-3A and Sentinel-3B, respectively. The BPNN-based calibration model exhibits a higher accuracy enhancement on OLCI/Sentinel-3 NIR water vapor observations, compared with other existing previous methods in previous studies. This could be because we have considered multiple dependence parameters in association with the measurement performance of OLCI-retrieved NIR water vapor, while the previous research did not do so.

Tropical cyclones are low pressure systems created by a local heating of the sea surface, which triggers an intense convective activity involving large amounts of water vapor. A deadly tropical cyclone was formed in the early April 2021 over the Savu Sea and was given the name Seroja. This study aims to understand the characteristics of Seroja based on the precipitable water vapor (PWV) as derived from the Global Navigation Satellite Systems (GNSS) data. Observations at 22 permanent GNSS stations, as well as in situ meteorological sensors, in the Nusa Tenggara region from 1 March 2021 until 30 April 2021 were used for this purpose. Surface pressure values dropped significantly by more than 20 hPa, while relative humidity increased and temperature was reduced. Furthermore, during the formation of Seroja, PWV values gradually increased, reaching its peak around the time of when the cyclone was at its closest. After the cyclone had passed, PWV values decreased rapidly, dropping from around 70 mm to below 20 mm at CKUP. PWV pre-Seroja tends to be larger than post-Seroja, whereas post-Seroja PWV was more variable.

The AERONET program is a federation of ground-based sun-photometer networks and operates in partnership with various national and international agencies, universities, institutions, research groups, and individual scientists and stakeholders. It provides worldwide observations of spectral aerosol optical depth, inversion products (microphysical and radiative properties), and precipitable water across diverse aerosol regimes. In addition, other network components enable measurements of normalized water-leaving radiance, solar flux, and marine aerosols. For the past three decades, AERONET has been providing long-term, continuous, and readily accessible public-domain datasets. The network has been growing steadily, with currently about 600 active sites. The program also supports research and application field campaigns around the world. The program continues to refine instrumentation, measurement techniques, and algorithms through research and development. AERONET has recently started making nighttime aerosol measurements using direct moon observations. The program's biggest strength is its imposition of standardized of instruments, calibration, quality control, processing, and data distribution. This presentation will provide an overview of the program with examples of available datasets, current research, and applications. We will also highlight important upcoming satellite missions, field campaigns, and new research results from the AERONET team at NASA Goddard. The presentation will highlight ongoing data analysis focused on the Asia-Pacific region by the GSFC team.

The GEMS validation campaign was conducted in the Seoul Metropolitan Area (SMA) from October to November, 2021. Twenty-four research teams participated in four platforms: ground remote sensing, chemical sonde, ground and airborne in situ chemical analysis, and chemical transport modeling. In particular, 5 Pandora and 5 MAX-DOAS were deployed within SMA, and 3 Car-DOAS were operated throughout SMA. Preliminary results showed that diurnal variations in the NO2 and HCHO columns in ground remote sensing matched well with those in GEMS. Car-DOAS yielded pollution maps in detail, which were used for subgrid variability analysis. The various shapes in the vertical profile of NO2 retrieved from MAX-DOAS and directly measured from the airborne NO2 monitor demonstrate the importance of A prior in GEMS VCD calculation.

The intensive field observations were conducted in central to southwestern Taiwan to clarify the physical and chemical processes and emission sources that lead to the serious PM_2.5 problem. The deployment of the atmospheric observations included the wind profiler, radiosonde, unmanned aerial vehicle (UAV), lidar, and surface datasets. These observations provide the wind fields, temperature, humidity, and aerosol information with high temporal and vertical resolutions in the planetary boundary layer (PBL). In addition, the PBL ensemble data assimilation system was developed based on the Weather Research and Forecasting Local Ensemble Transform Kalman Filter (WRF-LETKF) framework coupled with the Community Multiscale Air Quality (CMAQ) model. We applied WRF and CMAQ models, and observations, to characterize the PBL vertical structures, evolutions, and air pollutants transport and dispersion processes. The preliminary analysis of the observation and model results demonstrate that the development of the PBL depth was limited due to the enhanced atmospheric stability in the PBL and the strong subsidence process from the layer above. The coastal sites revealed a shallow PBL depth; yet, the inland sites revealed a well-mixing PBL structure during the day. The lidar and radiosonde revealed distinct layer structures of the PM2.5 concentrations. The model can reproduce the observed features and assist in discussing the air pollution transport processes.

A recently developed GRASP/Component approach (GRASP: Generalized Retrieval of Atmosphere and Surface Properties) was applied to AERONET (Aeronet Robotic Network) sun photometer measurements in this study. Unlike traditional aerosol component retrieval, this approach allows the inference of some information about aerosol composition directly from measured radiance, rather than indirectly through the inversion of optical parameters, and has been integrated into the GRASP algorithm. The newly developed GRASP/Component approach was applied to 13 AERONET sites for different aerosol types under the assumption of aerosol internal mixing rules to analyze the characteristics of aerosol components and their distribution patterns. The results indicate that the retrievals can characterize well the spatial and temporal variability of the component concentration for different aerosol types. A reasonable agreement between GRASP BC retrievals and MERRA-2 BC products is found for all different aerosol types. In addition, the relationships between aerosol component content and aerosol optical parameters such as aerosol optical depth (AOD), fine-mode fraction (FMF), absorption Ångström exponent (AAE), scattering Ångström exponent (SAE), and single scattering albedo (SSA) are also analyzed for indirect verifying the reliability of the component retrieval. It was demonstrated the GRASP/Component optical retrievals are in good agreement with AERONET standard products [e.g., correlation coefficient (R) of 0.93 – 1.0 for AOD, fine-mode AOD (AODF), coarse-mode AOD (AODC) and Ångström exponent (AE); R = ~ 0.8 for absorption AOD (AAOD) and SSA; RMSE (root mean square error) < 0.03 for AOD, AODF, AODC, AAOD and SSA]. Thus, it is demonstrated the GRASP/Component approach can provide aerosol optical products with comparable accuracy as the AERONET standard products from the ground-based sun photometer measurements as well as some additional important inside on aerosol composition.

Changes in urbanization, industry, and biomass burning in South, Southeast, and East Asia, have led to rapidly changing emissions of gasses and aerosols. These changes are due to missing and misrepresented sources, extreme events, and time-varying sources. Due to few ground stations outside urban areas in China, Korea, Japan, Singapore, and Thailand, this work uses a new, model-free approach in combination with daily OMI measurements of NO₂ and UV, to identify and quantify emissions where measurements are available. An emphasis is made in regions that have undergone rapid or significant change. Using observations from the same platform takes advantage of consistent spatial and temporal coverage while taking advantage of the fact that NO₂ and BC are co-emitted from the same processes and under the same thermodynamic conditions. The BC mass, size, and number loadings are all computed using AERONET SSA and a core/shell-constrained MIE model. Through the application of the mass-conserving model-free approach constrained by first-order approximations of wet and dry deposition, chemical aging, and dynamic transport, the computed number, size, and mass emissions profiles are computed. Differences between existing a priori emissions from FINN and EDGAR are compared with the top-down emissions database and the ranges of allowable physical, chemical, and dynamical terms. All products and analyses are done grid-by-grid in both space and time, with an emphasis on extreme events previously identified from OMI NO₂ columns in 2016.The results demonstrate a significant underestimation in rural areas in Myanmar, Northern Thailand, Laos, and Northeast India, as well as in suburban areas around cities in Southeast Asia. However, some highly developed areas have overestimated emissions, likely due to more strict environmental enforcement, including Shanghai, Singapore, and Taibei. The uncertainty and day-to-day variability are examined in further detail, supporting that the inversion results are statistically significant in the regions mentioned.

Simultaneous observations of carbon dioxide (CO₂), the major greenhouse gas (GHG), and nitrogen dioxide (NO₂), a tracer for fossil fuel combustion, help to identify and quantify anthropogenic CO₂ sources. Ship-, aircraft-, and ground-based observations together with satellite observations of GHG and other trace gases aim to better understand changes in their atmospheric concentrations. To achieve this goal, the global coverage of in situ observations by public and private networks is expanding, and new satellite missions are scheduled like that of GOSAT-GW. By using an approach to integrate ship data of the Ship-of-Opportunity program and aircraft data of the Comprehensive Observation Network for Trace gases by Airliner, we set up a framework to evaluate satellite observations of the column-averaged dry-air mole fractions of, for example, CO₂ (XCO₂) over the ocean. Although the applicability to CH₄ is currently limited due to the sparseness of in-situ data, we explore the potential of this approach for the future when the observation networks are expanded. Complementary to this approach to constrain anthropogenic CO₂ emissions and to validate CO₂ and NO₂ data of the GOSAT-GW mission, we are aiming to make continuous cargo ship-based observations of XCO₂, XCH₄, and XCO using a semi-automatic Fourier transform infrared (FTIR) spectrometer, combined with a UV spectrometer to measure the vertical column densities of NO₂ (VCD_NO2). The setup, developed by the Heidelberg University, consisted initially only of the mobile semi-automatic FTIR, and was tested for the first time on a cargo ship in the summer 2022. Currently, the Heidelberg University is integrating the UV spectrometer. Scheduled for end of this year, the novel setup will be deployed on a cargo ship operating along major anthropogenic emission sources of the Japanese East Coast. We will present the concepts, challenges, and perspectives for validating future satellite missions and monitoring anthropogenic emissions.

Nitrogen dioxide (NO₂) is a major air pollutant. Tropospheric NO₂ vertical column densities (VCDs) retrieved from sun-synchronous satellite instruments have provided abundant NO₂ data for environmental studies, but such data are limited by insufficient temporal sampling (e.g., once a day). The Geostationary Environment Monitoring Spectrometer (GEMS) launched in February 2020 will retrieve NO­₂ at an unprecedented high temporal resolution. Here we present a research product for tropospheric NO₂ VCDs, referred to as POMINO-GEMS. We develop a hybrid retrieval technique combining GEMS and TROPOMI observations as well as GEOS-Chem simulations to generate hourly tropospheric NO₂ slant column densities (SCDs). We then derive tropospheric NO₂ air mass factors (AMFs) with explicit corrections for the anisotropy of surface reflectance and aerosol optical effects, through pixel-by-pixel radiative transfer calculations. Prerequisite cloud parameters are retrieved with O₂-O₂-based cloud algorithm by using ancillary parameters consistent with those used in NO₂ AMF calculations. Initial retrieval of POMINO-GEMS tropospheric NO₂ VCDs for June–August 2021 exhibit strong hotspot signals over megacities and distinctive diurnal variations over polluted and clean areas. POMINO-GEMS NO₂ VCDs agree well with our POMINO-TROPOMI v1.2.2 product (R = 0.96, and NMB = 6.9%). Comparison with MAX-DOAS VCD data at five sites shows a small bias of POMINO-GEMS (NMB = –12.2%); however, the correlation for diurnal variation varies from -0.57 to 0.86, suggesting strong location-dependent performance. Surface NO₂ concentrations estimated from POMINO-GEMS VCDs are consistent with measurements from the Ministry of Ecology and Environment of China at 855 sites (NMB = – 5.6%, and R = 0.96 for diurnal correlation averaged over all sites). POMINO-GEMS will be made freely available for users to study the spatiotemporal variations, sources and impacts of NO₂.

Local environmental conditions in Southeast Asia, especially high humidity and frequent cloud cover, pose challenges in atmospheric monitoring both via modelling and when using satellite observations. We explore opportunities to address such challenges in the case of smoke haze dispersion, with the synergistic use of model outputs and satellite products. The Numerical Atmospheric-dispersion Modelling Environment (NAME) is an atmospheric pollution dispersal modelling tool capable of simulating and forecasting many atmospheric dispersion phenomena and associated physical and chemical processes. Himawari is a geostationary satellite providing high frequency multispectral observations. We use Himawari Aerosol Optical Depth (AOD) and Fire Radiative Power (FRP) products. We focus on 2019 and explore use of Himawari FRP, instead of the MODIS FRP assimilated by CAMS GFAS which is currently used as input to NAME. The aim is to evaluate the impact of the higher observation frequency of Himawari and the differences in saturation limit between the two instruments on the emission outputs of the model. We further compare model-derived AOD with Himawari AOD data, to get a view of the uncertainties of AOD estimation. Finally, we train a Machine Learning algorithm to translate AOD to column-integrated Particulate Matter emissions using NAME model simulations. This translation is traditionally done using smoke mass extinction coefficients measured in laboratory experiments or field expeditions. However, there is limited availability of such data in Southeast Asia, and their accuracy is further impacted by high levels of humidity in the region and plume aging. Use of model outputs in the ML algorithm could complement the lack of data and cover a wider range of local conditions. Based on our findings, we discuss potential synergistic use of satellite products and NAME outputs, for research regarding deriving smoke haze emissions in the region as well as for operational monitoring and forecasting of haze dispersion.

Understanding the sea surface wind structure during tropical cyclones (TCs) is the key for study of ocean response and parameterization of air-sea surface in numerical simulation. However, field observations are scarce. In 2019, three wave gliders were deployed in the South China Sea and the adjacent Western Pacific region, which acquired sea surface wind structure of eight TCs. Analysis of the field data suggests that the wave glider-observed surface winds are consistent with most analysis/reanalysis data (i.e., ERA5, CCMP and NCEP-GDAS) and SMAP. Both wave glider observations and analysis/reanalysis data indicate that TC wind fields induce an obvious increase in speed toward the sea surface together with a sharp change in direction, showing an asymmetric wind structure which is sensitive to TC translation speed and intensity. Larger mean values of wind speed and inflow angle are located on the right side along TC tracks. The inflow angle shows a highly dynamic dependence on the radial distance from the TC center, the TC intensity, as well as the TC-relative azimuth. Comparisons between field observations and theoretical models indicate that the most widely used, ideal TC wind profile models can largely represent the observed sea surface wind structure, but generally underestimate the wind speed due to lack of consideration of background wind. Moreover, simple ideal models (e.g., the modified Rankine vortex model) may outperform complex models when accurate information of TCs is limited. Wave glider observations have potential for better understanding of air-sea exchanges and for improvements of the corresponding parameterization schemes.

Eyewall Replacement Cycle (ERC) is often seen in TCs. ERC occurs when secondary eyewall forms outside the inner eyewall, and the inner eyewall disappears. ERC significantly impacts TC intensity, so revealing the mechanism is an important issue both scientifically and socially. Several mechanisms of secondary eyewall formation have been proposed. According to Huang et al. (2012), tangential wind enhancement associated with lower-level inflow causes secondary eyewall formation. The relationship between the mesoscale descending inflow (MDI) formed by diabatic cooling of stratiform precipitation areas and secondary eyewall formation has also been pointed out (Didlake et al. 2018). However, the detailed processes by which dry air inflow and diabatic cooling affect secondary eyewall formation through MDI are not yet well understood. Therefore, in this study, the role of dry air inflow from the middle and upper troposphere and diabatic cooling in secondary eyewall formation is investigated using numerical experiments. Idealized numerical experiments were conducted using the plane version of the nonhydrostatic icosahedral atmospheric model, NICAM. Control experiments confirmed the existence of dry air inflow in the middle and upper troposphere and the formation of downdrafts due to diabatic cooling. It was also confirmed that the mechanism of secondary eyewall formation by agradient force was working, as pointed out in previous studies. In a sensitivity experiment, we conducted an experiment to increase water vapor in the middle and upper troposphere outside the TCs. The results showed that the secondary eyewall formation was hindered and slowed down as the water vapor increased. We further conducted a realistic case study for Typhoon Haishen in 2020 and confirmed the above mechanism works for the ERC of Haishen.

Doppler weather radars are a powerful tool for investigating the inner-core structure of tropical cyclones (TCs). The Doppler velocity from a single radar has no information on the wind component normal to the radar beam. Therefore, closure assumptions are needed to estimate the circulations of the TCs from single-Doppler radar observations. The Generalized Velocity Track Display (GVTD) technique used to estimate the TC circulations adopts the closure assumption of no asymmetric radial winds in the TC vortex.
The present study proposes a new closure assumption introducing asymmetric radial winds to improve the axisymmetric-circulations estimation in TCs with asymmetric structure in the GVTD retrieval formula. Our new method can consider the asymmetric radial winds by using streamfunction based on the Helmholtz decomposition theorem of horizontal winds. As with GVTD, the new method retrieves TC circulations based on the Fourier decomposition of winds in the azimuthal direction and the least-square fit of the Doppler velocity from the single radar observation. The new method and GVTD are applied to analytical vortices and a real typhoon. For the analytical vortices with asymmetric winds in wavenumber-2 vortex Rossby waves, the axisymmetric tangential wind of V_T0 retrieved by the new method (GVTD) has a relative error of less than 2% (10%) near the radius of maximum wind speed. For the real typhoon, the GVTD-estimated V_T0 has periodical fluctuations with an amplitude of about 5 m/s near the elliptical eyewall. The period of the fluctuations is approximately synchronized with the counterclockwise rotating period of the elliptical shape of the eyewall, suggesting pseudo-signals due to the closure assumption in GVTD. The periodical fluctuations are largely reduced in the estimated V_T0 from the new method. We find that the new method can reduce the pseudo-signals of the GVTD-retrieved axisymmetric circulation in cases of asymmetric vortices.

Accurate prediction of rainfall distribution associated with tropical cyclone (TC) is very important for risk management but remains challenging due to complicated multi-scale interactions. Previous studies mostly focus on the precipitation associated with a single TC but little attention has been paid to the features of precipitation in binary tropical cyclone system. In this study, based on the satellite observations, a novel approach with rotated coordinate system is proposed to examine the binary TC interactions over the western North Pacific and the associated asymmetry of rainfall distributions. The results show that the asymmetric component of TC rainfalls strengthened with the decreasing of the separation distance between the two TCs. The results show that the asymmetric component of TC rainfalls enhances with the decreasing of the separation distance between the two TCs. A critical separation distance at about 2050 km for binary TCs could naturally emerge based on the relationship between the asymmetric rainfall component and the separation distance. This can be an observational reference for estimating the strength of binary-TC interaction and identifying binary TCs from two coexisting TCs. When two TCs become nearby, the asymmetric component of rainfall shows an increasing trend with rainfall significantly suppressed in Quadrant IV of the TC located to the west when orienting the two TCs in the west-east direction. The suppression becomes remarkable once the separation distance between the two TCs is within about 2,050 km. It is found that the convective activity in one TC is related to the deep-layer vertical wind shear (VWS) from its companion. Rainfall is enhanced downshear-left in a TC, consistent with a single TC embedded in an environmental VWS as found in previous studies. Our study suggests that the evolution of rainfall distribution under the binary-TC conditions could be potentially predictable at vortex scale.

Various aspects of tropical cyclones (TCs) fluctuate with the diurnal cycle. In this paper, diurnal variations in TC intensification and the growth of the radius of gale-force winds (34 kt; R34) over the Ocean for in global TCs were investigated using best-track data. Statistically significant diurnal variations were found for global TCs, with a maximum intensification over the period 03–09 local solar time (LST). No diurnal signals are detected for TC decay. Statistically significant diurnal variations are found for global TCs which experienced rapid intensification (RI, ≥30 knots within 24 h), with a maximum 6-h growth rate occurring at 03–09 LST. The highest rate of intensification and R34 growth rates at 03–09 LST for TCs were associated with the greatest coverage of very deep convective clouds with infrared brightness temperatures < 208 K at 03–06 LST. R34 growth is favored at the same time as TC intensification. This study suggests that nocturnal radiative cooling affects the intensity and outer region sizes.

Size of tropical cyclone (TC) is often asymmetric in nature. Yet, there is a lack of clean and intuitive definition/expression to specify the asymmetry of TC size. Here, we introduce a novel index, TC size asymmetry index (SAI), which indexes both the degree and pattern of the asymmetry systematically. In particular, the symbolic form of SAI is vividly designated for identifying the latter. The SAI proposes 1 quasi-symmetric pattern and 28 asymmetric patterns in total. The 41-yr (1979–2019) global climatology of SAI shows that the distribution of the degree of TC size asymmetry is trimodal. Elementarily, the degree and pattern of TC size are found to be TC intensity, TC movement, time, and space dependent. The introduction of SAI does not only give an insight into the subject of TC size asymmetry but also lays important foundation for future applications and research. Furthermore, besides meteorology, it could inspire other fields to index the geometric asymmetries of other kinds.

Tropical Cyclone Debbie (2017) made landfall near Airlie Beach on 28 March 2017 causing 14 fatalities and an estimated US$2.67B economic loss and was ranked as the most dangerous cyclone to hit Australia since TC Tracy in 1974. In addition to the extreme flooding as TC Debbie moved onshore and down the east coast of Australia, it intensified rapidly just offshore from Category 2 to Category 4 in approximately 18 hours and finally made landfall as a Category 4 TC, causing widespread and disastrous damage. A high-resolution WRF simulation (1-km horizontal, and 10-min temporal resolution) is used to analyze the inner-core structure and evolution during the offshore rapid intensification period in the current conditions and potential future change. In current condition, Debbie’s a rapid intensification (RI) stage is characterized by three rounds of eyewall breakdown into mesovortices and re-development events. Each round of breakdown and re-establishment brings high potential vorticity and equivalent potential temperature air back into the eyewall, re-invigorating eyewall convection activity and driving intensification. The potential future changes in the inner-core structure and eyewall evolution will also be discussed using WRF with the Coupled Model Intercomparison Project Phase 6 (CMIP6) perturbed conditions to better assess the possible TC intensity change under different climate change scenarios. 

Using the tropical cyclone best track data from Shanghai Typhoon Institute of CMA of 1949-2020 and the reanalysis data of the European Centre for Medium-Range Weather Forecasts of 1991-2020, and the joint EOF analysis of u and v components of the wind field on the 200 and 850 hPa, the characteristics of the larger-scale circulation at onset of the rapid-intensification typhoon are summarized, and the evolutions of environmental dynamic and thermal conditions during the ±12 h period at the onset of typhoon rapid intensification are further analyzed. The results indicate that the lower level of the main circulation of EOF decomposition is the convergence pattern of monsoon trough at the onset of the typhoon rapid intensification, and the circulation is conducive to the low-level water vapor transport of typhoon. The upper-level circulation has obvious typhoon outflow channels, and the characteristic can be used as a typical circulation for the rapid intensification forecast. The thermal conditions such as sea surface temperature, water vapor and convective instability, as well as dynamic conditions such as environmental vertical wind shear and the strength of upper-level outflow all can generally reach the fitness range of conditions that are favorable for typhoon intensification. However, the values of above environmental factors have not changed significantly or suddenly during the transition from the slow intensification process to rapid intensification process. Even some extreme cases show that the change tend of environmental factors towards to the unfavorable conditions. These research results provide a reference for the prediction of rapid intensification and further study of typhoon in the future. As for the unfavorable conditions in RI some cases, whether there are other favorable factors to offset the negative effects of these conditions and what are the corresponding physical processes, these issues need to be further studied in the future.

NOx (NOx = NO + NO₂) is a major component of air pollution with substantial spatiotemporal variations. Current emission inventories lag in time for several years and can hardly capture the fine-scale spatial pattern of NOx emissions. Here we develop a novel algorithm (PHLET) to retrieve NOx emissions at high spatial resolutions (≤5 km) based on tropospheric NO₂ vertical column densities (VCDs) retrieved from satellite instruments. The algorithm derives the lifetimes and emissions of NOx at individual locations by explicitly accounting for the nonlinear chemistry and horizontal transport. We apply the algorithm to obtain NOx emissions for China at 5 km resolution in summer (JJA) 2012–2022 based on POMINO NO₂ VCD data for OMI and TROPOMI instruments. Our PHLET data reveal many fine-scale emission details and considerable small-to-medium-scale sources missing in current anthropogenic inventories. Those missing sources are associated with human activities reflected in road network and Tencent user location data. Our PHLET data also show substantial inter-regional differences in interannual variations and trends of NOx emissions over the past decade, which are also absent in current inventories. Although the total NOx emission in China has been decreasing, the contribution of emissions from the western provinces to the national total has been growing and has exceeded the contribution of the eastern coastal provinces. Besides the anthropogenic sources, our PHLET retrieval discovers large NOx emissions from natural lakes on the Tibet Plateau that are previously unknown. Such emissions are likely a result of the anammox process under rapid warming over the plateau as an unaccounted feedback between climate change, lake ecology and nitrogen emissions. Overall, our NOx emission retrieval serves as a crucial tool to detect and quantify anthropogenic and natural emissions at fine scales, in support of air quality and climate modeling and targeted emission mitigation.

Organosulfates (OSs) as important constituents of atmospheric organic aerosol (OA) are widespread in ambient environments. Although the sources, formation process, and chemical composition of OSs have been diversely studied, OSs formed by anthropogenic emissions are still little known. In this study, the molecular compositions of OSs in atmospheric PM_2.5 samples collected from a winter measurement campaign (SEISO-Bohai) at Jingtang Harbor, a typical traffic environment, are characterized by ultra-performance liquid chromatography (UHPLC) coupled to electrospray ionization ultrahigh resolution mass spectrometer (UHRMS) Orbitrap. The changing trends of the port OS chemical compositions are observed from one complete haze pollution episode, which happened in this campaign. Along with the increased degree of haze pollution, the relative abundances of OSs have been apparently added and the molecule structures have become more complex. The processes of oxidation and fragmentation drive the OS formations in the haze environment. Then, the potential precursors from ship emissions are identified based on the “OS precursor map” created by previous study. The high molecular weight and low degree of unsaturation and oxidization of OSs are suggested to be mainly derived from ship intermediate/semi-volatile organic compound (I/SVOC) emissions. These OSs have appeared the wide range of molecule weight and good chemical homogeneity in the clean aerosol samples. In addition, our study also finds that ship emissions should further facilitate the production of OSs under the haze pollution condition. 

Air pollution in India is a growing concern, which has significant adverse effects on human health and ecosystems. However, there are very limited observations of air pollutants, and assessments of specific effects of air pollution are hindered. This study aims to simulate air pollutants in India from 2015 to 2019 using the Community Multi-scale Air Quality (CMAQ) model at the resolution of 36 × 36 km². The meteorological fields are generated by the Weather Research and Forecasting (WRF) model. The anthropogenic emissions are generated from Emissions Database for Global Atmospheric Research (EDGAR) and the MIX Asian emission inventory. Emissions are adjusted to simulation years with different adjustment coefficients for different sectors and states. Biogenic and wildfire emissions are also provided to CMAQ. After the model results are validation, the levels, health and ecosystem effects are analyzed. The study would reveal the importance of controlling particulate matters and ozone in India and provide information for designing effective control strategies.

With decreases in anthropogenic precursors, natural precursors may become more important in O₃ pollution. However, the biogenic volatile organic compounds (BVOCs) change and its impact remain unclear due to the lack of long-term measurements of VOCs. In this study, we emphasize the increasingly important role of BVOCs in O₃ pollution by analyzing the long-term measurements in Hong Kong obtained during 2013–2019. Driven by the warming temperature, biogenic isoprene at a suburban site in Hong Kong increased 0.05 ± 0.02 ppbv/yr (9%/yr), which is in sharp contrast to the decreases in anthropogenic precursors during study period. Detailed chemical modeling shows that increased BVOCs enhanced local O₃ production by 25%, and the most obvious effect was shown in summer. Increased BVOCs also affected the non-linear relationships between O₃ and anthropogenic precursors, i.e., increasing the O₃ sensitivity to nitrogen oxides (NOx) and decreasing the O₃ sensitivity to anthropogenic VOCs (AVOCs). Despite changes in precursors, the O₃ formation remains in VOC-limited regimes at Tung Chung. Joint control of AVOCs and NOx (at a ratio greater than 1.2) would help avoid exacerbation of O₃ pollution and reduce NO₂ pollution. Our findings also suggest the BVOCs increase likely have occurred in the larger Pearl River Delta (PRD) region, and thus the results from our study in Hong Kong may have implications for developing AVOCs/NOx reduction measures in the PRD region and beyond.

Long-chain unsaturated fatty acids (uFAs), such as oleic acid, undergo rapid degradation via heterogeneous reactions with atmospheric oxidants upon emission. The oxidation mechanism and kinetics have been extensively studied in laboratory experiments. However, quantitative knowledge of degradation rates under real-world atmospheric conditions is scarce. We obtained the nighttime decay rates of three cooking-related uFAs using a relative rate approach applied to bihourly measured data in urban Shanghai. The estimated lifetime of oleic acid was 6 h under conditions of ∼12 ppb ozone and 60%–100% relative humidity encountered at our urban location or an inferred ∼2 h at a higher ozone level of ∼40 ppb. The decay rates of elaidic and linoleic acid are determined to be 0.62 and 1.37 that of oleic acid, respectively. This work provides the first kinetic data pertaining to real-world conditions. They are valuable for constraining the modeling of heterogeneous aging of ambient organic aerosols.

Polycyclic aromatic hydrocarbons (PAHs) are unavoidably derived from combustion processes, and are contaminants of global concern because they increase the risk of lung cancer and are detrimental to human health and the ecosystem. While high concentrations of PAHs were already measured in 2008, future changes in energy use, land use, and climate policy may alter the PAHs concentrations. Integrating a global atmospheric chemistry model, a lung cancer risk model, and plausible future emissions trajectories of PAHs, we assess how global PAHs and their associated lung cancer risk will likely change in the future. Benzo(a)pyrene (BaP) is used as an indicator of cancer risk from PAH mixtures. From 2008 to 2050, the population-weighted global average BaP concentrations under all RCPs consistently exceeded the WHO-recommended limits, primarily attributed to residential biofuel use. Peaks in PAH-associated incremental lifetime cancer risk shift from East Asia (4×10^-5) in 2008 to South Asia (mostly India, 2-4×10^-5) and Africa (1-2×10^-5) by 2050. In developing regions of Africa and South Asia, PAH-associated lung-cancer risk increased by 30-64% from 2008 to 2050, due to increasing residential energy demand in households for cooking, heating, and lighting as the rapid population growth, as well as the continued use of traditional biomass use, increases in agricultural waste burning, and forest fires. With the stringent air quality policy, PAH lung-cancer risk substantially decreases by ~80% in developed countries. Climate change is likely to have minor effects on PAH lung-cancer risk compared with the impact of emissions. Future policies, therefore, need to consider efficient combustion technologies that reduce air pollutant emissions, including incomplete combustion products such as PAH.

Aiming to control the rising Particulate Matter (PM) emissions, many researchers highlighted the importance of identifying PM sources within and transboundary of the city. This study investigates the seasonal and spatial heterogeneity of PM concentration and chemical characteristics. The source apportionment technique was performed to identify the potential sources' contribution in Vijayawada city. Based on the land use land cover (LULC) pattern, the atmospheric PM₁₀ and PM_2.5 samples were collected at multiple monitoring sites with Traffic, Commercial, Industrial, Residential, and Background characteristics during winter and summer. The highest average PM concentration (PM₁₀: 154±42 µg/m³, PM_2.5: 82±21 µg/m³) was measured at the industrial site in winter followed by traffic, commercial, background, and residential sites with an estimated difference of (20%, 14%), (31%, 30%), (36%, 34%) and (36%, 25%), respectively. A similar trend was observed in the summer season, with the highest concentrations at the industrial site (PM₁₀: 86±24 µg/m³, PM_2.5: 52±15 µg/m³) followed by traffic, commercial, background and residential sites. Secondary Inorganic Aerosols (SIA) (7-35%, 6-24%) and crustal elements (4-35%, 11-18%) were observed to be the significant components in PM₁₀ and PM_2.5 mass size fraction, respectively. Low OC/EC ratio values ranging from 0.98-2.95 indicate a high proportion of gasoline and diesel emissions in the urban atmosphere. Chemical Mass Balance (USEPA CMBV5.0) model identified vehicular, industrial, construction, road dust, SIA, biomass burning, and coal combustion sources, respectively, with a significant variation in seasonal and spatial contributions. Vehicular emission is one of the primary sources, contributing 47-16% in PM₁₀ and 51-22% in PM_2.5, followed by the combined contribution of biomass and coal combustion, which varies from 35-13% and 31-20% in PM₁₀ and PM_2.5, respectively. High biomass and coal combustion emissions were attributed to several industrial units which use coal, wood, briquettes, etc., for fuel requirements.

Accurate modeling of PM_2.5 is essential for effective air quality management and mitigation. Chemical transport models (CTMs) like the Community Multiscale Air Quality (CMAQ) model, are widely used for PM_2.5 simulation through atmospheric processes of dispersion, deposition, and chemical reactions. However, CMAQ often suffers from biases due to limitations in model structure, uncertainties in initial and boundary conditions, and insufficient representation of meteorological conditions and source contributions. Traditional methods for model bias diagnosis usually rely on empirical and priori assumptions and require extensive sensitivity tests with high demands on computational resources, such as Monte Carlo methods or Latin hypercube sampling. Recently machine learning (ML) methods have been widely used in environmental science researches due to their simple structure, fast speed and ability to deal with no-linear relationships. ML also provides a new perspective on the identification of simulation biases. However, as a complex multi-phase mixture, it is still challenging to diagnose biases in PM_2.5 simulations using ML methods. In this study, we plan to diagnose the drivers of the model bias in simulating surface PM_2.5 concentrations based on the lightGBM model, an efficient ensemble ML method. The individual components of PM_2.5 will be simulated using CMAQ and the sectoral sources of PM_2.5 will be tracked using source apportionment methods. CMAQ simulation bias will be diagnosed from multiple perspectives, including meteorology, PM components, and sectoral sources. The results of this study can provide new ideas for CTMs model biases diagnosis, deepen the understanding of CMAQ simulation biases, and provide information for model improvement.

Monsoon systems are transporting water vapour and energy across the globe, making them a central component of the global circulation system. Changes in different forcing parameters have the potential to fundamentally change the monsoon characteristics as indicated in various paleoclimatic records and as projected by the latest generation of general circulation models in the context of the Coupled Model Intercomparison Project Phase 6 (CMIP6). Here, we use 32 CMIP6 models to analyze the model projections for the Indian summer monsoon with particular focus on the seasonal summer monsoon rainfall, its interannual variability and the occurrence of extremes. Besides, we use the Atmosphere Model version 2 developed at the Geophysical Fluid Dynamics Laboratory (GFDL-AM2) and couple it with a slab ocean to analyse the monsoon's sensitivity to changes in different forcing parameters on a planet with idealized topography. This Monsoon Planet concept of an Aquaplanet with a broad zonal land stripe allows to reduce the influence of topography and to access elementary meridional monsoon dynamics behind the competing effects of different forcings.

The Asian summer monsoons are globally significant meteorological features, creating a strongly seasonal pattern of precipitation. The stability of this hydrological cycle is of extreme importance for ecosystems and the livelihoods of a large share of the world’s population. Simulations are performed with an intermediate complexity climate model, PLASIM, to assess the future response of the Asian monsoons to changing concentrations of aerosols and greenhouse gases. The radiative forcing associated with aerosol loading consists of a mid-tropospheric warming and compensating surface cooling, which is applied to India, Southeast Asia and East China, both concurrently and independently. The primary effect of increased aerosol loading is a decrease in summer precipitation in the vicinity of the applied forcing, although the regional responses vary significantly. The decrease in precipitation is only partially ascribable to a decrease in the precipitable water, and instead derives from a reduction of the precipitation efficiency, due to changes in the stratification of the atmosphere. When the aerosol loading is added in all regions simultaneously, precipitation in East China is most strongly affected, with a quite distinct transition to a low precipitation regime as the radiative forcing increases beyond 60 W/m². The response is less abrupt as we move westward, with precipitation in South India being least affected. By applying the aerosol loading to each region individually, we are able to explain the mechanism behind the lower sensitivity observed in India, and attribute it to aerosol forcing over East China. Additionally, we note that the effect on precipitation is approximately linear with the forcing. Doubling carbon dioxide levels acts to increase precipitation and weakening the circulation over the region. When the carbon dioxide and aerosol forcings are applied simultaneously, the carbon dioxide forcing partially offsets the surface cooling and reduction in precipitation associated with the aerosol response.

Because increased greenhouse gas emissions considerably warm and moisten the Earth's atmosphere, one may expect an increase in monsoon precipitation during the historical period. However, we find the observed Northern Hemisphere land summer monsoon (NHLM) precipitation has significantly decreased since 1901. Simulations from Coupled Model Intercomparison Project Phase 6 (CMIP6) well reproduce global warming and the drying of NHLM since the industrial revolution when forced by observed external forcings. Result from single forcing experiment shows that the anthropogenic aerosol (AA) dominates the Northern Hemisphere (NH) monsoon precipitation drying, while the greenhouse gases (GHG) largely control surface warming. Thus, the NH monsoon precipitation responds to AA more sensitively than the GHG. The AA can more effectively modulate downward solar radiation reaching the surface, decreasing evaporation and weakening monsoon circulations by reducing the interhemispheric temperature difference and land-ocean thermal contrast, albeit with the same efficiency of the thermodynamic effect in the two forcings. Our result indicates the future intensive reduction of aerosol emission may rapidly recover the NH monsoon precipitation.

The Indian Ocean summer aerosol regime is of greater relevance to the regionally-important monsoon system of South Asia. The effects of aerosols such as black carbon (BC) on climate and buildup of the monsoon over the Indian Ocean are poorly constrained. Uncertain contributions from various natural and anthropogenic sources impede our understanding. Here, we use multi-year (2012-2017) observations of BC and its isotope fingerprint at a remote island observatory in the northern Indian Ocean to constrain loadings and sources during the little-studied monsoon season. Carbon-14 data pin down a largely fossil (65±15%) BC source domain. Occasional plumes from African savanna fire contribute up to (50 ± 5%) of the loadings over the summertime Indian Ocean. We estimate that the mass-absorption cross-section for this region is 7.6 ± 2.5 (m²/g), with a tendency to increase with savanna fire input. Simultaneous measurement of air and rain allow us to calculate washout ratios of BC and other chemical species. The washout rates are much lower for BC than OC and inorganic ions such as sulfate, implying a longer atmospheric lifetime for BC. The wet deposition flux for BC during the high-loading winter was three times higher than during the wet summer, despite much less precipitation in the winter. Taken together, the combustion sources, longevity, and optical properties of BC aerosols over the summertime Indian Ocean are different from the more-studied winter aerosol, with implications for chemical transport and climate model simulations of the Indian monsoon.

The rise of the trends of extreme rainfall events across all the major regions of India since 1901 has been revealed in our recent study. This machine-learning driven data analysis also suggested several potential drivers including urbanization and other land-use changes behind the observed trends. However, their causal relations need to be carefully examined by using, e.g., advanced models. Specifically, it has been indicated that urbanization can modify the water cycle and precipitations, either through the modification of land-use, or through effects induced by the emissions of anthropogenic aerosols. The thermodynamical perturbations induced by the presence of urban land-use, including the urban heat island effect, are known to induce rainfall modification due to perturbation of the flow and enhancement of the convective activity. However, this impact has yet to be clarified in a large scale, highly energetic system like the Asian Monsoon system. Using the high resolution meso-scale atmospheric model Meso-NH coupled with an urban-module, we have investigated the impact of urban land use on the heavy precipitation events during the Indian Summer Monsoon. The results of this study will be presented and discussed.

A high rate of deforestation has occurred in the Maritime Continent (MC) during recent decades due to the rapid growth of the local economy. MC deforestation is known to have a considerable influence on the local climate. However, its possible teleconnections to other regions are less understood. In this study, the influence of MC deforestation on precipitation over southern China is investigated using both reanalysis data and state-of-the-art climate models. The results show that MC deforestation could strengthen the late winter and early spring precipitation decline over southern China during 1979–2019. The enhanced regional convection due to MC deforestation leads to anomalous northward shifting of the tropical meridional circulation, with the ascending branch at 0°–10° N and descending at 20°–30° N compared with climatological ascending (10° S–0°) and descending (10°–20° N) branches. Such circulation change suppresses the moisture convergence and the development of convection over southern China. Our results suggest that, in addition to the local effects of deforestation, a further investigation of the remote impacts is essential for a thorough understanding of the climate influences of ongoing MC deforestation.

The classical understanding of monsoon onset implies a sudden increase in precipitation and sustainable rains. However, initial rain often gets stalled after monsoon onset for a week or even longer, causing disaster for farming. Here, I show that there are two types of critical transitions to monsoon: a direct transition, with a sudden increase in precipitation, and a two/multiple-step transition, with a dry spell after the initial rain. I present evidence that 70% of the last 47 years show a two/multiple-step transition, which went overlooked. Significantly, the second type of transition prevails under climate change. I uncover that the cause of rainfall cessation is the hidden phenomenon of intermittence emerging between two successive phase transitions. The new theoretical finding opens a door for the universal definition of local monsoon onset. I show how to evaluate the effect of climate change on the transition to monsoon in every state of the Indian subcontinent. 

Exposure to ambient air pollution, including fine particulate matter (PM_2.5) and trace gases like ozone (O₃) and nitrogen dioxide (NO₂) at the ground level, poses serious threats to environmental quality and public health, significantly increasing the risk of death. Satellite remote sensing allows for generating spatially continuous PM_2.5 data, but current datasets have overall low accuracies with coarse spatial resolutions limited by data sources and models. Air pollution levels in developing countries like China have experienced dramatic changes over the past couple of decades. To reveal the spatiotemporal variations, artificial intelligence, including machine and deep learning, is extended by considering the spatiotemporal heterogeneity of air pollution to generate long-term and high spatiotemporal-resolution outdoor air pollutants from big data that integrate ground-based measurements, satellite remote sensing products, atmospheric reanalysis, and model simulations. The application and fidelity of the dataset are demonstrated by analyzing their spatial distributions and temporal variations of surface air pollution, exposure risk to the public, and the COVID-19 pandemic. These novel products have been widely employed to address a variety of atmospheric, environmental, ecological, and public health issues.

Reactive nitrogen gases including NO_x and HONO strongly affect the atmospheric oxidization capacity through the contribution to the hydroxyl radicals (OH) and ozone (O₃). Although stringent anthropogenic NO_x emission controls policies have been implemented in past years, observations show that the surface O₃ concentrations in China are still increasing. Soils are important sources of emissions of NO and HONO that have been overlooked in previous studies; however, there is still a lack of research on the quantitative effects of soil nitrogen emissions on O₃ pollution in China. Since atmospheric HONO sources are not well understood, and the default HONO formation mechanism (NO + OH → HONO) always severely underestimates HONO observations and atmospheric oxidation capacity as a result, we first added HONO source from soil bacteria and other four potential sources (traffic emissions, NO₂ heterogeneous reactions on ground and aerosol surfaces, and inorganic nitrate photolysis in the atmosphere) in the WRF-Chem model. Soil NO_x emissions are also estimated by a new mechanistic parameterization, the Berkeley Dalhousie Iowa Soil NO Parameterization (BDISNP).In this study, the improved WRF-Chem model coupled with a new soil reactive nitrogen emissions (SoilNr) scheme is applied to quantitatively assess of the effects of soil NOx and HONO on ozone formation separately, and their combined effect on ozone formation in the North China Plain. The results of this study highlight the previously underappreciated important role of SoilNr on O₃ pollution and provide scientific basis for designing ozone pollution regulation strategies.

Here we apply a state-of-art global chemical transport model GEOS-Chem High Performance (GCHP) to understand the sources contributing to Chinese background (CNB) daily maximum 8 h average (MDA8) ozone, and to identify the driving factor of its interannual variability from 2015 to 2019. The five-year-mean CNB ozone is estimated as 37.8 ppbv, showing a general west-to-east downward spatial gradient. The national-mean CNB ozone is the largest in summer (42.5 ppbv), but distinct seasonality can be seen at different regions. Using the tagged ozone technique, we show that the high background levels in western China are due to abundant transport from the free troposphere and adjacent foreign regions, while in eastern China, domestic ozone formation near the surface from natural precursors is also important and exhibits intensive seasonal variation. We find the greater importance of biogenic volatile organic compounds (VOCs) over soil NO_x to ozone as reported in previous studies is reversed when domestic anthropogenic emissions are turned off, reflecting a more NO_x-sensitive ozone chemical regime in a “clean” atmosphere. The interannual variability (IAV) of CNB ozone shows the peak in summer, with standard deviation values during five years of ~5 ppbv over Qinghai-Tibet Plateau (QTP) and >3.5 ppbv over vast eastern China. CNB levels in QTP are found to be well correlated with horizontal circulation anomalies at 500 hPa, while in the east, year-to-year changes in soil NO_x emissions dominate the IAV of CNB ozone. We also explore the role of El Nino-Southern Oscillation (ENSO) in modulating the IAV of CNB ozone over southern China in spring, and find that the El Nino (La Nina) event has opposite effects on Southeast China and Southwest China.

Aerosol optical effects can trigger complex changes in solar shortwave radiation in the atmosphere, resulting in significant impacts on the photochemistry and vertical structure of ozone. This study provides observational evidence of aerosol absorbing and scattering effects on modifying the shortwave radiation and ozone profiles in the low troposphere. Using field vertical measurements and observation-based model simulations, we demonstrated that absorbing aerosols decreased shortwave radiation, resulting in substantial inhibition of ozone production throughout the boundary layer (BL). A similar inhibition effect occurred within the lower BL under sufficient scattering aerosols. However, the scattering augmentation effect played an additional role in enhancing the photolysis rate and promoting ozone generation in the upper BL. Hence, the observational evidence as well as our model simulations disentangled the radiative effects of different types of aerosols on the vertical structures of ozone.

Surface ozone (O₃) pollution has become a major public health concern in China. To accurately estimate the spatial-coverage O₃ from sparse ground-truth data, we propose a two-stage deep learning model that combines convolutional neural networks (CNN), deep neural network (DNN), and integrated gradients (IG). This model is able to monitor large-scale O₃ dynamics with greater temporal and spatial accuracy than that achieved by current models (R² = 0.78 and RMSE = 18.35 μg/m³). Using the integrated IG, we are also capable of interpreting the contribution of nearby cities to O₃ in a targeted site (i.e., Beijing) even under dust storms conditions, which traditionally limits model accuracy. In clean days, especially during summer when O₃ concentrations are often high, the surroundings have positive scores ranging from 0.05 to 0.11, indicating that these areas enhance O₃ formation. Conversely, during dust storms, the surrounding dust cells have negative IG scores, ranging from -1.43 to - 0.01, indicating that these areas inhibit the formation of O₃. This study provides a novel strategy to extract spatial dependence among predictors to estimate O₃ with high accuracy while obtaining spatial-coverage interpretation. It also contributes to further our understanding of O₃ pollution dynamics, and is applicable to the monitoring of other noxious air pollutants.

We present 3-year (2017-2019) tower-based measurement of lower boundary layer (up to 500m) ozone and examine its interaction with surface ozone from the Canton Tower in Guangzhou, the core megacity in South China with severe ozone pollution. Measurements of ozone, CO, NO₂, and meteorological parameters are available at 10m, 118m, 168m, 488m. We find that ozone concentrations increase with altitude, with ozone concentration higher by 1.8-3.3 times at 488m level than that at the surface. The diurnal cycle of ozone is consistent throughout the lower boundary layer, Regional air quality model (CMAQ) simulations show that chemical loss and dry deposition are the main sinks of surface ozone contributing to 55 ppbv·hr⁻¹, shaping the strong ozone vertical gradients. On ozone polluted days, the average ozone is higher by 1.5~2.2 times at 488m level than that on clean days. The ratio of nighttime to daytime ozone range from 69%~90% at 488m level, suggesting that a large proportion of daytime ozone can be stored in the nighttime residual layer. We also find significant positive correlation coefficients between 488m nighttime ozone and the following day surface MDA8 ozone, indicating nighttime ozone in the residual layer is a critical ozone factor in forecasting surface ozone in the following day. The tower-based measurements capture the ozone decreases at the 488m level with increases at lower altitude during the nocturnal ozone enhancement (NOE) event (defined as nighttime ozone concentration increases by more than 5ppbv in one of any two adjacent hours), supporting that the enhanced vertical mixing between the surface and the residual layer is an important contributing factor to NOE events. Our study highlights the value of long-term tower-based measurements for understanding the coupling between air pollution and boundary dynamics.

Atmospheric aerosols can influence ozone (O₃) by modulating photolysis rates (aerosol–photolysis interaction, API) and through meteorological feedback (aerosol–radiation feedback, ARF). This study attempts to assess the impact of API and ARF on surface O₃ and fine particulate matter (PM_2.5, aerodynamic diameter ≤2.5 µm) during a pollution episode in November-December 2018 in megacity Delhi, India using the Unified Inputs (initial/boundary conditions) for WRF-Chem (UI-WRF-Chem) model. The UI-WRF-Chem uses MERRA-2 data to provide both meteorological and chemical initial and boundary conditions for the regional WRF-Chem model. Our model reasonably well captured the observed diurnal variations of surface O₃ and PM_2.5 with a good correlation (r = 0.52–0.90) in megacity Delhi. API considerably lowers the surface photolysis rates (18–20%) and reduced surface O₃ concentration by 6% over the study region. The ARF shows a small increase in O₃ concentration (2%) and as a result, the combination of ARF and API contributes to an overall reduction in surface O₃. Aerosol-induced solar dimming (50 Wm^-2) due to ARF led to cooling (1 K) at the surface, suppresses the development of the planetary boundary layer (180 m), and further hinders the PM_2.5 dispersion, resulting in an enhancement in surface PM_2.5 concentration by 24% in megacity Delhi. Contrarily, the lower abundances of atmospheric oxidants caused by the API constrain the secondary aerosol formation, thereby lessening the ARF effects on PM_2.5. The combined effect of API and ARF results in a net increase (16%) of surface PM_2.5. The API and ARF have important implications on the mitigation of higher pollution loading in Delhi, particularly during the post-monsoon and winter seasons.

Halogen radicals (Cl, Br, and I) can strongly affect oxidation capacity as a product of photochemical reactions in atmosphere. Several Studies from 3-D chemical transport models and field observations have discovered the formation of nitrogen oxides and ozone through halogen radicals over a variety of regions and weather conditions. In addition, during KORUS-AQ campaign from 08 May to 12 June 2016, many studies measured not only nitrogen oxides and ozone but halogen species of nitryl chloride (ClNO₂) and chloride (Cl₂) as a precursor of chlorine radicals. However, current knowledge of halogen chemistries has not been fully understood due to insufficient study in South Korea. Therefore, this study aims to facilitate halogen chemistry including chlorine, bromine, and iodine species and evaluate the impacts of halogen chemistry on the atmospheric species during the KORUS-AQ campaign for the first time in South Korea. The modification was incorporated into the CMAQ modeling system as follows: i) preparing emissions (Cl₂, HCl, HBr, and Br₂ for anthropogenic source; BR₂, I₂, and HOI and other inorganic species from GOCI chlorophyll-a for natural source;) and ii) modifying/adding 139 gaseous, 7 aqueous, 18 heterogeneous reactions. In conclusion, impact of chlorine emission (5990Mg of HCl and 451Mg of Cl₂ in EXP1) enhances ozone (~1.6%) and OH mixing ratios (~5.4%), and reduces nitrate concentrations (~6.2%) in South Korea. The newly updated chlorine chemistry can capture ClNO₂ and Cl₂ mixing ratios (EXP2). Furthermore, ozone destruction and formation mechanisms exhibited by bromine and iodine chemistry in the ocean (~2.3% in EXP3) and updated bromine chemistry over land (~0.3% in EXP4), respectively. Collectively in this study, halogen emissions and modified halogen chemistry reproduced the levels of ClNO₂, nitrate, and ozone in the Korean peninsula and can help further understand halogen radicals in the CMAQ model simulation.

How anthropogenic forcing could change tropical cyclones (TCs) is a keen societal concern owing to its significant socio-economic impacts. However, a global picture of the anthropogenic aerosol and greenhouse gas effect on TCs has not yet emerged. Here we show that anthropogenic aerosol emission can reduce northern hemisphere (NH) TCs, represented by genesis potential index (GPI), but increase southern hemisphere (SH) TCs primarily through altering vertical wind shear and mid-tropospheric upward motion in the TC formation zones using the single anthropogenic forcing experiments from the 14 Coupled Model Intercomparison Project phase 6 (CMIP6) models. These circulation changes are driven by anthropogenic aerosol-induced NH-cooler-than-SH and NH-increased versus SH-decreased meridional (equator to mid-latitudes) temperature gradients. The cooler NH produces a low-level southward cross-equatorial transport of moist static energy, weakening the NH ascent in the TC formation zones; meanwhile, the increased meridional temperature gradients strengthen vertical wind shear, reducing NH TC genesis. The opposite is true for the SH. The results may help to constrain the models’ uncertainty in the future TC projection. Reduction of anthropogenic aerosol emission may increase the NH TCs threat. Further, we quantify the relative contributions of anthropogenic aerosol and greenhouse gas (GHG) to global TCF. We find that the two forcings have comparable but opposite impacts on GPIs due to their influences on the TC environment, leading to an insignificant change in GPIs in the historical period (1850-2014). Notably, the aerosol radiative forcing’s intensity is only about one-third of that of GHG, suggesting a more effective modulation of aerosol forcing on GPIs. The stable global TC frequency during the past decades could be attributable to the similar pace of the two anthropogenic emissions. The results highlight that a reliable global TC projection depends on both the aerosol and GHG emission policies.

Long-term historical meteorological observations are necessary for understanding climate variability. European and US ships sailed in the vicinity of Japan waters before the weather station network was established in Japan during the late Edo period in the eighteenth and nineteenth centuries. Many historical meteorological observation records are stored in libraries and repositories around the world and are slowly being recovered and used in weather and climate research. Such historical weather observations are being recovered worldwide through “Data Rescue” activities under the international ACRE (Atmospheric Circulation Reconstruction over the Earth) initiative. However, the spatial range of these weather data can be limited. In this study, we focus on the ship log weather records made on vessels sailing along Japan waters during this period. The oldest weather records in the vicinity of Japan were found to be recorded on the ship of the third expedition of James (Caption) Cook in 1779. During the eighteenth-century weather records came mostly from the Expedition cruises. In the nineteenth century, weather records were found from the US Navy and other ships came to Japan to open the country to the wider world. We focus on three tropical cyclone (TC) events in the vicinity of Japan during the period from 21 to 25 July 1853 observed by seven US Naval Japan Expedition of Perry’s fleet, on 23 and 24 September 1856 observed by Medusa of Dutch Navy ship, and on 15 and 16 August 1863 during the bombardment of Kagoshima in Japan observed by eleven UK Navy ships. Tracks of TCs are analyzed based on the ship log weather records.

The destructive potential of a tropical cyclone (TC) is primarily determined by its intensity and outer size. Although TC intensification has been researched extensively, the growth rate of its outer size remains obscure. This prompts us to develop the concept of rapid growth of outer size (RG) of TCs. RG is defined as an increase of at least 75 km in the gale-force wind radius within 24 hr using an objective anomaly detection algorithm. RG is intrinsically linked to the life cycle of the outer size and comprises most of the peak for large TCs (>300 km) in the distribution of lifetime maximum size. Compared with rapid intensification, RG is a more dangerous change in the TC structure, leveling up the destructive potential more rapidly. This is the first attempt to reveal the importance of RG to the outer size climatology, life cycle, and destructive potential of TCs.

Variabilities in tropical cyclone (TC) activity are commonly interpreted in individual TC basins. We identify an antiphase decadal variation in TC genesis between the western North Pacific (WNP) and North Atlantic (NA). An inactive (active) WNP TC genesis concurs with an enhanced (suppressed) NA TC genesis. We propose that the transbasin TC connection results from a subtropical east–west “relay” teleconnection triggered by Atlantic multidecadal oscillation (AMO), involving a chain atmosphere–ocean interaction in the North Pacific. During a negative AMO phase, the tropical NA cooling suppresses local convective heating that further stimulates a descending low-level anticyclonic circulation in the tropical NA and eastern North Pacific as a Rossby wave response, inhibiting the NA TC genesis. Meanwhile, the anomalous southwesterly to the western flank of the anomalous anticyclonic circulation tends to weaken the surface evaporation and warm the SST over the subtropical eastern North Pacific (southwest–northeast-oriented zone from the tropical central Pacific to the subtropical west coast of North America). The SST warming further sustains a cyclonic circulation anomaly over the WNP by local atmosphere–ocean interaction and the Bjerknes feedback, promoting the WNP TC genesis. This transbasin linkage helps us interpret the moderate amplitude of variations in TC genesis frequency in the Northern Hemisphere.

Most tropical cyclones (TCs) generated over the eastern North Pacific (ENP) do not make landfall. Consequently, TCs in this basin have received less attention, especially those that occur away from the mainland. Furthermore, there have been few studies of the climatic effects of ENP TCs. This study explores the feedback relationship between ENP TCs and the intensity of the El Niño–Southern Oscillation (ENSO), including El Niño and La Niña events, from the perspective of accumulated cyclone energy (ACE). Observational and modeling results indicate that the ENP ACE 3 months earlier can still affect the intensity of El Niño and La Niña events, although the SST persistence is main contributor. Thereinto, the impact of ENP TCs on El Niño appears to be approximately equal to that on La Niña. Moreover, this impact is independent of the persistence of the sea surface temperature (SST) in the Niño 3.4 region and the Madden–Julian Oscillation. Generally, the greater the ENP ACE, the stronger the El Niño, and the smaller the ENP ACE, the stronger the La Niña, this is especially the case for those TCs that develop over the July‒September period. In addition, results show that the ENP TCs modulate ENSO intensity by changing anomalous zonal wind at the low-level atmospheric layer. And the joint impacts of the low-level zonal wind anomalies on the Walker circulation and the east-west thermocline gradient lead to the time characteristics that ENP TCs lead ENSO intensity by about 3 months.

In addition to sea level rise, other effects of global warming have already been observed. The intensification of tropical cyclones (TC) is also no exception. This study shows how TCs will change in the future using MPI (Maximum Potential Intensity) theory, which estimates the maximum development of TCs for given climatological environmental conditions. This study presents future changes in TCs based on HighResMIP (High-Resolution Model Intercomparison Project). This projection focuses on the intercomparison of TC intensities, providing high-resolution data and estimating the effect of atmosphere-ocean coupling. The scenario-based projections show intensified TC trends in the monthly mean MPI in the Western North Pacific, although each mode of HighResMIP has different characteristics. There is also likely to be a spatial error between MPI and the maximum development value of TCs from track data of HighResMIP. MPI mostly depends on SST (Sea Surface Temperature). Therefore, it is important to reveal the north-south difference. By clearly quantifying this difference, we can accurately estimate TC intensity in each area over the long term and significantly contribute to predicting future changes in maximum storm surge risk. The detail of the results will be presented at the conference.

The effect of model resolution on the simulation of tropical cyclone (TC) landfall frequency in East Asia [including the South China Sea (SCS), Taiwan and coastal areas of East China (TWCN) and Japan (JP)] was investigated by comparing Atmospheric Model Intercomparison Project (AMIP) type simulations on the basis of 50-km High Resolution Atmospheric Models (HiRAMs) and 25-km HiRAM. The number of TC landfalls in the TWCN region was realistically simulated by the 50-km HiRAM ensemble model. However, fewer (more) TCs were steered westward (northward) toward the SCS (JP) because of an overestimation of the monsoon trough in the western North Pacific (WNP). The overestimation created a low-level cyclonic circulation anomaly in the WNP, which substantially modified steering flow. Consequently, more (less) TC made landfall in JP (SCS). The overestimation of the monsoon trough in model was primarily resulted from compounding factors, including the AMIP type simulation, upscale feedback of TCs to mean flow and the monsoon flow–topography interaction in the Indochina Peninsula Mountains and Philippine. First, the SST was negatively correlated with precipitation in the WNP during the typhoon season for the observation. Conversely, the SST–precipitation relationship was positive in the AMIP run. Second, the upscale feedback of TCs to mean flow (monsoon trough) was overestimated, which in term contributed to the overestimation of monsoon trough. Third, the model underestimated the mountain lifting effect in the Indochina Peninsula and Philippine. Overall, the aforementioned biases were substantially improved by increasing model’s horizontal resolution from 50-km to 25-km HiRAM.

Climate change due to greenhouse gases has fueled more powerful tropical cyclones (TCs). However, their characteristics and future changes of rainfall strength (RS) and rainfall area (RA) of TCs in regional scales are not fully revealed yet. Here, using ultra-high-resolution coupled simulations, we investigate the dominant factor in rainfall characteristics due to landfalling TCs in the North Indian Ocean (NIO) and western-North Pacific (WNP) and their future change in response to doubling and quadrupling atmospheric CO₂ concentrations. As CO₂ increases, RS increases more than RA in the NIO, but the opposite changes indicate in the WNP. We demonstrate that RS is highly related to the lifetime maximum intensity, landfall intensity, and latent heat flux (LHFLX), while RA mainly depends on LHFLX, relative humidity at 600 hPa, and vertical wind shear over the WNP. Our results suggest the need to establish regional-scale adaptation strategies for future rainfall change in landfalling TCs.

The Asian summer monsoon has been recognized in recent decades for its importance in modifying atmospheric composition based on data from modern Earth observing satellites. To understand the influence of Asian monsoon convection on the gas phase chemistry and aerosol loading in the upper troposphere and lower stratosphere (UTLS), a layer of significant climate sensitivity, airborne in-situ measurements are necessary. The ACCLIP campaign was motivated by these needs and was designed to obtain a large suite of trace gas and aerosol measurements in the Asian summer monsoon UTLS outflow over the Western Pacific. After a two-year postponement due to the COVID-19 pandemic, the campaign was successfully carried out during the summer of 2022, with operations based in Osan, South Korea. The campaign used two research aircraft, the NCAR Gulfstream V and the NASA WB-57, and it conducted 29 research flights with measurements ranging from 100 m above sea level to ~19 km altitude across a broad area of the western Pacific (15°N-43°N, 125°E-155°). The US led project received strong international collaboration, particularly from Asia, with multiple regional teams participating with airborne, balloon-borne, and ground-based measurements. This overview will focus on the trace gas measurements, highlighting the novel UTLS observations of species impacting ozone chemistry in the UT, very short-lived substances (VSLS) relevant for stratospheric ozone, and the species contributing to aerosol formation.

The Asian tropopause aerosol layer (ATAL) was thicker than other regions at the same latitude due to the strong confinement effect of the Asian summer monsoon anticyclone. The size distribution of the particles requires further measurements. Aerosol profiles were measured by balloon-borne sensors (Cobald, POPS) launched from Lhasa (29.66 °N, 91.14 °E), Golmud (36.48 °N, 94.93 °E), Kunming (25.01 °N, 102.65 °E) China, from 2019 to 2022 over the Tibetan Plateau at the part of the SWOP (Sounding Water vapor, Ozone, and Particle) campaign. The measurements combined with backward trajectories show that the volcano Raikoke (48°N, 153°E) in June 2019 and the dust storm in March 2021 over the Taklamakan desert have significantly impacted on the aerosol layer in the upper troposphere and lower stratosphere (UTLS). The backscatter ratio at wavelength 455 nm of the volcanic plume and dust storm was 0.1 higher than the ATAL. The particle number density in the volcanic plume is 30 cm^-3, higher than the ATAL and dust storm (10 cm^-3) in the lower stratosphere, with particle diameters centered around 0.42-3.4 μm. In contrast, the dust storm has a high density of up to 100 cm^-3 in the upper troposphere with particle diameters less than 0.42 μm.

We present our study on the intraseasonal and interannual variability of CO and aerosols in the upper troposphere (UT) that are the results induced by the variability of the Asian summer monsoon dynamics. We use the NASA global model GEOS simulations that incorporates emissions from anthropogenic, biomass burning, volcanic, and other natural sources to simulate CO and aerosols from 2000 to 2022 that are evaluated with satellite and aircraft observations. Model experiments separating source types (anthropogenic, biomass burning, volcanic) and source locations (East Asia, South Asia, and Southeast Asia) are used to identify the origin, trends, transport pathways, and spatial/temporal variabilities of CO and aerosols in the UT, and the meteorological data from NASA MERRA-2 reanalysis are used to assess the Asian summer monsoon anticyclone intensity and size associated with climate variability to understand the response of atmospheric composition to the ASM dynamics.

Global tight restrictions of the movement have severely affected civil aviation during COVID-19 lockdowns, resulting in noticeable reductions in aviation-related nitrogen oxides (NOx) emissions worldwide. As the formation and depletion of ozone (O₃) are mainly driven by its precursors volatile organic compounds (VOCs) and NOx, the emission reductions of NOx from aircrafts are expected to affect ozone concentrations at both the surface and upper troposphere. Given the different effects of tropospheric and stratospheric ozone on human health and climate, the potential changes in ozone concentrations at the surface and upper troposphere are of interest to us. In this study, we quantified the location- and time-resolved emission reductions of aviation-related NO_x due to the COVID-19 lockdowns based on open-access air traffic activity data. The Single Column Atmospheric Model Version 6 (SCAM6) was used to simulate the responses of tropospheric ozone to abrupt changes in aviation NO_x emissions in 2020. We shall discuss what these simulations reveal about the role of aviation emissions in atmospheric composition and global climate change.

This study aims to explore the impact of the Asian Monsoon on atmospheric ozone (O₃) and Carbon monoxide (CO) in the Western Pacific (WP) region, using measurements collected from August 2022 at the site of Koror, Palau (7.34°N, 134.47°E). The measurements of O₃ and CO are conducted using a solar absorption Fourier transform infrared (FTIR) spectrometer. O₃ and CO concentrations in WP are simulated by the GEOS-Chem global 3-D chemistry transport model. Balloon sondes of O₃ were launched two to four times per month in Koror to get the O₃ profiles from the surface to around 30 km. CO and O₃ measurements from different instruments and model simulation results are compared. The results will provide a deeper understanding of the relationship between the Asian Monsoon and atmospheric O₃ and CO in the WP region, using data from a specific time period. Comparing the measurements with simulations of the GEOS-Chem model offers insights into the dynamics of atmospheric composition and tests how well the current model’s understanding of WP is. Furthermore, this study provides an evaluation of the emission processes and source regions of O₃ and CO pollution in the WP region.

Winter cloudy days (CDs) in China exhibits strong spatiotemporal variability which has a large impact on agriculture, transportation and solar photovoltaic power industry. While the physical mechanism of CDs variability over China remains unclear, it is unknown what are the future changes of CDs under global warming. Here we reveal the spatiotemporal feature of the leading mode of winter CDs in China and its two independent formation mechanisms. The future changes of winter CDs under global warming are further projected using the optimal state-of-the-art models which are capable in simulation the two formation mechanisms in historical period. Results show that 1) the leading mode of winter CDs presents a homogeneous pattern over China. The positive CDs anomaly is related to the anomalous southerly wind over the western flank of the lower level North Pacific anomalous anticyclone and the southeastern flank of lower level Asian anomalous cyclone, which is related to the Eurasian Rossby wave train and the convection over Maritime Continent; 2) the optimal models projects a significant decrease of CDs in Tibetan Plateau and southern China, and nearly unchanged CDs over northern China under global warming; 3) The warming trend of surface air temperature in Arctic (around Barents Sea) is the possible reason for the decreasing winter CDs in Tibetan Plateau and southern China in the future.

Particulate matter (PM), which causes severe problems in human health, has become an important global issue in recent years. However, the climatology of annual variations and the interannual variations in PM level are still not fully evaluated. In our research, we find that the vertical motions of the East Asian monsoon system can affect the development of boundary layer height and then regulate the annual variation in PM over Taiwan. By understanding the annual variation in PM climatologically, the PM pollution season in Taiwan, from October to the following April, can be delineated. Then, we further define five phases of the PM pollution lifecycle that are similar to the well-defined East Asian summer monsoon lifecycle: onset (PM₁₀ onset date, PMOD), active (November to January, NDJ), break (between the end of January and early February), revival (February to April, FMA) and retreat (PM₁₀ retreat date, PMRD). After the precise definition of the PM pollution lifecycle, the interannual variation in PM level is clearer. Both the starting (PMODs) and ending (PMRDs) dates of PM pollution seasons are earlier during El Niño episodes than during La Niña episodes, in particular a significant 20-day difference between their starting dates. For the active phase (NDJ), climatological PM pollution development does not show distinct features under the two different El Niño-Southern Oscillation (ENSO) episodes. On the other hand, the influence of El Niño and La Niña on PM pollution during the revival phase (FMA) is significant. In summary, the climatology of PM pollution in wintertime is dominated by the annual seasonal cycle, but in the seasonal transition periods, October and March are significantly modulated by ENSO.

Secondary organic aerosols (SOAs) comprise a substantial portion of urban PM_2.5 concentrations, and can be formed through gaseous reactions of volatile organic compounds (VOCs). To mitigate urban PM_2.5 by reducing SOAs, this study investigates VOC-borne SOAs and explores the associated emission sources via a box chemical reaction model (PyCHAM), incorporating >110 measured VOCs. The VOC-borne SOAs (173.2±146.7 ng/m³) on average account for approximately >20% of total SOAs in PM_2.5. Anthropogenic (AVOCs) and biogenic VOCs (BVOCs) account for ~85% and ~7% of the total VOC-borne SOAs, respectively. Synergistic effects of both AVOCs and BVOCs lead to ~10% of the formed SOAs. We investigated the potential overestimation of biogenic SOAs (BSOAs) due to the inclusion of VOCs that are likely emitting from volatile chemical products (VCPs), such as fragrance chemicals in household cleaning products. Excluding six potential VCP-borne VOCs decreases the BSOAs by more than 60%. This indicates anthropogenic SOAs (ASOAs) could account for >85% of total VOC-borne SOAs with lesser contribution from BVOCs in the warm humid tropical urban environments.

Atmospheric black carbon (BC) has a large yet highly uncertain contribution to global warming. When mixed with non-BC/coating material during atmospheric aging, the BC light absorption can be enhanced through the lensing effect. Laboratory and modeling studies have consistently found strong BC absorption enhancement, while the results in ambient measurements are conflicting, with some reporting weak absorption enhancement even for particles with large bulk coating amounts. Here, from our direct field observations, we report both large and minor absorption enhancement factors for different BC-containing particle populations with large bulk non-BC-to-BC mass ratios. By taking insights into the measured coating material distribution across each particle population, we find the level of absorption enhancement is strongly dependent on the particle-resolved mixing state. Our study shows that the greater mixing-state heterogeneity results in the larger difference between observed and predicted absorption enhancement. We demonstrate that by considering the variability in coating material thickness into the optical model, the previously observed model-measurement discrepancy of absorption enhancement can be reconciled. The observations and improved optical models performed here highlight the importance of mixing-state heterogeneity on BC’s radiative forcing, which should be better resolved in large-scale models to increase confidence when estimating the aerosol radiation effect.

Ammonia (NH₃) emission reduction has been advocated for its potential to mitigate PM_2.5 air pollution, yet emission quantifications at city levels are limited. Here we develop high-resolution (3 km) bottom-up emission inventories of agricultural NH₃ in the Beijing-Tianjin-Hebei (BTH) region and traffic NH₃ in Beijing for the year 2016. Then the WRF-Chem model is used to evaluate the NH₃ and PM_2.5 concentrations against ground-based and satellite observations. Our estimated annual BTH agricultural NH₃ emissions (625 Gg) and Beijing’s traffic emissions (7.8 Gg) are within the ranges of published inventories. However, simulated NH₃ concentrations are significantly lower than observations during August in urban Beijing, nevertheless wintertime underestimations are much more moderate. Further evaluation and sensitivity experiments show that biases in meteorology or regional transport cannot explain such discrepancies. Using measurements as constraints, our inversed NH₃ inventory indicates both agricultural and non-agricultural NH₃ emissions in Beijing during August should increase by ~5 times to match NH₃ and PM_2.5 observations. Current underestimations may stem from the missing power sector, urban green space emissions, the lack of representation of industrial hotspots, and uncertainties in traffic emissions. Our study highlights that denser and more frequent urban NH₃ observations are urgently needed to constrain and validate bottom-up inventories.

Air pollution constitutes one of the major global threats to both the environment and human health. Pollutant emissions from various agricultural activities cause significant environmental impacts are highly understudied and call for the preparation of a detailed emission inventory. Emission inventories are a useful tool for comprehending the sources of pollutant emissions and their overall effects on the environment. Most nations and regions do not include entire agricultural operations during the process of developing an emission inventory (EI) for their agricultural sector. Additionally, there hasn't been enough work done to improve our understanding of agricultural emissions’ effects on the ecosystem. Listing emissions by source and quantity for each tracked pollutant enables defining their primary emission sources. However, comprehensive datasets associated with emissions and agricultural activities are often missing. The lack of data is mainly because of the false perception that the agriculture sector emits much smaller pollutant(s) quantities compared to other sources. Data non-availability impedes the ability to develop appropriate policies (that could lower the emissions), leading to decreased interest for collecting relevant data by pertinent authorities which consequently results in less research on agricultural emissions. Instead of using models based on laboratory research, gathering activity data (such as fertilizer application, crop burning, equipment usage, pesticides, livestock management, tillage, etc.) from field operating conditions helps in determining the actual real-world agricultural emissions and their significant sources. Large deficits in activity data causes high degrees of uncertainty in the results of EI. This article identifies and delineates the influencing parameters to develop an EI for the agriculture sector of a region. The requirement for the collection of apposite data, types of influencing data, data sources, the parametric format in which data must be collected, and how each parameter is attributable to the emissions are elucidated in detail.

To address the issue of poor air quality in India, a number of emission reduction scenarios were devised, each specific to a certain source. However, many of these scenario-based model simulations overlooked the impact of changes in Land Use Land Cover (LULC) and meteorology on air quality. The study endeavors to assess the influence of these two factors on concentrations of Short-Lived Climate Pollutants (PM_2.5, OC, and BC). The simulations were performed using the WRF-Chem V3.8.1 model for the years 2019 and 2024, with emissions remaining constant (2019 emissions) for both years, and only the LULC and meteorology being altered. The results indicate that greater emission reductions are necessary in light of the effect of LULC and meteorology, to achieve the clean air goals set forth in India's national action plan.

This study aims to quantify the contributing factors to changes in winter PM_2.5 concentrations in South Korea from 2019 to 2021. We consider various factors, such as meteorological variability, reduced anthropogenic emissions in China, and South Korea's Seasonal Particulate matter Management (SPM) plan. To determine the individual impact of these factors on PM_2.5 concentrations, we used nested versions of GEOS-Chem and updated the anthropogenic emissions with observational data. Our simulations captured well the spatial distribution of observed surface PM_2.5 concentrations in China and South Korea. Our findings indicate that meteorological variability was the most significant factor affecting PM_2.5 concentrations over the past three winters, despite its high level of monthly variability. Furthermore, reductions in anthropogenic emissions in China and the implementation of South Korea's SPM also contributed to changes in PM_2.5 concentrations. Since various factors have a complex effect on changes in winter PM_2.5 concentration in South Korea, the methodology used in this study will contribute to establishing effective environmental policies in the future.

The air quality in the Indian subcontinent has been a growing concern in recent years, with particulate matter (PM) being one of the major pollutants. PM_2.5, in particular, has a significant impact on human health and the environment as it can penetrate deep into the respiratory system. PM_2.5 has been linked to several health issues including respiratory and cardiovascular disease, while Water-Soluble inorganic Ions (WSII) contribute to the acidity and salinity of the air. In this study, we aim to investigate the seasonal variation and contributing sources of PM_2.5 and 13 associated WSII (Na+, NH⁴⁺, K⁺, Mg²⁺, Ca²⁺, Li⁺, F^-, Cl⁻, Br^-, NO₂^-, PO₄^3-, NO₃⁻ and SO₄²⁻) in two non-attainment Indian cities, Alwar and Amritsar. The study regions are selected owing to the unique meteorological conditions, population density and industrial activities. The study employs a combination of field measurements and comparative analysis to understand the sources and seasonal patterns of PM_2.5 and WSII in these cities. Initial analysis of PM_2.5 winter samples shows Nitrate (22.078 μg/m³) and Sulphate (17.408 μg/m³) to be the dominant anionic species and Ammonium (13.046 μg/m³) and Sodium (5.452 μg/m³) to be dominant cationic species for both day and night respectively in Alwar city. The total average day anionic concentrations for the same period were observed to be 31.37 μg/m³ and night concentrations to be 52.52 μg/m³ with total observed average day cationic concentrations to be 15.23 μg/m³ and night concentrations to be 23.29μg/m³.

Models underestimated ozone, formaldehyde, peroxyacetyl nitrate (PAN) and alkyl nitrate (AN) concentrations in Seoul during the 2016 joint NIER/NASA Korea United States-Air Quality (KORUS-AQ) field study. This indicates needed improvements in emissions and chemistry of volatile organic compounds (VOCs) in a region where photochemistry is radical-limited. Total peroxynitrates (PNs) were twice as large as PAN during the campaign, in contrast to previous studies where missing PNs were <20%. Observations of speciated ANs could only explain 3% of total ANs. This points to the need to evaluate model VOC emissions and chemical mechanisms producing ANs and PNs. We find that model emissions severely underestimate PAN-precursors such as ethanol and methyl ethyl ketone, which contribute approximately 30% and 10% to the PAN budget, respectively. These species may be underestimated from vehicles, or from non-combustion sources such as solvents or cooking. We use the GEOS-Chem chemical transport model to investigate the impacts of improvements in VOC emissions and chemistry on ozone. Scaling emissions of individual VOC species based on KORUS-AQ observations increases surface ozone by ~10 ppb in Seoul, while addition of chemistry to produce missing ANs and PNs reduces ozone in Seoul by ~1 ppb but results in transport of ozone precursors to downwind regions.

The relationships between emissions and PM2.5 pollution are governed by nonlinear chemistry. Therefore, understanding the role of nonlinear chemistry in the PM2.5 formation pathways is critical for policy-making. Here we analyze the sensitivity of PM_2.5 in Beijing to emissions from different precursor species and different regions using the adjoint of the GEOS-Chem chemical transport model, with a focus on the changes in sensitivity in response to precursor emission changes of COVID-19 lockdown and 2013-2017 control policies. This allows us to diagnose the direct effects versus nonlinear chemical effects and analyze the differences and commonalities between long-term and short-term emission reductions. We find significant changes in the sensitivities for secondary inorganic aerosols in response to emission reductions. In the absence of such nonlinear chemical effects, the emission reductions of COVID-19 lockdown and 2013-2017 control policies would reduce the February mean PM_2.5 by 10.7 μg/m³ and 63.6 μg/m³ respectively in Beijing. However, an increase of 6.3 μg/m³ occurred for each emission reduction scenario due to the increased sensitivities of nitrate to NO_x and NH₃ emissions, reflecting changes in atmospheric oxidation capacity and gas-particle partitioning. These chemical effects accelerate aerosol formation and offset the effectiveness of emission reductions. Overall, our study highlights the importance of considering nonlinear chemical effects when assessing the effectiveness of emission reduction measures for highly polluted areas.

The influence of the intraseasonal Indo-western Pacific convection oscillation (IPCO) on the absence of typhoons in July 2020 over the western North Pacific (WNP) was explored. While observation analysis shows that necessary conditions such as sea surface temperature (SST) and vertical wind shear in July 2020 meet the basic requirement of or even are conducive to the formation of typhoon, the unprecedented absence of typhoon over the WNP occurred in July 2020, and it is the first time that no typhoon in July since 1951. Additionally, significant differences were found in the number of typhoons in July between the different phases of the intraseasonal IPCO, and the number in the positive phase of the intraseasonal IPCO was significantly higher than that in the negative phase of the intraseasonal IPCO. In July 2020, the intraseasonal IPCO was in a strong negative phase, with the third lowest index in history and had the strongest inhibition effect on convection over the WNP on record, leading to large-scale circulation anomalies. The strongest descending movement on record inhibited the upward transport of water vapor and the development of cumulus convection, thereby reducing the release of latent heat of condensation and making it difficult to form a typhoon warm-core structure. In addition, the geopotential height increased over the WNP, and the western Pacific subtropical high moved southerly, which inhibited typhoon formation. Simultaneously, the South China Sea monsoon trough weakened significantly, with increased negative vorticity anomaly in the response scale, which hindered disturbance generation. The lowest genesis potential index confirmed that the large-scale circulation anomaly caused by the intraseasonal IPCO had an unprecedented restraining effect on typhoon generation, leading to the absence of typhoons over the WNP in July 2020.

A large body of observational and modelling studies has pointed out that two-way interaction of the Pacific Ocean (PO) with regions outside of the PO, such as the tropical Atlantic (AO) and Indian Ocean (IO), plays an important role in tropical climate variability and predictability, and multiple interactions among the three basins are now known as tropical interbasin interaction (TBI). Here, the impact of TBI upon characteristics and predictability of sea surface temperature (SST) in the tropics are assessed by applying a linear inverse modelling (LIM) framework that uses SST and sea surface height anomalies in the tropical Pacific (PO), Atlantic (AO), and Indian Ocean (IO). The TBI pathways are shown to be successfully isolated in stochastically-forced simulations that modify off-diagonal elements of the linear operators. The removal of TBI leads to a substantial increase in the amplitude of El Niño-Southern Oscillation (ENSO) and related variability. Partial decoupling experiments that eliminate specific coupling components reveal that PO-IO interaction is the dominant contributor, whereas PO-AO and AO-IO interactions play a minor role. A series of retrospective forecast experiments with different operators shows that decoupling leads to a substantial decrease in ENSO prediction skill especially at longer lead times. The relative contributions of individual pathways to forecast skill are generally consistent with the results from the stochastically-forced experiments. We will also discuss some typical examples that highlight the importance of TBI on regional SST variability.

Synoptic-scale disturbances are prominent over the tropical western North Pacific during boreal summer and they play an important role in the weather and climate of East and Southeast Asia through causing extremely heavy rainfall events, seeding tropical cyclone genesis and interacting with low-frequency activities. Those disturbances are generated over the equatorial western-central Pacific and propagate northwestward to the tropical western North Pacific. Thus, understanding the changes and factors of synoptic-scale disturbance intensity over the tropical WNP is an important topic. The intensity of synoptic-scale disturbance is closely related to the El Niño–Southern Oscillation that modulates the seasonal atmospheric fields over the source regions, along the propagation paths, and over the impact regions of the synoptic-scale disturbances. In this talk, we will present analysis of interannual variations of synoptic-scale disturbance intensity over the tropical western North Pacific during boreal summer and the associated factors. Evidences for the asymmetric and nonlinear response of synoptic-scale disturbance intensity to El Niño and La Niña events will be presented along with physical explanations based on observational analysis and numerical model experiments.

This study examines the climate response to a sea surface temperature (SST) warming imposed over the southwest Tropical Indian Ocean (TIO) in a coupled ocean-atmosphere model. The results indicate that the southwest TIO SST warming can remotely modulate the atmospheric circulation over the western North Pacific (WNP) via inter-basin air-sea interaction during early boreal summer. The southwest TIO SST warming induces a “C-shaped” wind response with northeasterly and northwesterly anomalies over the north and south TIO, respectively. The northeasterly wind anomalies contribute to the north TIO SST warming via a positive Wind-Evaporation-SST(WES) feedback after the Asian summer monsoon onset. In June, the easterly wind response extends into the WNP, inducing an SST cooling by WES feedback on the background trade winds. Both the north TIO SST warming and the WNP SST cooling contribute to an anomalous anticyclonic circulation (AAC) over the WNP. The north TIO SST warming, WNP SST cooling, and AAC constitute an inter-basin coupled mode called the Indo-western Pacific ocean capacitor (IPOC), and the southwest TIO SST warming could be a trigger for IPOC. While the summertime southwest TIO SST warming is often associated with antecedent El Niño, the warming in 2020 seems to be related to extreme Indian Ocean Dipole in 2019 fall. The strong southwest TIO SST warming seems to partly explain the strong summer AAC of 2020 over the WNP even without a strong antecedent El Niño.

The climate systems over the Indian and Pacific oceans interact with each other at the interannual time scales. The Dipole Mode Index(DMI) is found to lead the Niño3 index by more than one year. Traditional understanding of this precursory relationship is thought through the atmospheric bridges, a.k.a. the Walker Cell variability. Latest studies suggest that the oceanic channel process, i.e. the Indonesian Throughflow (ITF) variability, plays a dominant role in the inter-basin interactions, which quickly emerges as a hot research topic. However, due to the concurrent ENSO and Indian Ocean Dipole (IOD) events in history, the effects of an IOD on the evolution of ENSO are contaminated by the repercussions of the ENSO. In 2019, a strong IOD took place in the Indian Ocean, with the tropical Pacific in a neutral state throughout that year, which provides an opportunity to test the effectiveness of the oceanic channel dynamics. A strong La Niña event indeed took place at the end of 2020, the dynamics of which are investigated using the China National Marine Environmental Forecasting Center (NMEFC) operational seasonal forecasting system. The strong subsurface cooling in the Indian Ocean is found to propagate to the eastern equatorial Pacific Ocean through the Indonesian seas and induce a strong La Niña event at the end of 2020, suggesting the dominance of the oceanic channel in the inter-basin forcing leading to the outburst of the 2020/2021 La Niña. In comparison, experiments initialized with only surface temperature anomalies over the tropical Indian Ocean show that the atmospheric bridge alone is unable to induce the onset of the 2020/2021 La Niña. Forecasting experiments of historical ENSO events with and without the IOD initializations suggest that both the delayed feedback of ENSO and the Indo-Pacific oceanic channel dynamics are important in forecasting the ENSO.

Considering that the subtropical highs and tropical convections are observed as negative and positive vorticities respectively, the large-scale features of the atmospheric environment can be effectively represented using streamfunctions as defined by the Laplacian. By investigating the geographical patterns of streamfunctions from different modes of environmental variability, this study conceptualizes how the subtropical high expands and the region for tropical convections migrates in the western North Pacific. It is confirmed that, owing to the expansion of the subtropical high, the limited ocean area for tropical convections even bounded by the equator becomes narrower in the “La Niña mode” than that in the “El Niño mode”. This study finds that a warmer environment is likely to further expand the subtropical high to the west, and then the westernmost shift in the region for tropical convections appears in the “warmer La Niña mode”. A linear perspective suggests that every warmer La Niña environment could be one that people have scarcely experienced before.

In this work we examine a methodology to study the origin of systematic circulation biases associated to the mean-state West Pacific subtropical high (WPSH) in the Met Office Unified Model (MetUM). MetUM exhibits robust biases, including a weakening of the anticyclone and a location too far east, which leads to an underestimation of the southwesterly monsoon flow over East Asia and contribute to seasonal precipitation errors in the area. We study the development of the errors in an ensemble of equivalent NWP hindcasts and, using a semigeotriptic (SGT) balance model, we link the circulation errors to physical processes. With this tool we show that most of the circulation errors in the WPSH are corrected when tropical convection occurs in the right location. We then examine how MetUM deep convection biases in the region arise and find a large drying of the boundary layer by convection that is balanced mainly by local surface fluxes. In places with low exchange coefficient (places with light surface winds), the surface fluxes are not able to support deep convection over a long time and the convection error is established.

The El Niño-related anomalous western North Pacific anticyclone (WNPAC) shows different latitudinal extensions during the El Niño decaying summer, which determines the moisture transport to different regions and leads to distinct climate impacts over East Asia. It is known that both the north Indian Ocean (NIO) sea surface temperature (SST) and the tropical North Atlantic (TNA) SST can generate a WNPAC in summer. However, the difference between the NIO SST-forced WNPAC and the TNA SST-forced WNPAC has hardly been noted before now. This study shows that the NIO SST warming makes the WNPAC contract southward, whereas the TNA SST warming makes the WNPAC extend northward. The NIO SST warming generates the WNPAC via a Kelvin wave response. Owing to the limited domain of Kelvin wave activity, the Kelvin wave-induced suppressed convection over the western Pacific is confined south of 20°N, resulting in the WNPAC being concentrated in the low latitudes. In contrast, the TNA SST warming generates the WNPAC via a Rossby wave-induced divergence/convergence chain response over the Pacific. The Rossby wave-induced suppressed convection over the central-eastern Pacific north of the Equator leads to enhanced convection on its southwest side, which further generates the low-level anomalous divergent winds over the western North Pacific and suppresses convection there. In this process, the suppressed convection over the western North Pacific is pushed more northward, thus producing a WNPAC extending northward. Further study finds that there are good precursors for predicting the WNPAC latitudinal extension based on the El Niño spatial pattern and the NIO/TNA SST intensity in the previous winter and spring.

Surface ozone forecast accuracy is limited by weather forecast uncertainties, which are not quantitatively represented in current air quality forecast systems. We developed an ensemble surface ozone forecast (2DCNN-ESOF) system using 2-D convolutional neural network and ensemble weather forecasts, and we applied the system to 216-h ozone forecasts in Shenzhen, China. The 2DCNN-ESOF’s skills were comparable to or better than other current operational systems and fulfilled the Chinese mandate for air quality level forecast accuracies up to 144-h lead time. Additionally, the 2DCNN-ESOF enabled an “ozone exceedance probability” forecast given the range of possible weather outcomes. Half of the ozone forecast errors were due to weather forecast uncertainties, which would induce a 7.4 ± 1.4 μg m^-3 uncertainty in forecasted ozone concentrations at 24-h lead time even with perfect emission estimates and chemical mechanisms. Our ensemble forecast framework can be applied to the operational forecasts of other meteorology-dependent environmental risks.

Machine learning has become a powerful tool to establish models for predicting the future PM_2.5 concentration, which can help the policy makers to take control measures to protect on human health and promote sustainable urban development. The spatial information of adjacent sites can well reflect the regional pollution pattern, however it has not been widely investigated in most existing prediction models. In this study, a novel hybrid model is used to select appropriate adjacent site information and add it into the deep learning prediction model as additional features. This hybrid model combined long and short-term memory neural network (LSTM) and a convolutional neural network (CNN) of 1 * 1 specific convolution kernel to solve the challenges of time series data prediction and discrete data aggregation. The experimental results showed that the novel hybrid model presented the highest prediction accuracy and the lowest error (R²=0.92-0.94, RMSE=8.54-8.93μg/m³, MAE=5.76-6.52μg/m³) compared with single models (R²=0.81-0.92, RMSE=9.01-13.24μg/m³, MAE=6.79-9.34μg/m³). The comparison with other similar research sets shows that the deep learning model significantly improves the ability to capture peak PM_2.5 concentration by adding information from appropriate adjacent sites.

The study utilized a Recurrent Neural Network architecture with Long short-term memory to predict PM_2.5 levels in Delhi. Four different models were developed, each incorporating different combinations of data inputs. The models that utilized PM_2.5 data alone or in conjunction with gaseous pollutants exhibited the best performance in predicting PM_2.5 concentrations during both hourly and daily forecasts. The models were also evaluated for their performance during specific events characterized by a drastic increase in PM_2.5 concentrations. The evaluation revealed that the models that incorporated gaseous pollutants displayed superior performance compared to others during such episodes. The results demonstrated that the models were effective in providing accurate predictions, but only for forecast windows limited to twelve hours for hourly forecasts or seven days for daily forecasts. Due to the difficulty in predicting gaseous pollutants, it is more practical to employ models that only utilize PM _2.5 data or PM _2.5 and meteorological information. The outcomes of this research can serve as decision-making tools for regulators seeking to implement timely interventions in heavily polluted cities, thus imparting a sense of assurance in the potential of these models.

Air quality is a pressing issue worldwide, with 90% of people exposed to unhealthy air. Chemical transport models are frequently used in the absence of dependable monitoring networks. These models integrate meteorology, emissions inventory, and atmospheric chemistry to forecast air quality. However, due to the high computational power necessary for each operation, such systems may not be the most efficient or effective early warning systems. An alternative to traditional modeling is Machine Learning models, which require little time and computational resources after being trained. Physics-based machine learning is a developing field that combines Partial Differential Equations, simulated data, and models in a data-driven approach that penalizes physically inconsistent results. Another way to utilize Machine Learning is by parameterizing a physical model. In our study, we utilized 4 years of simulated WRF-Chem data over the Indian subcontinent to apply a physics-based machine learning approach to improve upon the existing model. Our findings reveal that the trained model is efficient and performs well.

The simultaneous retrieval of aerosol optical/microphysical and surface parameters have been a challenging work due to the limited information of satellite observations. The common optimized methods with iterative Radiative Transfer (RT) calculations are usually time-consuming and have to retrieve all the unknowns with different information content. In this study, we developed a flexible aerosol algorithm framework for satellite measurements based on physical-based deep learning (PDL) method. By pre-training of RT simulations, all the unknown aerosol parameters can be retrieved independently by modeling their respective function with the whole satellite observations. Moreover, our method can utilize the abundant priori information such as existing surface and aerosol products, which can also provide an effective constraint for very abnormal values. By applying the PDL algorithm to satellite measurements of multi-spectral, multi-angle, and polarization such as MODIS, MISR, and POLDER-3, the retrieval results have robust high-accuracy compared with AERONET products. With a high efficiency in both computation and information utilization, the PDL algorithm with flexible framework gives a competitive selection for operational aerosol retrieval with various satellite measurements.

The aerosol fine-mode fraction (FMF) is valuable for discriminating natural aerosols from anthropogenic ones. However, most current satellite-based FMF products are highly unreliable over land. Here, we developed a new satellite-based global land daily FMF dataset (Phy-DL FMF) by synergizing the advantages of physical and deep learning methods (Figure 1) at a 1° spatial resolution covering the period from 2001 to 2020. It was extensively evaluated against AERONET FMF retrievals, revealing its higher accuracy (RMSE= 0.136) based on 361089 validation samples; 79.15% of the data fell within the ±20%EE envelope) and generally good agreement with AERONET FMF with respect to its values, trends, and frequencies. Phy-DL FMF showed superior performance over alternative deep learning or physical approaches (such as the spectral deconvolution algorithm presented in our previous studies), particularly for forests, grasslands, croplands, and urban and barren land types. By examining Phy-DL FMFs from 2001 to 2020, we found a general decreasing trend around the globe, which was not revealed by AERONET point-scale measurements. However, both Phy-DL and AERONET FMFs showed significant increasing trends in FMF over the western USA and India. The new dataset captured high-level FMFs (> 0.80) over southern China, South Asia, eastern Europe, and the eastern USA. The FMFs were consistently < 0.3 in Northwest China, the Saharan region, and southern South America, indicating coarse-particle desert emissions. The findings of various evaluations, especially the attempted explanations of the spatiotemporal variations and long-term trend changes, suggest that this newly developed dataset is sound, more accurate and thus useful for investigating the impact of fine-mode and coarse-mode aerosols on the atmospheric environment and climate, especially in gaining a deeper insight into fine-mode aerosols. The datasets can be downloaded from https://doi.org/10.5281/zenodo.5105617.

Atmospheric aerosols play a crucial role in the Earth's climate system by affecting radiative forcing, cloud formation, and precipitation patterns. Air quality in Asia is a growing concern due to the region's rapid industrialization and urbanization. Anthropogenic activities such as burning fossil fuels and deforestation can release large amounts of aerosols into the atmosphere, which can have negative impacts on air quality and human health. The Geostationary Environment Monitoring Spectrometer (GEMS), onboard GEO-KOMPSAT-2B (GK-2B) satellite, is the first air quality monitoring sensor in geostationary earth orbit launched in 2020. GEMS measures the hyperspectral radiances with 0.6 nm spectral resolution in ultraviolet and visible ranges over the Asia region during the daytime to provide hourly air quality information. We have updated the aerosol retrieval algorithm based on the optimal estimation method for GEMS. The aerosol retrieval algorithm for GEMS uses 6 channels in ultraviolet and visible wavelengths, which have the advantage of measuring aerosol absorption and height information. We present GEMS aerosol retrieval results for high aerosol loading cases over Asia. The results show that the GEMS AOD has the advantage of allowing retrieving of aerosol over bright surfaces due to dark surface reflectance in UV. The GEMS AOD and SSA are validated against ground-based AERONET. The GEMS ALH is compared to data from the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP). The GEMS aerosol product is in good overall agreement with ground measurement and satellite data products. In addition, we conducted an error analysis of GEMS aerosol optical properties. The retrieval can have errors associated with various factors. Sun–earth–satellite geometry, and assumed aerosol model, clouds, and other atmospheric effects. Conducting an error analysis of aerosol optical properties is essential to identify the major contributors to errors and to improve the accuracy of the retrieval algorithms.

This study develops an improved algorithm for retrieving atmospheric columnar aerosol components from optical remote sensing data. This is achieved by using the complex refractive index (CRI) of a multi-component liquid system in the forward model and minimizing the differences with observations. The aerosol components in this algorithm comprise five species combining eight sub-components including black carbon, water-soluble and water-insoluble organic matter, inorganic salt (ammonium nitrate, AN), sea salt (sodium chloride, SC), dust-like (DU), and aerosol water content in fine and coarse modes (AWf and AWc). The calculation of the complex refractive index (CRI) in the multi-component liquid system allows the separation of the water soluble components (AN, WSOM and AWf) in the fine mode and the sea salt (SC) and water content (AWc) in the coarse mode. The uncertainty in the retrieval results is analysed based on the simulation of typical models, showing that the complex refractive index (CRI) obtained from instantaneous optical-physical inversion compares well with that obtained from chemical estimation. The algorithm is not only used for ground-based remote sensing, but is also transferred to satellite inversion algorithms to obtain larger area-wide observations. Based on the POLDER observations, the study inverted the spatial distribution of aerosol optical, microphysical and chemical parameters in the North China region and made a preliminary comparison with ground-based observations. The comparison shows that the column mass concentration of aerosol black carbon mass concentration is consistent with that observed on the ground, and the MAE is 0.38. The mass ratios of black carbon, inorganic salt and organic aerosol (8%, 49%, 43%) are in good agreement with the mass ratios of the ground chemical sampling analysis results (11%, 49%, 40%).

Current aerosol remote sensing without using multi-view satellite sensors still fails to retrieve aerosol components directly, such as black carbon (BC) and organic carbon (OC) simultaneously due to the limitations on available observations. In this study, A new general carbonaceous aerosol retrieval strategy for geostationary single-view satellite observations is implemented based on a method of critical reflectance, to retrieve the parameters of BC and extra OC concentration from measured radiance without prior quantification of Aerosol Optical Thickness (AOT). Random sampling consensus (RANSAC) is also used in the strategy to reduce the influence of clouds and anomalous pixels on the retrievals. An initial validation and application applied to Himawari images over the North China Plain (NCP) show that the carbonaceous aerosol retrievals are highly consistent with the expectations in terms of spatial and temporal patterns. The retrievals of BC and OC concentration follow the daily fluctuations of the Aethalometers observations and feature small differences in mean values throughout the month. Additionally, the AOTs at 0.55 µm can be also satisfactorily reproduced based on the carbonaceous aerosol retrievals, resulting in a mean absolute error of 0.089 and a correlation coefficient of 0.870. The main errors in the method arise from shell model assumptions, RANSAC fitting bias, inconsistent clear reference AOT in LUT, and geometric inconsistency between the clear reference and hazy images. This work indicates that the BC and OC concentration in high resolution can be acquired through the geostationary single-view remote sensing for further air quality and climate studies.

For the last few years, South Korea has launched two geostationary earth orbit satellites, Geo-Kompsat-2A (GK-2A) and Geo-Kompsat-2B (GK-2B). GK-2A carries a single sensor named Advanced Meteorological Imager (AMI), which has spectral bands from visible to infrared for observations of weather variables such as clouds. GK-2B carries two sensors. One of them is an ultraviolet-visible hyperspectral spectrometer named Geostationary Environment Monitoring Spectrometer (GEMS), and the other is a band (visible-shortwave infrared) observing ocean color imager named the 2nd Geostationary Ocean Color Imager (GOCI-II). In the aspect of aerosol optical properties products, each geostationary instrument has its own instrument specification. The visible radiance capability of AMI and GOCI-II is sensitive to the size of the airborne aerosols, while ultraviolet observation by GEMS is sensitive to the light-absorption of aerosols. Moreover, a dark surface in the ultraviolet spectrum enables aerosol retrievals over arid regions. In this study, we present a statistically fused product of three aerosol optical depth (AOD) from AMI, GEMS, and GOCI-II that are retrieved by individually optimized algorithms. Because of the algorithm characteristics such as surface reflectance estimation and observation geometry, each AOD product has its own systematic retrieval bias according to surface vegetation and observation time. Therefore, we proceed with a bias correction based on comparison with Aerosol Robotic Network (AERONET) measurements before the AOD fusion. Statistical fusion uses Maximum Likelihood Estimation (MLE) method for the sake of weighting based on retrieval errors evaluated with AERONET measurements. We analyzed satellite and AERONET data of 2021 to calculate bias and retrieval error. AOD fusion algorithm was applied to the individual AOD products in 2022, and validation result showed that individual error characteristics were much improved by statistical aerosol fusion.

Recently, air pollution has been worsening globally and, in particular, more severe in Asia. South Korea launched and has been operating the first geostationary environmental satellite, Geostationary Korea Multi-Purpose Satellite (GK)-2B, to monitor the atmospheric environment since February 2020. The Geostationary Environment Monitoring Spectrometer (GEMS) onboard the GK-2B has spectral bands within 300~500nm from ultraviolet (UV) to visible (VIS) wavelength ranges specialized in observing micropollutants in the atmosphere. Like human eyes, the three red-green-blue (RGB) bands are helpful for monitoring and intuitively understanding the atmospheric environment, including the movement, diffusion, and extinction of clouds and aerosols. However, the GEMS cannot provide RGB true-color images because the GEMS has hyperspectral bands within only the blue band range. Notably, the central wavelength of the Advanced Meteorological Imager (AMI) blue band corresponds to the GEMS blue band. Thus, this study presents a deep-learning-based method to simulate virtual radiances at the virtual GEMS RGB bands similar to the neighboring AMI sensor with the real RGB bands onboard the GK-2A satellite. We adopted the data-to-data translation method using the AMI three RGB bands as training and test datasets. This study used AMI data from 2020 to 2021 for model development and validation. The GEMS data during 2021 were used for model application as input data. As a result, the proposed model showed excellent results with a high correlation coefficient of over 0.95 in all AMI RGB bands, in addition to qualitative agreements between the observed AMI RGB and model-generated GEMS RGB data. This study can significantly contribute forecasters to monitoring and intuitively understanding the atmospheric environment in Asia.

Tropical cyclone (TC) motion determines areas affected by the TC, and has a consequential effect on the severity of TC impacts. This is because TC motion modulates TC characteristics by exposing it to environments that regulate the subsequent TC behaviors, including TC intensity and structure. In the existing literatures, the impact of global warming on TC translation speed (TCTS) has mixed signals and receives different opinions. Studies have shown that the change in TC frequency by latitudes, in addition to TC’s steering flow, influences the mean TCTS. Analysis of historical observation data indicated that multi-decadal variability can substantially contribute to slow down the mean TCTS. In this study, TC motion is assessed through the track-cluster analysis of four ensemble projections from the High-Resolution Atmospheric Model during a historical period and late 21^st century under RCP 8.5 scenario at 25-km resolution. The two analysis periods share the same year-to-year variability of sea surface temperatures. Changes identified herein are solely attributed to the global warming trend. The impact of varying interannual variabilities under global warming on TC activity is beyond the scope of this study. For TCs in the western North Pacific, this study presents results of six clusters, stratified by TC genesis zones and moving patterns. The analysis better allows exploration of the intra-cluster environmental favorability for TC formation and TC steering flow pattern(s). In a warmer future, most of the projected results (19 of 24 outcomes, composed of the analysis of six clusters and projections under four SST warming patterns) show faster intra-cluster mean TCTS, but only four out of these 19 incidents report a statistically meaningful increase (0.4 – 0.75 m s^-1). The analysis suggests a faster TC-moving trend under global warming, however, with limited statistical significance. The meeting presentation will include discussions on relevant physical causes.

Previous studies have linked interannual variability of tropical cyclone (TC) intensity in the North Atlantic basin (NA) to Sahelian rainfall, vertical shear of the environmental flow, and relative sea surface temperature (SST). In this study, the contribution of TC track changes to the interannual variations of intense hurricane activity in the North Atlantic basin is assessed through numerical experiments. It is found that that observed interannual variations of the frequency of intense hurricanes during the period 1958–2017 are dynamically in agreement with changes in the large-scale ocean/atmosphere environment. Track changes can account for ~50% of the interannual variability of intense hurricanes, while there is no significant difference in individual environmental parameters between active and inactive years. The only significant difference between active and inactive years is in the duration of TC intensification in the region east of 60°W. The duration increase is not accounted for by the slow-down of TC translation. In active years, a southeastward shift of the formation location in the region east of 60°W leads TCs to take a westward prevailing track, which allows TCs to have a longer opportunity for intensification. On the other hand, most TCs in inactive years take a recurving track, decreasing the duration of intensification. This study suggests that the influence of track changes should be considered to understand the basin-wide intensity changes in the North Atlantic basin on the interannual time scale.

This study investigated the interdecadal change of tropical cyclone (TC) activity over the Northwestern Pacific (NWP) in the early-2010s. On the western boundary of the NWP, the interdecadal change of TC activity exhibited a meridional tripole pattern. In contrast to depressed activity over the northern South China Sea (SCS) and Taiwan, TC activity became active over the southern SCS and to the north of Shanghai after the early-2010s. This study focus on the northern NWP, in the recent decade, frequent TC occurrence has brought devastating disasters to East China, Korea and Japan. This work examined the influence of SSWs on the interdecadal change in TC activity. During the 2011–2021 period, SSWs tended to propagate northward, which lead to more TC tracks turning to affect the northern NWP and the surrounding countries. In contrast, the westward-propagating SSWs before the early-2010s were more likely to favor westward moving TCs.

Recent idealized simulations have shown that a system of binary tropical cyclones (TCs) induces vertical wind shear (VWS) in each TC, which can subsequently modify the tracks of these TCs through asymmetric diabatic heating. This study investigates these three-dimensional effects in the western North Pacific using the best track and ERA5 reanalysis data. The TC motion was found to deviate systematically from the steering flow. The direction of deviation is clockwise and repelling with respect to the midpoint of the binary TCs with a separation distance of more than 1000 km. The large-scale upper-level anticyclonic and lower-level cyclonic circulations serve as the VWS for each TC in a manner consistent with the idealized simulations. The VWS of a TC tends to be directed to the rear-left quadrant from the direction of the counterpart TC, where the maxima of rainfall and diabatic heating are observed. The potential vorticity budget analysis shows that the actual TC motion is modulated by the diabatic heating asymmetry that offsets the counterclockwise and approaching motion owing to horizontal advection when the separation distance of the binary TCs is 1000–2000 km. With a small separation distance (<1000 km), horizontal advection becomes significant, but the impact of diabatic heating asymmetry is not negligible. The above-mentioned features are robust, while there are some dependencies on the TC intensities, size, circulation, duration, and geographical location. This research sheds light on the motion of binary TCs from the three-dimensional perspective that has not been previously explained by a two-dimensional barotropic framework.

Forecasting typhoon tracks is important from the viewpoint of disaster prevention, because the damage area of a typhoon depends greatly on the typhoon track. Although the accuracy of typhoon track forecast is improving year by year on average, it is often difficult to forecast typhoons with complex track patterns (hereafter “stray typhoons”). It may be possible to improve forecast accuracy for stray typhoons by identifying the conditions that cause their complex movements and understanding their effect on typhoon tracks. However, there is no established quantitative definition of a stray typhoon, making statistical analysis difficult. In this study, we aim to quantitatively define and extract stray typhoons to statistically analyze the environmental conditions required for stray typhoons to exist. To extract stray typhoons, a numerical value called steering angle is introduced. Steering angle represents how much the steering wheel is turned from the perspective of the “driver” of a typhoon. When the typhoon passes through coordinates P(n), P(n+1), and P(n+2) every 6 hours, the steering angle is defined as the angle changed from path P(n) → P(n+1) to path P(n+1) → (n+2). Next, we extract stray typhoons using this steering angle based on observations. We identify 31 stray typhoons, whose absolute value of the steering angle is in the top 1%, the maximum wind speed exceeds 25 m/s, and the velocity of movement exceeds 5 kt. We will also discuss the usage of these indices to determine the large-scale field preferable for the existence of stray typhoons.

Understanding the impact of climate change on tropical cyclones (TCs) has become a hot topic. The slowdown of TC translation speed contributes greatly to the locally accumulated TC damage. While the recent observational evidence shows that TC translation speed has decreased globally by 10% since the mid-twentieth century, the robustness of the trend is questioned by other studies as effects of changes in observational capability can strongly affect the global trend. Moreover, none of the published studies considered dependence of TC slowdown on TC intensity. This is the caveat of these analyses as the effect of TC slowdown is closely related to TC intensity. Here, we investigate the relationship between TC translation speed trend and TC intensity, and reveal possible reasons for the trend. We show that the global slowing trend without weak TC moments (≤ 17 m s^-1) is about double of that with weak TC moments in a recent study. This is because the slowing trend is dominated by strong TCs’ trend. Stronger (weaker) TCs tend to be controlled more by upper-level (lower-level) steering flow, and the calculated trend of upper-level steering flow is much larger than that of lower-level steering flow. This may be an important reason for the large difference between the slowing trend without weak TC moments and that with weak TC moments. Furthermore, the changes of TC tracks (including inter-basin trend and latitudinal shift), which are partly attributed to data inhomogeneity, make a much larger contribution to the slowing trend, compared with the weakening of tropical circulation, which is related to anthropogenic warming.

Tropical cyclone (TC) translation speed (TCS) is closely related to TC disasters. Recently, the long-term changes in TCS are extensively studied. To date, however, little is known about the multidecadal variability of TCS over the western North Pacific (WNP). By using multiple observational and reanalysis datasets, this study investigated the multidecadal variability of the WNP TCS and the underlying physical mechanisms. The results show that the mean TCS over the WNP presents robust multidecadal variability during the past seven decades. The multidecadal variability of the basin mean TCS is dominated by the TCS over the extratropics. Further analysis shows that the Atlantic multidecadal oscillation (AMO) is mainly responsible for the WNP TCS multidecadal variability. For the tropical mean TCS, AMO induced steering flow anomalies are responsible for the TCS multidecadal variability. In contrast, for the extratropics, AMO positive (negative) phases lead to favorable (unfavorable) large-scale environmental conditions for maintaining TCs, which results in longer (shorter) action time of the mid-latitude steering flow on TCs and leads to higher (lower) TCS. The shift from negative to positive phases of the TCS anomalies during the recent three decades strongly offsets the long-term slowdown trend of TCS, leading to the inconsistent TCS trends before and after the 1980s. Our results help to complete the physical picture for how TCS responds to internal and external forcings, and also provide new evidence for the recent controversial slowdown of TCS.

Complementing the bottom-up methods, the emissions inventories of greenhouse gases (GHGs) can be derived using atmospheric inverse models, providing integrated constraints on surface fluxes from all sectors/processes. For these models a variety of observations are used, including those from ground-based, ship, aircraft and satellite platforms. In particular, recent improvements in the capability of satellite observations of atmospheric composition are providing great advances on spatial resolutions. Among several plans to launch GHG and air quality (AQ) observing satellites in near future, a plan is in progress in Japan to launch the “Global Observing SATellite for Greenhouse gases and Water cycle (GOSAT-GW)”, that will make observations of carbon dioxide (CO₂), methane (CH₄), and nitrogen dioxide (NO₂) at a horizontal resolution of 3 km or less. The missions of GOSAT-GW include (1) monitoring of whole atmosphere-mean concentrations of GHGs, (2) validation of nationwide anthropogenic emissions of GHGs, and (3) detection of GHGs emissions from large sources, such as megacities and power plants. We will provide an overview of the mission/project and some highlights on the potential role in the Asia-Oceania region, in particular, how we can support the mitigation policies on climate change as well as air quality at both international and national levels.

Global emissions from industrial and transport activity, in addition to land change for intensive agriculture, have continuously increased since the pre-industrial era, driven mainly by economic and population growth. Such activity has generated immediate and long-term impacts on the local and regional atmospheric environments, especially on air quality in Southeast Asia (SEA). For instance, over Singapore, air pollution levels are strongly affected by atmospheric circulation, and Its variability is controlled by meteorological conditions and large-scale circulation patterns, including monsoon dynamics and the occurrence of biomass burning over the SEA. Therefore, accurate daily observations of NO₂ and O₃ made from satellites (OMPS, OMI, and TROPOMI) are critical to our ability to quantify and understand the local and regional air quality environment, particularly from the Geostationary Environment Monitoring Spectrometer (GEMS); the first instrument in the geostationary constellation (GEMS, TEMPO, and Sentinel-4) to produce hourly dataset measurements. Nevertheless, the SEA is especially challenging for satellite observations since: the sensitivity of retrievals at near-surface levels can be reduced in environments with a high degree of cloud cover, heavy particle pollution, changes in the O₃ profile within the boundary layer, viewing geometry angle, biomass burning and stratospheric intrusion events. To better understand and reduce the biases between ground-based (Pandora spectrometer) and gridded (satellite and atmospheric models) trace gas datasets, we performed an exploratory data analysis using preliminary NO₂ and O₃ products from the Pandora spectrophotometer, GEMS instrument, and GEOS-CF numerical model. The analysis will focus on: (a) Identifying NO₂ and O₃ diurnal spatiotemporal biases using remote sensing data, (b) characterizing meteorological patterns at SEA, and (c) evaluating uncertainties in air quality numerical models. The outcomes of this study will significantly reduce uncertainties of NO₂ and O₃ algorithm retrievals from GEMS.

Atmospheric fine particulate matter (PM_2.5) in many megacities in Asia have been of great concern due to high emission intensity and frequent pollution episodes. In recent years, a few database of PM_2.5 concentration and its composition have been developed, such as Tracking Air Pollution in China (TAP), China High Air Pollutants (CHAP), and High-resolution Air Quality Reanalysis Dataset over China (CAQRA-aerosol). Based on these database, the concentration of PM_2.5 and its main chemical components (e.g., sulfate, nitrate, ammonium, organic carbon, and black carbon) in China have shown a decreasing trend, but the rates vary by species and regions. In this study, concentrations of chemical species from these database are compared with the monitoring data at an urban site in Beijing during 2016-2019.Besides the trend analysis, sources of PM_2.5 were investigated based on online measurements and receptor model in Beijing. With high-resolution (1h) measurement of multiple components of PM_2.5 in Beijing from 2016 to 2019, positive matrix factorization was applied to quantify the contribution of different sources to PM_2.5 to examine what sources contributed to the decrease of PM_2.5 concentration. Our results showed that the decrease of coal combustion to PM_2.5 was the most significant one for primary sources, from 9.85 μg/m³ in 2016 to 1.57 μg/m³ in 2019. Although all contributing sources of PM_2.5 exhibited a decreasing trend, the relative importance of secondary source increased significantly (from 37% in 2016 to 46% in 2019). During the past decade, many studies about air quality have been conducted in China, with abundant and valuable data available. An on-going project supported by the National Natural Science Foundation of China aims to compile all air quality related data into a database. In this presentation, this project, to be completed by the end of 2023, will be also briefly introduced.

The objective of this study is to assess the spatiotemporal variability of the important compositions and organic markers in three urban areas in Taiwan during 2016-2021. Hi-Vol sampling of PM_2.5 for 7-10 days in summer and winter has been conducted in Taipei, Taichung, and Kaohsiung, the three largest cities in Taiwan. After pre-treatments, filters were analyzed for important compositions and organic markers. Ionic species, sugars, and sugar alcohols were analyzed by Dionex ICS3000 and Thermo ICS5000. Elemental carbon (EC) and organic carbon (OC) were analyzed by a Semi-Continuous OC-EC Field Analyzer. Water-soluble organic carbon (WSOC) was analyzed by Total Organic Carbon Analyzer. Polyaromatic hydrocarbons (PAHs), nitro-PAHs, and important organic markers for cooking, biomass burning, biogenic emission, and secondary organic aerosols were analyzed by Ultra Performance Liquid Chromatography (UPLC, Sciex/Shimadzu) and a triple quadrupole mass spectrometer (MS/MS, Sciex, API5500 Plus). UPLC-MS/MS methods were optimized for the best performance. It was found that PM_2.5 were 13.5 ± 5.7, 17.3 ± 7.0, and 26.0 ± 8.0 µg/m³ in Taipei, Taichung, and Kaohsiung, respectively, during 2016-2021. Ionic species accounted for 48% on average in all three cities, with 6.3 ± 3.7, 8.6 ± 5.0, and 12.9 ± 6.0 µg/m³, respectively. OC accounted for 20-23%, with 2.3 ± 1.2, 3.5 ± 1.1, and 4.2 ± 1.1 µg/m³, respectively. EC accounted for 5-7%, with 0.63 ± 0.34, 0.93 ± 0.42, and 1.31 ± 0.44 µg/m³, respectively. For biomass burning markers, levoglucosan were 11.7 ± 9.2, 35.4 ± 24.4, and 37.0 ± 21.5 ng/m³, and galactosan 16.6 ± 21.7, 4.9 ± 3.5, and 8.6 ± 7.1 ng/m³, respectively. The levoglucosan/OC values were higher in winter compared to those in summer. Source apportionment in different cities/seasons and the COVID-19 impacts will be discussed in the presentation.

Increasing levels of hygroscopic aerosols are one of the major reasons behind irregular rainfall and changed radiative forcing which plays a significant role in climate change. Dicarboxylic acids, which are known for their deliquescence and hygroscopic nature, can act as cloud condensation nuclei (CCN) and ice nuclei (IN), due to which rainfall patterns can be disturbed. Also, the importance of DCAs analysis can be linked to its ability to act as potential organic molecular markers for various anthropogenic and biogenic sources. Despite its significance, very few studies deal with the optimization of the protocol for qualitative and quantitative analysis of DCAs using Gas Chromatography-Mass Spectrometry (GC-MS). In the present study, we have optimized the extraction of DCAs from aerosol samples collected on a quartz filter, by employing several organic solvents with differing relative polarities. Extraction efficiencies of organic solvents were evaluated at different temperatures and pressures using an advanced energized dispersive extractor (EDGE, CEM Corporation, USA). The high polarity and low levels of dicarboxylic acids demand a derivatization step prior to GC analysis to reduce the polarity of the compounds. BSTFA (N, O-bis-(trimethylsilyl)trifluoroacetamide) + TMCS (trimethylchlorosilane) was chosen as the derivatizing reagent, and reaction conditions (temperature, amount of BSTFA, conc. of TMCS) were optimized to give maximum conversions. Separation of compounds was done on HP-5 column with Helium as the carrier gas. Protocol was finalized by selecting the operating parameters of GC-MS in selected Ion monitoring (SIM) mode that reduces the total run time while maintaining a good resolution of peaks. Aerosol samples from northeast region of India have been analyzed using proposed method and DCAs were detected. The average concentration of total DCAs was found 103.75 ng/m³ in the northeast region of India where Pimelic and Suberic acids were found dominant among all the DCAs.

Formaldehyde (HCHO) is an important trace gas that affects the abundance of HO₂ radicals and ozone, leads to complex photochemical processes, and yields a variety of secondary atmospheric pollutants. In a 2021 summer campaign at the Dianshan Lake (DSL) Air Quality Monitoring Supersite in a suburban area of Shanghai, China, we measured atmospheric HCHO by a commercial Aero-Laser formaldehyde monitor, methane, and a range of non-methane hydrocarbons (NMHCs). Ambient HCHO showed a significant diurnal cycle with an average concentration of 2.2 ppbv. During the time period with the most intensive photochemistry (10:00-16:00 LT), secondary production of HCHO was estimated to account for approximately 69.6% according to a multi-linear regression method based on ambient measurements on HCHO, acetylene (C₂H₂), and ozone (O₃). Average secondary HCHO production rate was estimated to be 0.73 ppbv h^-1 during the whole campaign, with a dominant contribution from reactions between alkenes and OH radicals (66.3%), followed by OH radical-initiated reactions with alkanes and aromatics (together 19.0%), OH radical-initiated reactions with OVOCs (8.7%), and ozonolysis of alkenes (6.0%). An overall HCHO loss, including HCHO photolysis, reactions with OH radicals, and dry deposition, was estimated to be 0.49 ppbv h^-1. Calculated net HCHO production rates were in relatively good agreements with the observed rates of HCHO concentration change throughout the sunny days, indicating that HCHO was approximately produced by oxidation of the 24 hydrocarbons we considered at the DSL site during the campaign, whereas calculated net HCHO production rates prevailed over the observed rates of HCHO concentration change in the morning/midday hours in the cloudy and rainy days, indicating a missing loss term, most likely due to HCHO wet deposition. Our results suggest the important role of secondary pollution at the suburb of Shanghai, where alkenes are likely key precursors for HCHO.

A better understanding of marine halogen emission on atmospheric oxidation is crucial for air chemistry and environment of GBA. The OH radical and O₃ are the two key species to indicate the atmospheric oxidation in atmospheric chemistry. In the marine atmosphere, OH levels could be significantly affected by the halogen species emitted from the ocean. However, due to the complicated interactions of halogens with OH through different pathways, it is not well understood how halogens influence OH and even what the sign of the net effect is. In addition, whether oceanic emissions can affect the O₃ level notably has not been fully understood. Therefore, in this study, we aim to quantify the impact of marine-emitted halogens (including Cl, Br, and I) through different pathways on OH and Ozone in the summer by using WRF-CMAQ model with process analysis and state-of-the-art halogen chemistry in GBA. Results show that the net change of P_OH is controlled by the competitions of three main pathways (OH from O₃ photolysis, OH from HO₂ conversion, and OH from HOX, X=Cl, Br, I) through different halogen species. Sea spray aerosols (SSA) and inorganic iodine gases are the major species to influence the strengths of these three pathways and therefore have the most significant impacts on P_OH. In terms of O₃, we found an unexpected hourly and MDA8 O₃ increase on polluted days, because the activation of particulate chloride (Cl⁻) in sea salt aerosol (SSA) is effective due to the high level of dinitrogen pentoxide (N₂O₅) that is formed from the reactions of O₃ and nitrogen dioxide (NO₂). Our results show that marine-emitted halogen species have notable impacts over the ocean and potential impact on coastal atmospheric oxidation by species (SSA, inorganic iodine, and halocarbons), processes (chemistry, radiation, and deposition) and main pathways.

The largest fraction of PM2.5 in China during severe haze is composed of nitrate, which has led to strict control of nitrogen oxide (NO_x) emissions believed to be an effective measure to combat air pollution. However, this notion was challenged by the persistent severe haze pollution observed during the COVID-19 lockdown, when NO_x levels decreased significantly. This study provides direct field evidence that reduced nitrogen oxides (NO) during the shutdown activated nighttime nitrogen chemistry, driving severe haze formation. Dinitrogen pentoxide (N₂O₅) heterogeneous reactions dominate particulate nitrate (pNO₃^-) formation during severe pollution, explaining the higher-than-normal pNO₃^- fraction in PM2.5 despite substantial NO_x reductions. Nocturnal nitrogen chemistry during the lockdown period was completely different from normal conditions in Beijing, but it may vividly depict future scenarios if NOx emissions are strictly controlled without simultaneous control of O₃. Therefore, our results suggest that nocturnal nitrogen chemistry is becoming increasingly important in urban and suburban areas worldwide. More attention needs to be paid to the complex influence of NO, NO₂, VOC, and O₃ on nighttime chemistry when formulating future emission control strategies.

CS₂ is the main precursor of COS in the atmosphere, with about 30-75% coming from CS₂ oxidation. The highly regional distribution and fast vertical concentration attenuation characteristics indicate CS₂ has strong surface removal pathways. Current understanding suggests CS₂ converts to COS through OH oxidation while the photochemistry pathway is neglected in previous model studies. Nevertheless, the CS₂ photo-excitation reaction, which is the initial reaction of the photo-oxidation pathway, has sufficient absorption cross-section in the band range 280-380 nm, implying that the photo-oxidation pathway may also play an active role even at the surface. In this study, we constructed a 1D model of the revised CS₂ reaction network with the addition of the photo-oxidation pathway and extended it to a sulfur cycle. The daytime-weighted zenith angle and solar constant are applied to counteract the spatial-temporal variation and simulate the global average solar radiation. All sulfides concentration in the model reproduced the field measurements or other model estimations. The sulfur budget of the sulfur cycle is determined and the addition of the new pathway has a relatively minor change (1.5%) on the product ratio between COS and SO₂. However, the flux analysis reveals the photo-oxidation pathway and the OH-oxidation pathway contain near-magnitude sulfur fluxes in the CS₂ reaction network and that 15.8% of sulfur flux passes through the photo-oxidation pathway under global average solar irradiance condition. This proportion ranges from 8.1% to 18% depending on the local solar radiation intensity, demonstrating that the photo-oxidation pathway could be a sink for CS₂ in the atmosphere.

In recent years, the concentration of ozone (O₃) in typical urban agglomerations in China has generally increased. This study carries out a three-year observation in the active photochemical reaction period (August-November) in Shenzhen, a megacity in the Pearl River Delta (PRD). Our study selects the photolysis reaction rate constant of NO₂ (j[NO₂]) and O₃ (j[O¹D]) as representative factor of photolysis rate. Principal coordinate analysis (PCoA) is first applied to identify the major factors influencing daily maximum 8-h average O₃ (MDA8-O₃) concentration. Then, the MDA8-O₃ concentration fitting equation is established by a stepwise multiple linear regression (MLR). In result, sensitivity test based on the fitting equation shows that temperature (+35.8%), photolysis rate of j[NO₂] (+11.1%), relative humidity (-10.4%) and photolysis rate of j[O¹D] (-9.5%) have more effect on the concentration of MDA8-O₃ with per factor perturbation (25% change), while increments of △CO (6.9%) and NO₂ (2.7%) have less effect. The insignificant effect produced by NO₂ suggests that Shenzhen is in a transition regime for the O₃-VOC-NO_x sensitivity and thus the reduction of both VOC and NO_x will be effective for O₃ control. In addition, it is found that PM_2.5 promoted O₃ formation by scattering light with wavelength dependence to increase △P (5.87×10³j[NO₂]-2.17×10⁶j[O¹D], indicating the net effect on flux of actinic radiation), which provides a new explanation for the synergistic formation of PM_2.5 and O₃ usually observed in the PRD region. Finally, the method establishes here is based on parameters easily available (e.g., without VOC) and can be widely applied in other regions. The results of this study will help better understand the current ozone formation mechanisms in Shenzhen and promote forecasting ozone pollution more accurately in PRD and similar regions.

The first step towards reducing the particulate matter (PM) load is its quantification scientifically, thereby highlighting the importance of source apportionment. Of the various ways, the emission inventory developed using the ground-based activity data is highly accurate. Developing such emission inventories is challenging because it takes a lot of resources and time. In this study, we focus on developing a gridded emission inventory for PM10 and PM2.5 at 300m × 300m resolution using ground-based activity data for the seven non-attainment cities (Jalandhar, Patiala, Mandi Gobindgarh, Khanna, Naya Nangal, Dera Bassi, and Dera Baba Nanak) of Punjab, a state in India. Further, a comparison with other databases (two global and one regional) EDGARv5, ECLIPSEv6b, and SMoGv1 emission inventories has been performed to highlight the inconsistencies or gaps that have not been accounted for in the other databases. It is observed that out of the seven cities, Jalandhar was the highest contributor to the total PM10 and PM2.5 emissions which is estimated to be 9307 and 4296 tons/year respectively whereas Dera Baba Nanak was the least contributor with total PM10 and PM2.5 emissions of 136 and 92 tons/year respectively. The primary contributing sources are mainly vehicular, road dust, and industrial emissions for all the cities. The inferences of the study will help the policy-makers to make better-informed decisions on reducing emissions and protecting public health.

This study investigated the inte-transport of atmospheric speciated mercury (ASM) between harbor and urban areas in Kaohsiung, Taiwan. Gaseous elemental mercury (GEM), gaseous oxidized mercury (GOM), and particle-bound mercury (PBM) were sampled at four selected sites in harbor and urban areas in four seasons. The spatiotemporal variation, transport routs, and potential sources of ASM were further resolved. Field measurement results indicated that the seasonal average concentration of ASM was ordered as: winter>fall>spring>summer. The concentrations of GEM in harbor and urban areas were 7.36±2.14 and 6.07±1.70 ng/m³, those of GOM in harbor and urban areas were 360±258 and 286±198 pg/m³; and those of PBM in harbor and urban areas were 584±386 and 448±275 pg/m³, respectively. In addition, gas-solid partition of ASM indicated that GEM was the main species at four sites and GOM accounted for only 2.6-6.0% of total atmospheric mercury (TAM). We found that the urban area of Kaohsiung was frequently affected by mercury species emitted from the harbor area. Backward trajectories and wind roses showed that polluted air mainly came from the north in winter. Poor atmospheric dispersion conditions led to the accumulation of local and oversea pollutants and the increase of ASM concentrations. In spring, polluted air transported mainly from the northwest. It was presumed that they were attributed from ocean evaporation and ship exhausts. In summer, clean air transported mainly from industrial complex in the southeast.

Eastern China was extremely wet in summer 2020, which is found to be related to the potential delayed effects of the Indian Ocean Dipole (IOD). Additional knowledge is warranted to improve our understanding of detailed mechanisms of such an effect. In this study, we compared physical processes associated with delayed effects of the IOD and El Niño–Southern Oscillation (ENSO) on summer precipitation. Partial correlation and composite analysis reveal that ENSO modulates precipitation mainly over the Yangtze River Valley, whereas IOD benefits precipitation farther north. Both IOD and ENSO can stimulate anticyclonic circulation over the western North Pacific (WNP) in the ensuing summer but with different spatial distributions related to the different sea surface temperature (SST) evolution processes. IOD is similarly followed by warming signals in the Indian Ocean, known as the “capacitor” effect, but the location is closer to Australia than that associated with ENSO. IOD also stimulates significant SST cooling anomalies over the equatorial Pacific during the ensuing summer, jointly contributing to the anomalous anticyclone over WNP. Numerical experiments confirm that combined effects of the Indian Ocean “capacitor” and equatorial Pacific cooling can generate an anomalous anticyclone with wider distribution in the meridional direction over WNP.

There is uncertainty in the future connection between ENSO and regional precipitation. Here we assessed the causal effects of ENSO on precipitation over Asia in the 2015-2100 period, using data from Coupled Modeling Intercomparison Project Phase 6 (CMIP6) models. Our results show that ENSO impacts on precipitation are mainly observed over parts of Southeast Asia and parts of central and western Asia, while these impacts are not significant over East Asia and South Asia.

Previous studies linked the increase of middle and low reaches of Yangtze River (MLRYR) rainfall to tropical Indian Ocean warming during extreme El Niño’s (e.g. 1982/83 and 1997/98 extreme El Niños) decaying summer. The present study finds the linkage to be different for the recent 2015/16 extreme El Niño’s decaying summer, during which the above-normal rainfall over MLRYR and northern China are respectively linked to southeastern Indian Ocean warming and western tropical Indian Ocean cooling in sea surface temperatures (SSTs). The southeastern Indian Ocean warming helps to maintain the El Niño-induced anomalous lower-level anticyclone over western north Pacific and southern China, which enhances moisture transport to increase rainfall over MLRYR. The western tropical Indian Ocean cooling first enhances the rainfall over the central-northern India through a regional atmospheric circulation, whose latent heating further excites a mid-latitude Asian teleconnection pattern (part of circumglobal teleconnection) that results in an above-normal rainfall over the northern China. The western tropical Indian Ocean cooling during the 2015/16 extreme El Niño is contributed by the increased upward latent heat flux (LHF) anomalies associated with enhanced surface wind speeds, opposite to the earlier two extreme El Niños.

Previous studies revealed that intraseasonal global climate variation is exerted by Madden-Julian Oscillation (MJO); meanwhile its year-to-year variability is strongly controlled by El Niño-Southern Oscillation (ENSO). In Thailand, previous studies indicated that rainfall deficit in summer season is interannually significantly linked to variation of ENSO warm phase (aka, El Niño). However, a compound effect between the MJO and the El Niño on the regional rainfall has not yet been understood. In this study, we, therefore, aimed at quantifying aspects of 20-60-day summer rainfall variation in Thailand in association with active MJO and El Niño. Daily rainfall data covering 1979-2019 were collected from the Thai Meteorological Department, whereas MJO (Oceanic Niño Index; ONI) was provided by the Bureau of Meteorology (National Centers for Environmental Prediction). Additionally, ERA-5 reanalysis is used for considering large-scale moisture transport. The result shows that the combined effect between MJO and the warm episode on Thailand’s rainfall is spatially varying depending on different MJO phases by which the 20-60-day rainfall anomalies in upper Thailand are above (below) normal at MJO phases 3-4 (6-7). For southern Thailand, above normal rainfall is also profound at MJO phase 3-4, while below normal rainfall lasts slightly longer than in the upper sub-region (phases 5-7). The intraseasonal variation of rainfall in the upper Thailand is strongly associated with moisture transport from Bay of Bengal, Gulf of Thailand, and South China Sea, while moisture coming from easterly, and westerly is a major control of the rainfall anomalies in the South. Hence, this study provides a fundamental understanding and characteristics of the combined effect between the climate drivers on intraseasonal rainfall variation and is useful for further development of sub-seasonal to seasonal (S2S) prediction in Thailand.

This study investigates the interdecadal variation of the Scandinavian (SCA) pattern and corresponding drivers during the boreal winter. It is found that the SCA pattern experiences a prominent regime shift from its negative to positive phase in the early 2000s based on several reanalyses. This interdecadal change contributes to an extensive cooling over Siberia after the early 2000s, revealing its importance for recent variation of climate over Eurasia. The outputs from 35 couple models within the Coupled Model Intercomparison Projection Phase 6 (CMIP6) are also analyzed. The results show that the interdecadal change of SCA is weak in response to external forcings but can be largely explained by internal variability associated with a change of precipitation over the tropical Atlantic. Further analysis indicates that the enhanced tropical convection induces poleward propagation of Rossby waves and further results in an intensification of geopotential height over the Scandinavian Peninsula during the transition to positive SCA phases. These findings imply a contribution of tropical forcing to the observed interdecadal strengthening of SCA around the early 2000s and offer an insight into the understanding of future climate change over the Eurasian continent.

This study investigates the variation in available potential energy of stationary eddies in the mid-troposphere over the midlatitude Northern Hemisphere (30˚ to 60˚N) from 1950 to 2021, based on the ERA5 reanalysis and its preliminary back extension. The variation is attributed mainly to change in the wavenumber-1 eddy. It is also noted that potential energy conversion from the zonal mean flow in early winter plays an important role in the 20-year variation. In addition, statistical analysis and numerical simulation of an ICTP model demonstrate that the potential energy conversion is controlled by the North Pacific Gyre Oscillation (NPGO), which induces anomalous meridional temperature advection in the mid-troposphere over Northeast Asia.

This study finds that the observed decrease in winter-spring Tibetan Plateau snow depth since 2000 has played an important role in weakening the correlation between rapidly-intensifying tropical cyclones over the western North Pacific and tropical Indian Ocean sea surface temperatures (SSTs). Tibetan Plateau snow depth modulates convective activity through changes in the South Asian High. Increased Tibetan Plateau snow depth promotes basin-wide tropical Indian Ocean cooling in spring through a Gill-type response. In the following seasons, downward latent heat flux anomalies associated with a weaker monsoon circulation contributes to warm SST anomalies over the western tropical Indian Ocean, thus favoring a positive phase of the Indian Ocean Dipole. This evolution of tropical Indian Ocean SST associated with anomalously high Tibetan Plateau snow depth potentially weakens the relationship between rapidly-intensifying tropical cyclone frequency and spring tropical Indian Ocean SST. When the effect of Tibetan Plateau snow depth is removed via partial correlation analysis, we find a significant relationship between spring tropical Indian Ocean SST and rapidly-intensifying tropical cyclones as well as corresponding large-scale environmental factors. The results of this study enhance understanding of changes in tropical cyclone intensity and have implications for seasonal forecasting of tropical cyclone intensity over the western North Pacific basin. This study also emphasizes the importance of Tibetan Plateau thermal forcing in atmosphere-ocean coupling.

Air quality has become a pressing public health issue globally due to increased urbanization and industrialization over the past several decades. One of the major contributors to poor air quality in urban areas is particulate matter (PM or aerosols). PM2.5, with an aerodynamic diameter of less than 2.5 μm, can cause respiratory and lung diseases and even premature death. It also plays a crucial role in atmospheric processes and is linked to climate change. Environmental agencies monitor and regulate particulate emissions by measuring PM2.5 concentrations on a hourly and daily basis. However, limited ground monitors are available, and recent advancements in satellite remote sensing and global numerical models offer the possibility of filling in measurement gaps. In this study, we plan to implement machine learning (ML) and deep learning (DL) methods to estimate surface PM2.5 at hourly and daily scales globally. We will combine data from ground networks, satellite platforms, and model outputs to develop an ML/DL algorithm that produces high-quality records over 22-years. Our initial analysis used one year of data and employed random forest and deep neural network algorithms. The results showed a mean bias close to zero, with a slope of 1.02 and an RMSE of 6.4 μgm^-3 globally. This study will present an analysis based on three years of data, and we will also test more advanced ML/DL algorithms (including CNN) to determine the best way to estimate PM2.5 at hourly and daily scales.

In recent years, surface ozone has observed an ascending trend in China despite of tremendous effort in reduction of emission, thus monitoring the surface ozone concentration with high accuracy and wide coverage is highly demanded. Previous studies have indicated that satellite observations can be used to retrieval the surface ozone estimation. However, the accuracy of the model estimations have not met the requirement to evaluate the spatial and temporal variations. This study proposes a machine-learning based model combining photochemical mechanisms of surface ozone aiming to better depict the patterns of surface ozone concentrations. Particularly, the incorporation of photochemical features including surface ultraviolet irradiance and nitrogen dioxide, which are among the feasible indicators and precursors of surface ozone formation, significantly enhances the model accuracy of surface ozone estimation. The proposed model achieves high accuracy (R²=0.853 and RMSE =17.09 μg/m³) with spatial continuity. Moreover, the model have overcome the critical challenges of the current model applications by promoting the effect of surface ozone estimation in regions with sparse distribution of surface monitoring sites.

Vehicle emissions have become a major source of air pollution in urban areas, especially for near-road environments, where the pollution characteristics are difficult to capture by a single-scale air quality model due to the complex composition of the underlying surface. This study promoted the capability of the emissions calculation and the air quality simulation to fine scales such as urban areas and street canyons. We developed a multi-scale coupling model (CMAQ-RLINE_URBAN) based on a street canyon flow scheme using Computational Fluid Dynamic and two machine learning methods. To estimate the influence of various street canyons on the dispersion of air pollutants, a machine-learning-based street canyon flow (MLSCF) scheme was established. It enables quantitative analysis on the effects of vehicle emissions on urban roadside NO2 concentrations at a high spatial resolution of 50m_50 m. The results indicated that compared with the Community Multiscale Air Quality (CMAQ) model, the hybrid model improved the underestimation of NO2 concentration at near-road sites with the mean bias (MB) changing from -10 to 6.3 μgm􀀀3. The MLSCF scheme obviously increased upwind concentrations within deep street canyons due to changes in the wind environment caused by the vortex. In summer, the relative contribution of vehicles to NO2 concentrations in Beijing urban areas was 39% on average, similar to results from the CMAQ-ISAM (Integrated Source Apportionment Method) model, but it increased significantly with the decreased distance to the road centerline, especially on urban freeways, where it reached 75%.

The moisture in the air is vital for reducing the PM2.5 concentration. We performed simulations using the random forest regression model to account for the effect of wetlands in lowering the PM2.5 level over some of the selected Indian cities in the north Indian plain which are known to be some of the most highly polluted cities in the world. The meteorological factors, such as 2 m air temperature, surface pressure, AOD, relative humidity, wind speed, boundary layer height, vertical airflow, and precipitation, in combination with surface greenness feature, i.e., NDVI, and proximity to wetlands were considered as potential covariates to model the PM2.5. The simulations, at an annual scale, suggest wetland proximity is more important than precipitation, surface pressure and wind speed, and even relative humidity. However, we found variability of the influence of wetland’s proximity at the seasonal scale, with the highest influence noted in the pre-monsoon season when thermal conditions are high as well as in the winter season when vapor pressure remains lower. We also found that above a PM level of 300 µg m^-3, the explanation of the regression model reduces to 57% (R² = 0.57), including an RMSE and MAE of 42.83 µg m^-3 and 29.45 µg m^-3, respectively. However, below 300 µg m^-3, the explanation increases to more than 60% (R² = 0.61), with an RMSE and MAE of 34.96 µg m^-3 and 25.84 µg m^-3. The modeling suggested a differential rate of increment around the wetlands. It follows a limited area influence (~ 3 km) within which the PM_2.5 remains low. The findings, indeed, suggest that restoration of wetlands within and surrounding the cities could be one of the effective nature-based solutions to curb the PM2.5 level while putting other pollution regulatory measures in place.

Since the implementation of various air pollution control policies, the concentrations of major air pollutants in the Yangtze River Delta (YRD) region have changed significantly, but the contributions of emissions and meteorological factors are unclear. Therefore, it is essential to decouple the effects of meteorology and emissions changes to air quality. This study applied a de-weather method based on machine learning technique to quantify the contribution of meteorology and emission changes to air quality from 2015 to 2021 in four cities in the YRD region. The results show that the significant reductions in PM_2.5, NO₂, and SO₂ emissions (57.2%-68.2%，80.7%-94.6%，81.6%-96.1%) offset the adverse effects of meteorological conditions, resulting in lower pollutant concentrations. The meteorological contribution of maximum daily 8-h average O₃ (MDA8_O₃) shows a stronger effect than others (23.5%-42.1%) , and meteorological factors promote the increase of MDA8_O₃ concentrations (4.7%) but emission changes overall result a decrease of MDA8_O₃ concentrations (−3.2%) in the four cities on average. NO₂ and MDA8_O₃ decreases more rapidly from 2019 to 2021, mainly because the emissions play a stronger role to reduce pollutant concentrations than 2015 to 2018. However, emissions changes have weaker reduction effects on PM_2.5 and SO₂ from 2019 to 2021 than 2015 to 2018. De-weather methods can effectively separate the effects of meteorology and emissions changes on pollutant trends, which helps to evaluate the real effects of emission control policies on pollutant concentrations.

The Madden–Julian oscillation (MJO) is a planetary-scale convective-coupled system that amplifies over the Indo-Pacific warm pool. How the MJO is amplified and maintained and why the MJO has planetary scale are the key issues for understanding the MJO dynamics. Using a theoretical model, this study shows that the boundary layer moisture convergence feedback (BLMCF) and the cloud-radiative feedback (CRF) are major sources for the amplification of MJO. The destabilizing effect of the BLMCF is augmented over the warm ocean due to higher background moisture content there, explaining why the MJO amplifies (decays) over the warm (cold) ocean. Moreover, it is found that only the BLMCF favors the growth of the MJO over planetary scales. Due to a small Coriolis force in tropics, theoretical studies of Madden–Julian oscillation (MJO) often assume weak temperature gradient (WTG) balance, which neglects the temperature feedback (manifested in the temperature tendency). Using the scale analysis, this study further indicates that the rotation effect is strong at the MJO scales, so that the temperature feedback (TF) is as important as the moisture feedback (manifested in the moisture tendency) that is often considered to be critical for MJO. The TF is shown to be critical for the maintenance of MJO over the warm pool. This is because the TF could boost the energy generation for the system, favoring the self-maintenance of the MJO. With the interaction between BLMCF and TF, more moist available energy is generated on planetary scales, explaining the planetary-scale selection of the MJO.

Shallow cumulus clouds play a crucial role in the Earth's energy budget by vertically redistributing momentum, heat, and moisture from the surface to the free atmosphere, but they cannot be explicitly resolved in current numerical models. Recent studies suggest that these clouds are composed of sticky rising thermals whose vertical velocity is mainly controlled by the buoyancy source and the drag due to the pressure perturbation. However, little is known about how the pressure drag of thermals can be related to that of a large ensemble of clouds, which is the focus of convection parameterization in climate models. This study takes the first step to bridge the gap between the pressure drag of a shallow cloud ensemble and that of an individual cloud composed of rising thermals. It is found that the pressure drag for a cloud ensemble is primarily controlled by the dynamical component. The dominance of dynamical pressure drag and its increased magnitude with height are independent of cloud lifetime and are common features of individual clouds except that the total drag of a single cloud over life cycle presents vertical oscillations. These oscillations are associated with successive rising thermals but are further complicated by the evaporation-driven downdrafts outside the cloud. The horizontal vorticity associated with the vortical structure is amplified as the thermals rise to higher altitudes due to continuous baroclinic vorticity generation. This leads to the increased magnitude of local minima of dynamical pressure perturbation with height and consequently to increased dynamical pressure drag. These findings could provide useful insights for a reasonable representation of pressure drag for shallow cumulus clouds in the convection parameterization to improve weather forecast and climate projection.

Radiative-convective equilibrium (RCE) is considered as the first approximation of the tropical atmosphere, which is often used as a fundamental framework to understanding climate change. With the same external forcing, numerical simulations indicate that RCE possesses multiple stable states. One is the ordinary RCE state with convection occurs randomly, and the other one is the state with bifurcation of dry and wet areas. In this study, we show that the ordinary RCE state could transfer to the other state through two phases. The first phase builds the secondary circulation and leads to the rapid intensification of dry areas, which is related to a fast and local instability revealed in previous studies. The second phase is a slow and global phase, during which dry areas expand and wet areas get wetter. A theoretical model is built to understand the mechanisms of the second phase. Based on this model, a new instability is proposed to explain the expansion of dry areas at the early period of the second phase. This work explains how two kinds of instabilities with different scales drives the ordinary RCE state to a new stable state, providing a novel pathway to understanding the evolution of RCE state and climate change in the real atmosphere.

Mesoscale convective systems (MCS) merged and sustained in the record-breaking heavy rainfall event during 19-21 July 2021 in Henan Province, China. Whether warming climate enhances the intensity or area of MCS is crucial for understanding the changing risk of extreme rainfall events under global warming. Ensemble simulations based on the Weather Research and Forecasting (WRF) is adopted and the human-induced temperature and humidity changes in the initial and boundary condition is estimated by ‘pseudo global warming’ approach which compares ALL-forcing and NAT-forcing in global climate model. An MCS-tracking algorithm indicates that the total rainfall volume and maximum rainfall intensity in the MCS region both enhance by ~10%. The spatial area of developing MCSs are 10%~40% larger in a warmer condition, but developed MCSs are in same sizes or smaller. Probability distributions of extreme rainfall grids inside MCS (larger than 100 mm/h) increase up to 50% in warming climate, which greatly aggravate the threat of flood. Furthermore, vertical velocity is enhanced by 10% and the cloud top is raised by 0.5 km, while the snow and ice expand horizontally and upward near the top of MCSs. Thus, the flood risk managements should take the changes of MCSs characteristics due to warming climate into account in reply to the future threat of extreme rainfall event.

The transition from mixed Rossby-gravity (MRG) waves to tropical depression (TD)-type disturbances is commonly observed over the western North Pacific (WNP). This study aims to understand the role of diabatic heating in the wave conversion process. Using the potential vorticity (PV) diagnostic equation, it is found that diabatic heating plays a crucial role in the growth of the wave through the release of latent heat from convection. In the MRG wave stage, it is affected by both advection of the background airflow and diabatic heating, while in the TD-type disturbance stage, it is mainly affected by diabatic heating. Further sensitivity numerical experiments suggest that both thermodynamic and dynamic processes are equally significant in the MRG-TD transition process, and that the process cannot be completed if either factor is absent. The key influences and pathways of diabatic heating on MRG-TD conversion are concluded, and the physical mechanisms affecting the wave conversion are supplemented.

Observations and numerical simulations show that there is typically a cold anomaly above the TC warm core. This cooling can work with the warm core to reduce the upper-tropospheric stability and affect the water vapor and momentum transport between the TC and the stratosphere. However, little is known about how the cold anomaly evolves during the life cycle of a TC and how it can be related to changes in TC intensity. This study examines the evolution of the cold anomaly throughout the lifetime of a TC with an idealized TC simulation. The cold anomaly emerges at the TC center before rapid intensification starts and continues to enhance during TC development. It moves radially and forms a cold ring around the TC until the mature stage. The advection, specifically the vertical mean motion, dominates the generation of the TLC, but the advection due to eddy motions is responsible for the cold anomaly near the center. Our findings demonstrate the importance of tropopause layer cooling on TC structure and intensity changes and indicate the necessity of additional observations of this cold anomaly to improve TC predictions.

The advanced capabilities of the new generation of operational weather satellite imagers in low-Earth orbit (LEO; e.g., VIIRS) and geostationary (GEO; e.g., ABI, AHI, etc.), having spectral and spatial capabilities analogous to the NASA Earth Observing System (EOS) MODIS, offer the opportunity to extend the high impact EOS MODIS dataset for clouds into the next decade and into the time domain. Such a merged, and consistent, LEO/GEO cloud product Program of Record (PoR) can enable enhanced climate and process studies by NASA investigators and the broader research community. In addition, this PoR is desired to provide critical synergy with the NASA Atmosphere Observing System (AOS), which is currently in formulation and is designed to address the Aerosols, Clouds, Convection, and Precipitation Designated Observables identified by the 2018 NASA Earth Science Decadal Survey. In a companion presentation (Platnick et al., also submitted to this session), we give an overview of a NASA research cloud optical property algorithm for the new GEO imagers that was developed to provide consistency with the NASA MODIS/VIIRS cloud continuity products CLDPROP. Here, we show results of our efforts to evaluate the consistency of this new GEO cloud dataset against the MODIS/VIIRS continuity dataset. We also will discuss ongoing challenges towards achieving LEO/GEO product consistency, including the impacts of fundamental differences in sensor specifications and/or orbits (e.g., spectral channel differences, spatial resolution/swath, viewing/solar geometries), differences in relative radiometric calibration, and forward radiative model issues.

Detection of atmospheric constituents such as cloud and aerosols with satellite observations is often a critical initial step in many remote sensing algorithms. Many traditional algorithms were developed based on underlying physics and hand tuned thresholds. Weakness of these methods is that it is challenging and time-consuming to develop algorithms across multiple instruments and locations. To overcome this weakness, we designed and trained a couple of Machine Learning (ML) based models for cloud and aerosol detection. Specifically, a Random Forest (RF) model, a Convolutional based Encoder-Decoder Neural Network, and a Recurrent Neural Network (RNN) are designed and trained with spectral, spectral/spatial, and spectral/spatial/temporal input, respectively. The first two models that require spectral and spatial input are designed for passive spectrometers such as VIIRS and MODIS; while the third model is designed for geostationary instruments, such as GOES-16/17 ABI. A hybrid training database is generated based on years of collocated satellite-satellite (e.g., CALIPSO/VIIRS and CALIPSO/ABI) and satellite-ground station (e.g., ABI/AERONET) data, and manually picked events with labels. In this presentation, we will introduce of the training database and compare the three different models.

Clouds cover most of the Earth, and a slight variation in their properties disturbs the radiation budget, modulates weather, and impacts climate change. Due to the strong interaction with shortwave and longwave radiation, any slight changes in the clouds have a potent effect on the climate system. Clouds are highly dynamic in space and time, and their proper representation in climate models remains challenging. Cloud properties such as cloud base height (CBH), cloud top height (CTH), cloud fraction, and vertical layer structure significantly affect the radiative balance, atmospheric circulations, and other meteorological processes. Therefore, monitoring and investigating these cloud parameters is essential for climate diagnosis and predicting future climate. This study deals with the investigation of cloud properties over two locations: Udaipur (24.6°N, 73.7°E) and Mount Abu (24.5°N, 72.7°E) in the Aravalli ranges of Western India using a ground-based Lidar (ceilometer), satellite, and reanalysis datasets. Over these regions, the average cloud base height for layers 1 and 2 (CBH1 and CBH2) follows the same seasonal pattern: low during monsoon and high during pre and post-monsoon. The cloud top height (CTH) information is obtained for the same period using Moderate Resolution Imaging Spectroradiometer (MODIS) instrument onboard Terra and Aqua satellites. The observed CTH ranges from about 0.2km to 19km. Furthermore, the classification of these observed clouds is performed using cloud top pressure and cloud optical thickness values as per the International Satellite Cloud Climatology Project (ISCCP) cloud classification. Over Udaipur, cirrostratus clouds were found to have a maximum occurrence (~ 36%) during the study period. ERA5 and MODIS-derived CBH doesn’t correlate well with the observed CBH from the ceilometer, indicating the limitation of reanalysis and satellite observations of cloud base height over the complex orographic region.

Earth observing instruments from multiple organizations, including NASA and NOAA in the United States, continue to provide multiple observations of aerosol plumes from wildfires. The Segmentation, Instance Tracking, and data Fusion Using multi-SEnsor imagery (SIT-FUSE) project at the Jet Propulsion Laboratory has developed an unsupervised machine learning framework that allows users to segment instances of objects like wildfires and smoke plumes in single and multi-sensor scenes from satellite instruments with minimal human intervention in low and no label environments. This is an important step toward allowing automatic smoke plume and wildfire instance tracking through time from multiple instruments. We will demonstrate this system with examples from the NASA/NOAA FIREX-AQ field campaign that took place in 2019. In addition, we will describe ongoing work to allow automatic feature identification and tracking using deep learning - specifically contrastive learning (CL) enhanced by the topological features of the object instances detected.

For environmental monitoring, the Geostationary Environment Monitoring Spectrometer (GEMS) measures backscattered radiances in the ultraviolet and visible spectral region (300-500 nm) over the Asia-Pacific region to retrieve key atmospheric constituents and aerosol properties eight times a day. The measured radiances and retrieved properties of GEMS have been collected in about two years of operation after completion of the in-orbit test in October 2020. There have been substantial efforts to calibrate the GEMS Level 1B data and here we briefly introduce the updates of calibration processes along with one of the studies regarding bad pixels of GEMS. Bad pixels occurred on the detector array could cause consistent information gaps and the issue becomes more problematic for hyperspectral measurements because only a few defective radiances in a spectrum can induce erroneous spectral features. To efficiently resolve the spatial discontinuity caused by bad pixels in the measured radiances and further in the retrieved properties of GEMS, we apply spectral replacement on the radiance level with machine learning models using multivariate linear regression and artificial neural network (ANN). Because the key information for the retrieval is originated from subtle spectral features, we more focus on whether the features can be successfully reproduced with the models trained with large datasets, the GEMS defect-free measurements. The results and conclusions can be found in the paper published with the identical title, and briefly said, the spectral replacement can be quite effective for retrievals (i.e., cloud centroid pressure) with some conditions such as input and output ranges or instrument artifacts. However, limitation still remains for the approach especially for low signals or very strong absorption features such as ozone. It indicates additional information would be needed to improve the methods if one pursues very high retrieval accuracy with the synthetic spectra.

NASA is preparing to launch the Tropospheric Emissions: Monitoring of POllution (TEMPO) mission into Geostationary Earth Orbit (GEO) in April 2023. TEMPO will provide hourly and sub-hourly daytime observations of aerosols and trace gases, including nitrogen dioxide, formaldehyde, sulfur dioxide, and ozone, at high spatial resolution (~2.0 x 4.75 km²) across a Field of Regard (FoR) covering greater North America. The hyperspectral ultraviolet-to-visible measurements from the TEMPO grating spectrometer will permit an ozone profile retrieval capable of monitoring the diurnal evolution of ozone in the planetary boundary layer. The non-standard or special scan operations of TEMPO at sub-hourly frequency (e.g., 2-10 minutes) over selected slices of the FoR will further enhance monitoring capabilities during air quality disasters, such as wildfires, volcanic eruptions, and dust storms. During the pre-launch phase of this mission, a large diversity of stakeholders and end-users have been engaged in the TEMPO Early Adopters Program, supported by the NASA Applied Sciences Program, which aims to maximize the societal benefit of TEMPO data after launch. Key objectives of the Early Adopters Program include preparing the user community for operational TEMPO data through early application of proxy products and designing products, tools, and mission planning activities to better meet user needs. This presentation will provide a TEMPO mission status update, details on data product developments, an Early Adopters Program overview, demonstrations of science applications enabled through TEMPO data, and insight into the special operations component of the mission.

Retrieving aerosol characteristics such as particle size, shape, and absorption properties is a big challenge in satellite-based remote sensing, mainly for the single-viewing-angle instrument. Multi-angle instruments such as Multi-angle Imaging Spectroradiometer (MISR) measures the reflected radiation from nine viewing angles, allowing the retrieval algorithm to accurately distinguish different aerosol microphysical and optical properties along with total aerosol loading in the atmospheric column. Recently, MISR upgraded the aerosol retrieval algorithm version 23 (V23) and enhanced the spatial resolution of the aerosol products to 4.4km, enabling the study of the aerosol characteristics at a finer scale. This study uses 20 years of MISR V23 aerosol products to comprehensively investigate aerosol optical and microphysical properties and their performance over Gulf Cooperation Council (GCC) countries using two decades of aerosol products from the MISR algorithm. The assessment of MISR AOD showed good agreement with ground-truth AOD from AERONET, with ~75% of retrieval falling within the expected error (0.05±0.2AOD) and a high Pearson’s correlation coefficient (R = 0.86). However, MISR relatively underestimates AOD at higher aerosol loading conditions (AOD>0.6) while overestimating coarse-dominated aerosol mixture (Angstrom Exponent <0.7). The 20 years mean MISR AOD showed higher aerosol loading dominated by coarse aerosol types over the Eastern province of Saudi Arabia and Rub’al Khali. Seasonally, the size-fractionated AOD showed that GCC countries dominated by coarse AOD (cAOD) and non-spherical aerosol particles mainly in Spring, while the contribution of small AOD (sAOD) to the total range from 42-45% for Winter and Autumn. Overall, the high spatial resolution of MISR aerosol data products has a great potential to identify an aerosol hot spot and constrain the aerosol types, mainly dust, across the highly polluted region in the Middle East.

The Dark Target (DT) retrieval algorithm was developed for Moderate-resolution Imaging Spectroradiometer (MODIS) to derive aerosol optical depth (AOD) over global land and ocean. Because of its relative simplicity and flexibility, it can be implemented on the Visible Infrared Imaging Radiometer Suite (VIIRS), or any imager with MODIS-like wavelength information. We are retrieving on VIIRS Suomi-NPP and NOAA-20 which are also in sun-synchronous low-earth orbit (LEO) to stitch together a long-term data record (now at 22 years and counting). More recently, DT has been ported to Full Disk imagery from sensors in Geostationary orbit (GEO), including Advanced Baseline Imager (ABI) on GOES-R series (East and West) and Advanced Himawari Imager (AHI) on the Himawari series. With these GEO sensors, we observe much of the globe every 10 minutes, thus providing the opportunity to characterize rapid aerosol changes and the aerosol diurnal cycle. Overall, the new GEO products look good, and compare reasonably well to existing LEO and to ground-based sunphotometer data. We discuss remaining challenges regarding relative sampling of GEO versus LEO as well as calibration. Considering the period when all 6 sensors are operating, we are imminently releasing a combined 4-year GEO / LEO dataset at 0.25°x0.25° resolution and 30-minute time intervals. Here, we introduce the new dataset including details about the product, expected accuracies, steps to acquire, and intended use. 

Long-term variations in precipitation during the major rainy period in Japan—the Baiu (June–July) and Akisame (September–October) seasons—are investigated using precipitation records from 44 weather stations in western to eastern Japan over the past 120 years (1901–2020). The total amount of Baiu precipitation has increased over the 1901–2020 period, mainly during the mid–late stages of the season (late June–July) over regions on the Sea of Japan side of the country. In contrast, the precipitation amount during the Akisame season has decreased, mainly during the mid-stage (late September–early October) over all regions. The frequency and intensity of heavy precipitation have generally increased in both seasons, but the trends are much stronger for the Baiu season compared to those for the Akisame season. A prominent positive trend, 23.5% per 100 years (18.1% per ℃), which is much higher than the Clausius–Clapeyron rate (approximately 7% per ℃), is observed for the Sea of Japan side of western Japan for the seasonal maximum 1-day precipitation total during the Baiu season. It may be noteworthy that the observed long-term trends differ greatly between the Baiu and Akisame seasons even though the statistical significances of the trends are not so high, because similar differences between the two rainy seasons are found in results of global warming simulations.

Accurate prediction of global land monsoon rainfall on a sub-seasonal (2-8 weeks) time scale has become a worldwide demand. Current forecasts of weekly-mean rainfall in most monsoon regions, however, have limited skills beyond two weeks, calling for a more profound understanding of monsoon intraseasonal variability (ISV). We show that the high-frequency (HF; 8-20 days) ISV, crucial for the Week 2 and Week 3 predictions, accounts for about 53-70% of the total (8-70 days) ISV, generally dominating the sub-seasonal predictability of various land monsoons, while the low-frequency (LF; 20-70 days)’s contribution is comparable to HF only over Australia (AU; 47%), South Asia (SA; 43%), and South America (SAM; 40%). The leading modes of HFISVs in Northern Hemisphere (NH) monsoons primarily originate from different convectively coupled equatorial waves, while from mid-latitude wave trains for Southern Hemisphere (SH) monsoons and East Asian (EA) monsoon. The Madden-Julian Oscillation (MJO) directly regulates LFISVs in Asian-Australian monsoon and affects American and African monsoons by exciting Kelvin waves and mid-latitude teleconnections. During the past four decades, the HF (LF) ISVs have considerably intensified over Asian (Asian-Australian) monsoon but weakened over American (SAM) monsoon. Sub-seasonal to seasonal (S2S) prediction models exhibit higher sub-seasonal prediction skills over AU, SA, and SAM monsoons that have larger LFISV contributions than other monsoons. These results suggest an urgent need to improve the simulation of convectively coupled equatorial waves and two-way interactions between regional monsoon ISVs and mid-latitude processes and between MJO and regional monsoons, especially under the global warming scenarios.

This study investigates the characteristics and long-term trends of southwesterly flow around southern Taiwan (hereafter, SWs) during mei-yu seasons (15 May - 15 June) from 1979 to 2022. The results show that the occurrence number of SWs had in general an increasing trend over this 44-year period, with a decadal oscillation by starting from a relatively small number in the 1980s and reaching a relative peak in the 2000s. The 4-year and 10-year periods show more power according to wavelet analysis. While the 4-year power period was more evident from 1995 to 2005, the 10-year power period was more evident before 2000. This posts a potential threat to Taiwan due to the increasing trend of heavy rainfall associated with the longer duration and higher moisture flux of the SWs events. The SWs activity was influenced by the long-term increasing trend of geopotential height and its decadal variability near Taiwan. When the intra-seasonal oscillation was evident, the weather system that mainly affected the occurrence of SWs was the low pressure system to the north of Taiwan, while when it was weak, the location of the western North Pacific subtropical high to the south of Taiwan was more important. In addition, the SWs index which was highly correlated with the precipitation during mei-yu seasons can effectively reflect the interannual variability of precipitation in Taiwan in periods of different lengths. These findings indicate that the SWs index can be used as a monsoonal precipitation index for Taiwan, especially southern Taiwan.

As the main water vapor channel entrance of the plateau, the water vapor flux convergence and precipitation in the Yarlung Tsangpo River Canyon region have been decreasing continuously since 1979, which has an important impact on the water storage of the plateau. Since the Lagrangian method is more responsive to the spatial and temporal variability of water vapor transport than the Eulerian method, this study uses the Lagrangian trajectory tracking model LAGRANTO to drive the ERA5 reanalysis data to track the water vapor transport in the canyon region for the past 40 years and obtain specific water vapor transport trajectories, and by analyzing the changes in the spatial and temporal distribution of the trajectories including trajectory location and trajectory height changes, as well as the interannual variability of water vapor flux on the trajectories By analyzing the interannual variation of water vapor flux on the trajectory and the interannual trend of water vapor flux on the trajectory, the main factors affecting the variation of water vapor transport in the southern part of the plateau are identified by combining the large-scale circulation such as monsoon and the interdecadal oscillation index for correlation analysis. At this stage, by comparing the differences in water vapor transport in typical wet and dry years, it is found that in addition to the source water vapor contribution and the influence of monsoon, the precipitation loss-related processes along the water vapor transport play a decisive role in the water vapor balance of the Yarlung Tsangpo River region, so the trend of water vapor transport changes in the past 40 years will be further analyzed and the main factors affecting water vapor loss will be studied, aiming to clarify the influence mechanism of water vapor transport changes in the southern plateau.

The mean Indian summer monsoon (ISM) rainfall as well as its duration have a profound impact on the agriculture practice in the country. Due to the recent increase in the surface temperature, global circulation patterns exhibit considerable changes which also affects the characteristics of ISM. The present study aims to find out any long-term changes in the monsoon onset and withdrawal times and the length of the monsoon rainy season that exists over different parts of India and the possible mechanisms behind it. During the last four decades, the trend analysis of ISM over south India and North-West (NW) India shows an early onset in both regions. However, the trends are statistically less significant. The monsoon withdrawal dates over NW India and south India show a statistically significant delay of about 6 days/decade and 3.25 days/decade, respectively. As a result, the monsoon season over NW India and south India shows a lengthening of about 7.8 days/decade and 3.5 days/decade, respectively. During the withdrawal phase of the ISM, a stronger monsoon low-level jet and an enhancement of the ISM rainfall have been observed in recent decades. The role played by factors such as Eurasian temperature, Indian ocean warming, Atlantic Multidecadal Oscillation (AMO) and Pacific Decadal Oscillation (PDO) on the ISM withdrawal is examined. The AMO has changed its phase from negative to positive in recent decades, particularly after about 1998, which might have played a key role in enhancing the meridional tropospheric temperature gradient. The stronger meridional tropospheric temperature gradient and the Eurasian surface warming observed in recent decades might played a key role in the delayed monsoon withdrawal over NW India. The delayed ISM withdrawal over the south of India is mainly attributed to the AMO phase change and the changes in the Indian Ocean.

Severe and extreme drought in southeastern mainland Asia (SEMA) had worsened drastically in 2010–2019, more than twice as frequently in the preceding decade. It is found that the spring rainfall has undergone a concordant positive-to-negative transition with the turning point occurred at the 2010, and can explain 43% of the overall regime shift towards exaggerated severely dry condition. Associated with the decadal precipitation change, the anomalous northeasterlies prevail over SEMA, resulting in weakened eastward moisture propagation from Indian Ocean as well as enhanced divergence. Meanwhile, there is a downward motion over SEMA. Such circulation pattern is remotely forced by teleconnection from the Tropical Western Indian (TWI) SST. TWI SST is negatively correlated with SEMA precipitation and highlights a regime shift around 2010, after which the TWI have persistent warm SST helping to maintain deficient SEMA precipitation. In terms of the physical mechanism, the heating in TWI warms the troposphere aloft and emanates wedge-shaped Kelvin wave with northeast flank traversing SEMA, where friction-driven northeasterly low-level wind and divergence emerge to block moisture penetration from the Indian Ocean. The low-level divergence is followed by descending motion in SEMA, suppressing convection and rainfall. Further, the simulated structure forced by TWI SST alone bears a close resemblance to the observed evidence, confirming the critical role of TWI. Finally, it is illustrated that ENSO and its diversity have modulating effect on SEMA precipitation as well as the coupling between TWI SST and SEMA precipitation, during the previous winter and the concurrent spring.

In this study, we used 40 years (1981–2020) of TC data and reanalysis data to investigate the relationship between the SCS TC activity frequency in May and the spring-to-summer transition of Asian monsoon systems. The results show clear decadal-scale variations of TC frequency with two active decades during the 1980s and 2000s, and two inactive decades during the 1990s and 2010s. The period of 2011–2020 was identified as a decade of the minimum TC genesis frequency over the SCS in May since 1981. The circulation and surface air temperature contrast during the earlier two decades is drastically different from the contrast during the later two decades. The difference can be understood as decadal-scale variations of two leading modes of the 40-yr March-June precipitation in the Asian-Australian-Pacific monsoon region. For the two earlier decades, the contrast of active and inactive SCS TC frequency in May can be explained by the difference in EOF2. The positive EOF2 corresponds to a wet and dry dipole pattern of the concurrent anomalies with enhanced convection over the eastern Indian Ocean and suppressed convection over the western Pacific warm pool. For the two later decades, the contrast can be explained by the difference in EOF1, which shows a meridional dipole pattern over the eastern Indian Ocean reflecting the northward movement of the ITCZ. Among four decades, the decade of 2001-2010 shows the earliest northward transition of the ITCZ and the most active SCS TC frequency in May.

During the last couple of years extreme climate-related events appear to have increased in devastating proportions in particular droughts and heatwaves. Major Rivers around the world are drying up as record breaking heatwaves take their toll. Europe recorded its worst drought in 500 years, drying up Rivers and bringing up archeological wonders and historic horrors, Roman Ruins to the surface, exposing World War II ships. Half of China hit by drought and worst heatwave in summer 2022. Siberia heating was unprecedented during the last several millennia. On the other hand, Pakistan recorded devastating and fatal floods during summer 2022. Heaviest rains in more than a century hit South Korea – worst in 115 years. While the summer monsoon rainfall during 2020 witnessed heavy rains over South Asia (in particular West-Central India) as well as over East Asia (in particular China’s Yangzte River Valley). On the other hand, summer 2022 witnessed severe floods over South Asia (Pakistan) and severe drought prevailed over East Asia (Yangzte River Valley). Possible factors leading to the contrasting behavior of Summer Monsoons 2020 and 2022 will be discussed at the AOGS2023 Conference.

The northeasterly wind under the active East Asian winter monsoon often results in the formation of stratocumulus cloud decks in the northeastern Yilan plain area and adjacent mountains in Taiwan. It is interesting to note that Yilan’s surface wind is southwesterly even though the prevailing large-scale wind is northeasterly. In some local areas in YiLan near the mountain, the stratocumulus cloud resulted in a 200 mm/day very large rainfall. To study the evolution of precipitation patterns, planet boundary layer (PBL) turbulence, and three-dimensional circulation characteristics during these conditions, the Yilan Experiment of Severe Rainfall (YESR) was conducted in November 2020, 2021, and 2022. The study utilized a rich collection of sounding observations made with the innovative and accurately calibrated Storm Tracker (ST) mini-radiosonde, using a flexible mobile strategy to understand the evolution of PBL structure during precipitating stratocumulus events, to capture the three-dimensional structure of the planetary boundary layer wind field. The continuous meteorological data collected during two northeasterly episodes during YESR2020 revealed the variability of local-scale wind patterns and the severe rainfall characteristics induced by stratocumulus clouds. Results showed the potential for a local-scale convergence line to form over the plain area of Yilan during Northeasterly conditions, with the precipitation hotspot located in the southern mountain region of Yilan where local winds displayed turbulence features. The severe rainfall of the two northeasterly episodes highlighted the presence of shallow cumulus under the stratus with pure warm rain processes. It highlights the importance of PBL structure over complex topography. Ongoing analyses of the interplay between the cooling/moistening of the sub-cloud layer, raindrop evaporation, and local circulation patterns will be presented.

This study investigates the heavily precipitating stratocumulus at the windward mountains and plains, focusing on the interaction between the terrain and the moist cloud-topped boundary layer. Such a phenomenon has been observed in the northeast of Taiwan, where rainfalls can be greater than 10 mm hr^-1 at some hotspots near the mountain under the stratocumulus-dominated environment with prevailing northeasterly in wintertime. To understand the related physical processes, we carried out large eddy simulations by the vector vorticity equation model (VVM) with idealized terrain configuration. The initial condition of the CONTROL simulation is the simplified Ishigaki Island sounding with a well-mixed boundary layer and homogeneous easterly below 1 km. The CONTROL indicates the heavy precipitation on the plain is related to the cold pool developed in the mountain region. The precipitation occurs mostly near the mountaintop, and the cold pool initiates in the foothill owing to evaporative cooling. When the cold pool intensifies, it moves upstream toward the plain, and the precipitation hotspot propagates with the cold pool. In the experiment with weaker low-level wind speed (WEAK), the cold pool develops much earlier and propagates upstream more easily. Inspired by the simulation results, field observations have been designed and carried out in November 2022 over Yilan to examine the evolution of the boundary layer structures of the precipitating stratocumulus and the signal of evaporative cooling in the sub-cloud layer. Our results highlight the significant role of the cold pool and its interaction with the background wind in the occurrence and propagation of heavy precipitation at the terrain and its nearby plain.

There are specific spatial rainfall patterns in northeast Taiwan (Yi-Lan area) under the East Asian Winter Monsoon condition. The large-scale wind field feature and the spatial distribution of water vapor flux with time can change such spatial rainfall characteristics. We analyzed the Yi-Lan area rainfall pattern variations in the past 60 years and noticed that there were obvious characteristics of terrain-locked precipitation patterns. The hot spot of rainfall occurred on the windward side of the background northeasterly flow and was decreasing toward the plain area. The average rainfall has an increasing trend, and it was more significant in the southern mountain region. We also used the reanalysis data (ERA5) to diagonalize the long-term variation of large-scale wind field features and water vapor distribution in different time periods. We noticed a significant increase in the water vapor content in the terrain's upper stream over the past 60 years. The background wind field has shifted from northerly-northeasterly to northeasterly-easterly in the past 30 years. Such changes led to a significant increase in the water vapor flux on the windward side of the terrain, which also increased the rainfall amount over the southern mountain area of Yi-Lan. However, there was not only an increasing trend in time but also showed an interdecadal oscillation signal. This decadal oscillation signal represented the background northeasterly wind speed variations. Based on the concept, we can estimate the long-term climate changes of winter rainfall in the Yi-Lan area for future climate projections by analyzing large-scale environmental variations.

Yi-Lan has a unique delta plain surrounded by mountains higher than 2000 m on three sides, and another side faces the Pacific Ocean. With a case study on 26 November 2021, this study attempts to investigate the local circulations at low levels of Yi-Lan area. With four-radar observations, Wind Synthesis System using Doppler Measurements (WISSDOM) is used to retrieve 3D wind fields. By illustrating wind structure below 1-km, different mechanisms of precipitation development can be examined. This study depicts different wind directions of the inflow to Yi-Lan causing different types of convergence and rainfall distributions. Furthermore, the inter-comparison between retrieved wind fields and ensemble simulations is conducted.

This study investigates the impact of global warming on the orographically locked diurnal convection in the summer season in Taiwan. We utilize the storyline approach to provide physically self-consistent narratives of possible future projections. The orographically-locked diurnal convection involves interactions between local circulation and the thermodynamic environment of convection, which are appropriately captured by an ensemble of TaiwanVVM large-eddy simulations. The 30 ensemble simulations are forced by a set of radiosonde observations covering the variability of the background environment. The relationships of convective updraft structures over orographic precipitation hotspots and their upstream environment are analyzed. The results reveal that strong convective updraft columns within the heavily precipitating, organized systems exhibit a mass flux profile that gradually increases with height through a deep lower-tropospheric inflow layer. Enhanced convective development is associated with higher upstream moist static energy (MSE) transport through this deep-inflow layer via local circulation, augmenting the rain rate by 35% in precipitation hotspots. In the pseudo global warming experiment set, the initial temperature profiles of all ensemble members are evenly elevated by 3 K, while the relative humidity profiles remain unchanged. The enhancement of mean precipitation intensity with respect to temperature is 6.53 %·K^-1, close to the Clausius–Clapeyron relation. Over the orographically-locked hotspots, the frequency of extreme rainfall increases by 65%, and their location expands toward the foothills and plains. The responses in the lifetime, structures, and propagation of the diurnal organized convection systems will be analyzed using the object-based tracking algorithm. The changes in buoyancy profiles and the MSE transport via local circulation will also be investigated to provide a comprehensive understanding of the physical processes influencing the orographically locked diurnal convection under climate change.

The objective of this study is to provide a conceptual framework to understand the diurnal convection over topography dominated by local circulation. The results are analyzed from the perspective of deep-inflow mixing of convection, focusing on relating convection strength with the moist static energy from the boundary layer inflow. We used the Vector Vorticity equation cloud resolving Model (VVM) to simulate the diurnal convection over an idealized terrain of a mountain island. The simulation is initialized with a simplified sounding representing typical summer weak synoptic conditions near Taiwan. We carried out two sets of sensitivity experiments by changing the mid-troposphere relative humidity and cloud condensation nuclei (CCN) concentration, respectively. In all simulations, the diurnal precipitation time series exhibits multiple peaks in time, the first two peaks, in which the developments are closely related to the evolution of sea-valley breeze circulations, show the deep-inflow mixing features, especially in the 2^nd peak with stronger local circulation and upstream energy transport. The sensitivity experiment with changing the mid-troposphere relative humidity shows that the convection in the 1^st precipitation peak reduces the environmental moisture difference, while the convection strength is more influenced by the boundary layer energy in the 2^nd peak. The environmental moisture determines the precipitation initiation time, duration of local circulation development, and the hydrometer contents in the atmosphere, which affect the low-level energy content, wind speed, and the amount of energy transport, further affecting the precipitation and convection intensity. The sensitivity experiment with changing CCN concentration confirms the effects of local circulation development prolonged by the delayed precipitation development on increasing energy transport and convection intensity. We concluded that the earlier precipitation peaks will modulate the environmental moisture, enhancing the influence of the boundary layer energy while reducing the influence of the mid-troposphere entrainment on estimating the convection strength.

The fog/low cloud during the cold season frequently blanket the eastern Taiwan mountains areas, known for the montane cloud forest. These fog/low clouds can provide significant water supplies to local areas and support the biodiversity of the montane cloud forest. However, it remains challenging to identify the appropriate temporal and spatial scales of these fog/low clouds due to limited ground observations in the complex topography. To address this issue, satellite remote sensing is involved to provide continuous, topography-unlimited temporal and spatial observations. In the view of the satellite from the top of the atmosphere, the mountain fog top can form a clear edge closely following the topographic features to represent the maximum height of the fog-occurring area. The objective of this work is to detect these mountain fog edges from the Himawari-8 satellite true-color image by applying the machine learning technique (U-net). The training data consist of the three visible bands observed at 8 am local time, with the fog edges in the fog-occurring cases serving as the training labels. The model performance is tested using true-color images at the same and different local times. The result shows that the model can capture the climatology of the fog edge hot spots at specific elevations, with an accuracy of over 93% for predicting fog-occurring cases at the same local time and nearly 70% for different local times. One of the characteristics of the current model is that the prediction has no false positive situations. The detection results using MODIS true-color images as input data will also be compared and discussed in this presentation. In the future, the images at different local times will be included in the model training, allowing for further investigation of the diurnal and inter-annual variations of fog/low cloud.

During the autumn of 2022, a record-breaking rainfall of over 7000mm was recorded at the Ximao Shan rain gauge in Yilan. The rainfall and average daily rainfall intensity were 2 to 3 times higher than the climatology in the region. To identify the daily synoptic weather types that contributed to the heavy rainfall, a machine learning-based weather typing method was utilized. The method successfully detected multiple weather types as the synoptic systems evolved. Ten days were identified as associated with remote rainfall events, caused by the interaction between the northeasterly monsoon flow and typhoons. These 10 days contributed to approximately 40% and 60% of the autumn rainfall in eastern Taiwan and the mountainous area of Yilan, respectively. The convergence of the northeasterly monsoon flow and the outer circulation of tropical storms passing through the northern South China Sea also led to enhanced rainfall in the mountainous area of Yilan. In the La Niña year of 2022, the environment may have been favorable for such weather types. Our research shows northern Taiwan's average autumn rainfall is 1.3 times higher in La Niña years compared to the climatology. Furthermore, the rainfall from remote rainfall events in 2022 was 1.6 times higher than the average of such events in other La Niña years, but the rain intensity was similar. The 10 days of remote rainfall events in 2022 were significantly more than the average of 3 event days during other La Niña years, resulting in record-breaking rainfall.

The compact, low-cost Storm Tracker was co-launched with commercial sounding systems at Jeju Island, South Korea, in the summer of 2021. The Storm Tracker collected high temporospatial profiles of temperature, pressure, humidity, wind directions, and wind speed. The performance of these observational metrics will be evaluated and compared with Vaisala RS-41 and GRAW DFM-09. There were 32 launches, and the quality of Storm tracker data has been controlled by AIQC algorisms, which are developed by training huge amounts of launches in advance at subtropical regions. The main objective of this study is not only to check Strom Tracker's performance but also to understand potential variances when launched in mid-latitude areas. The average pressure bias between RS-41 and DFM-09 sounding systems is ~0.75 hPa and ~3hPa with the Strom Tracker. For temperature, the average bias is 0.57^oC between RS-41 and DFM-09, with ~2^oC (0.6^oC) differences in the daytime (nighttime) compared with the Strom Tracker. The dew point reveals ~5^oC (2^oC) variances between the Strom Tracker and RS-41 (DFM-09), and ~2^oC between RS-41 and DFM-09. The wind speed and direction discrepancies observed from the Storm Tracker are ~2 m s^-1 and from 10 to 20^o compared with the other two sounding systems. Generally, the results show good agreement between these three sounding systems. However, a significant bias of temperature in the daytime should be paid more attention to with the Strom Tracker.