
HomeEssential information & links to all sections of the website. |
Program OverviewProgram summary.Gives the big picture. |
Scientific ProgramAll activities scheduled at the meeting. |
ShowcaseExhibitors’ recent innovations & industry advancements. |
IG01-A002
Divergent Response of Vegetation Photosynthesis to the Record-breaking Meiyu Event in 2020
Jiuyi CHEN#+, Bo QIU, Weidong GUO, Xin MIAO, Lingfeng LI
Nanjing University, China
Extreme precipitation events have posed a threat to global terrestrial ecosystems in recent decades. However, the response of terrestrial ecosystems to extreme precipitation in areas with various vegetation types and complex topography remains unclear. Here, we used satellite-based solar-induced chlorophyll fluorescence measurements, a direct proxy of photosynthetic activity, to assess the response of vegetation to the record-breaking extreme precipitation event during the East Asia monsoon season in eastern China in 2020. Our results demonstrate that vegetation was adaptable to moderate increases in precipitation, but photosynthetic activity was significantly inhibited by exposure to extreme precipitation because of insufficient photosynthetically active radiation and waterlogging. The responses of vegetation photosynthesis to extreme precipitation were regulated by both vegetation type and topography. Crops in the lowland areas in eastern China were severely damaged due to their higher vulnerability and exposure to extreme precipitation. The topography-induced redistribution of precipitation accounts for the modulation of vegetation response to extreme precipitation. Our research highlights the urgent need for effective management and adaptive measures of croplands under the elevated risk of extreme precipitation in the future.
IG01-A005
Mobile Network Data Analytical Modules of Population Movement Trend Around Points-of-interest Areas
Chia-Jung WU#+, Ming-Wey HUANG
National Science and Technology Center for Disaster Reduction, Taiwan
This study aims to improve the accuracy of risk assessment for the population affected by disasters, such as earthquakes, by cooperating with the government and using mobile-phone data provided by the telecommunication company. The mobile-phone data can reveal the temporal and spatial patterns of people’s movement and location in different areas. This can help estimate the number of affected people more precisely and identify abnormal crowd flow at specific locations. Based on this information, the competent authority can send real-time warning messages to those who are at risk. The mobile-phone data is presented in a grid unit of 500m×500m, and an analytic model is developed to compute the difference after an emergency event. We analyze the population changes after two natural events (one earthquake and one typhoon) and find that there are significant variations after the disasters. The comparison between the population of the background and those after the events can show the impact of the disasters. The change of mobile-phone data for the points of interest is consistent with the actual number of affected people reported by the government official at the time of the disaster. This confirms that the mobile-phone data can be a valuable tool to assist the commander in making decisions when responding to disasters.
IG01-A007
Dynamic Assessment of Spatiotemporal Population Distribution Based on Mobile Phone Data: A Case Study in Xining City, China
Benyong WEI#+, Guiwu SU, Wenhua QI
China Earthquake Administration, China
High-resolution, dynamic assessments of the spatiotemporal distributions of populations are critical for urban planning and disaster management. Mobile phone big data have real-time collection, wide coverage, and high resolution advantages and can thus be used to characterize human activities and population distributions at fine spatiotemporal scales. Based on six days of mobile phone user-location signal (MPLS) data, we assessed the dynamic spatiotemporal distribution of the population of Xining City, Qinghai Province, China. The results show that strong temporal regularity exists in the daily activities of local residents. The spatiotemporal distribution of the local population showed a significant downtown-suburban attenuation pattern. Factors such as land use types, holidays, and seasons significantly affect the spatiotemporal patterns of the local population. By combining other spatiotemporal trajectory data, high-resolution and dynamic real-time population distribution evaluations based on mobile phone location signals could be better developed and improved for use in urban management and disaster assessment research.
IG01-A014
Research on Seismic Risk Assessment Based on Residential Building Stock and Field Survey Results: A Case Study of 3 Cities, Shaanxi Province
Xia CHAOXU#+, Nie GAOZHONG, Wenhua QI, Li HUAYUE
China Earthquake Administration, China
The collapse and damage of buildings caused by earthquakes are the main reasons for casualties. High-precision building data is the key to improving the accuracy of earthquake disaster loss risk assessment. Although the field investigation method based on "township to township" can obtain more accurate building inventory data, considering the economy and timeliness, it is necessary to consider combining remote sensing and other diverse data to achieve the acquisition of building data. Based on the field survey data, combined with the Global Human Settlement Layer (GHSL) data, this paper obtains the proportion data of each type of building and its lethality level in each township based on the classification of building height, and realizes the calculation of the overall lethality level of the building level and township level on this basis. It is found that the fitting results between the calculated results and the field survey results are good, the error is within 0.15, and the fitting results R2 of Xian, Baoji and Ankang are 0.6552, 0.5788 and 0.5937 respectively. Based on this, this paper carries out the earthquake disaster loss risk assessment based on the building level, and finds that the risk of casualties caused by the same type of buildings in each city is different. Generally, the areas with high disaster loss risk in three cities are mainly distributed in urban areas, while the disaster loss risk in newly built areas of each city is relatively low, it can be found that Xi'an has the highest loss risk, while Baoji and Ankang are at the same level. Based on the method constructed in this paper, we can realize the quantitative assessment of earthquake disaster loss risk at the building level, so as to make the pre earthquake emergency preparation and post earthquake auxiliary decision-making more targeted and accurate.
IG01-A023
Urban Ground Deformation of the Greater Manila Area from 2014 to 2022 Using InSAR Time Series Analysis
Jolly Joyce SULAPAS1,2#+, Alfredo Mahar LAGMAY3, Audrei Anne YBAÑEZ1, Kayla Milcah MARASIGAN1, Julian Bernice GRAGEDA1
1University of the Philippines Resilience Institute, Philippines, 2University of the Philippines Nationwide Operational Assessment of Hazards (UP-NOAH), Philippines, 3University of the Philippines Diliman, Philippines
Land subsidence has become a significant global issue caused by either natural or anthropogenic factors, such as excessive groundwater extraction, rapid urbanization, and natural sediment compaction, and is exacerbated by climate change through rising sea levels. In the Philippines, it is recognized as a hazard that threatens Metro Manila and other urban areas. This paper presents updated vertical ground motion rates using Sentinel-1 interferometric synthetic aperture radar (InSAR) time series analysis from 2014 to 2022 through UK COMET’s LiCSAR products and the LiCSBAS package of Morishita et al. (2020). The results revealed a maximum subsidence rate of 10.8 cm/year in Bulacan Province coinciding with industrial and commercial complexes and evident as a contiguous distribution of large and expansive man-made structures. Such subsiding areas are observed as circular to elliptical deformation features in vertical motion maps co-located with zones of compressive motion in horizontal motion maps, most likely related to movements toward areas of peak subsidence. Additionally, these are usually centered on economic zones, tightly-packed residential areas, and technoparks. Towards the east and south of Metro Manila, elongated deformation is associated with the Valley Fault System where left-stepping en echelon faults and horst and graben occur. This is where damages to infrastructure (houses, buildings, road pavements, and walls) have been observed since the 1990s. There is a likely correlation between the splays and ground fissures of the fault and drawdown-induced subsidence. Monitoring this hazard is crucial as it increases the risk of floods, building and infrastructure damage, and economic loss, and exposes residents along the coast to worsening tidal incursions and storm surges due to climate change.
IG01-A026
Developing the Dynamic Crop Calendar of Paddy in Indonesia from 2001 - 2021
Amalia Nafisah Rahmani IRAWAN#+, Daisuke KOMORI
Tohoku University, Japan
According to the Food and Agriculture Organization (FAO), agricultural sectors absorb around 80% of drought’s direct impact. With the climate change impact indicated by the increase of temperature and the rainfall patterns that become more variable, could lead to more severe drought in some regions. This condition could increase the agricultural drought risk that might adversely affect food security. To improve the crop condition monitoring, crop management, and increase the crop production under changing climate, many researchers have developed the crop calendar dataset. The crop calendar contains information about when and where a specific crop is planted and harvested. However, there are two remaining gaps related to this crop calendar: 1) It was developed based on the specific period and then provided as a fixed calendar, thus cannot catch the diversity of cropping season under changing climates, and 2) It was provided on coarse spatial resolution, so it cannot capture the spatial variations, especially on the fragmented agricultural area. This study was conducted to develop the 1-km annual dry season crop calendar of paddy in Indonesia from 2001 – 2021.The crop calendar was developed using a model-based by utilizing the Normalized Difference Vegetation Index (NDVI) to monitor crop greenness that can identify crop growth. The NDVI dataset was obtained from Moderate Resolution Imaging Spectroradiometer (MODIS) through their MOD13A2 product with 16-days intervals from January 2001 – December 2021 with a total of 483 images. Then, the reconstruction of the NDVI timeseries was conducted using the ST-Tensor method, which is suitable for a cloudy and rainy region like Indonesia. After the agricultural area was isolated based on the land cover map from the Ministry of Forestry and Environment, the start of season (planting date), the peak of season, and the end of season (harvesting date) were determined.
IG04-A003
Interpretable Predictions of Chaotic Dynamical Systems Using Dynamical System Deep Learning
Mingyu WANG+, Jianping LI#
Ocean University of China, China
Making accurate predictions of chaotic dynamical systems is an essential but challenging task with many practical applications in various disciplines. However, the current dynamical methods can only provide short-term precise predictions, while prevailing deep learning techniques with better performances suffer from model complexity and interpretability. Here, we propose a new dynamic-based deep learning method, namely the dynamical system deep learning (DSDL), to achieve interpretable long-term precise predictions by the combination of nonlinear dynamics theory and deep learning method. As validated by three chaotic systems with different complexities, the DSDL significantly outperforms other dynamical and deep learning methods. Furthermore, the DSDL also reduces the model complexity and realizes the model transparency to make it more interpretable. We firmly believe that the DSDL is a promising and effective method for comprehending and predicting chaotic dynamical systems in the real world.
IG04-A005
Robust Prediction of Chaotic Systems with Random Errors Using Dynamical System Deep Learning
Zixiang WU+, Jianping LI#
Ocean University of China, China
To predict nonlinear dynamical systems, we recently proposed a novel method called Dynamical System Deep Learning (DSDL), which is based on the theory of dynamical system reconstruction and relies on data for model training. The robust prediction of real-world dynamical systems using observational data with errors is a practical issue. This study primarily investigates the robustness of DSDL in the presence of random errors in real training data. The performance of DSDL is tested on two example systems, namely the Lorenz system and the Kuramoto-Sivashinsky system. The results demonstrate that compared with different traditional deep learning methods and the fourth-order Runge-Kutta method (RK4), respectively, DSDL exhibits high accuracy and stability even in the presence of errors. In terms of accuracy, the DSDL results are comparable to the results of numerical solution method taking RK4 as an example and outperform the results of traditional deep learning. Notably, as the magnitude of errors decreases, the advantages of DSDL become more pronounced, indicating its ability to effectively utilize dynamic information within data sets. Moreover, unlike traditional methods, DSDL does not rely on original equations or hyperparameter optimization and is insensitive to random errors present in training sets. This dynamic-driven method provides greater potential for enhancing the predictive capabilities when analyzing Earth system observational data.
IG04-A007
Insights Into Indian Agriculture: A Deep Learning Approach to Understand the Climatic and Anthropogenic Influences on Rice Yield
Nairit SARKAR#+, Sujata RAY
Indian Institute of Science Education and Research Kolkata, India
Agricultural production, characterized by a prolonged cycle, is susceptible to various internal and external uncertainties. Variations in rainfall, extreme temperatures, fertilizer consumption, and irrigated areas can significantly impact crop yields. Rice serves as the predominant dietary staple for more than 60% of the world's population, playing a crucial role in ensuring food security. India, a significant contributor to world rice production with a 21.5% share, necessitates a comprehensive examination of the possible consequences of climate change and anthropogenic contributions to its rice yield. This study examines the agricultural dynamics of rice production in Indian districts, focusing on the trends of climatic factors and anthropogenic inputs. The non-parametric Mann-Kendall (MK) test, its modified version, and the Theil-Sen estimator were employed to identify and quantify significant trends. The study showed that despite declines in gross cropped area (GCA) across most districts, positive trends in production prevailed due to increased yield. Spearman's correlation test was employed to investigate one-to-one correlations between the factors and the rice yield to detect the most dominant contributor. The correlation test indicates that elevated fertilizer consumption and increased fraction of irrigated land in many districts significantly correlate to the rise in yields. Despite notable climatic changes over the past three decades, the correlation between rice yield and variations in annual and seasonal rainfall and extreme temperatures has proven mostly insignificant. The multivariate analysis using Artificial Neural Network (ANN) provided more insights than traditional correlation coefficients in finding the key contributing factors. The top three models, which exhibited the best fit, collectively established that anthropogenic inputs exerted the most substantial influence on the fluctuations in rice yield in India over the past three decades.
IG04-A009
Microearthquakes Detection Using a Machine Learning Analysis in Gangwon Province, South Korea
Byeongwoo KIM1#+, Minhwan KIM2, Tae-Kyung HONG1
1Yonsei University, Korea, South, 2Kyung Hee University, Korea, South
Gangwon province is located in the northeastern South Korea and characterized by mountainous terrains. The region experienced strong ground motions from a magnitude 4.8 earthquake in 2007, bringing concerns on potential seismic damage. Detection of scattered microearthquakes is challenging for conventional methods such as short-term to long-term average ratio analysis and matched filter analysis of which sensitivity is limited by ambient noise and station proximity. In this study, we detect offshore microearthquakes. We employ a machine learning-based approach for microearthquake identification to overcome the limitation. Utilizing the trained PhaseNet machine learning model, we successfully detect P and S phases in the seismic records and identify microearthquakes based on coherent phase arrivals across stations. We collect seismic data from 175 stations deployed in the northeastern area of South Korea since 2021. We select areas of relatively high seismic activity. We additionally collect seismic data from ~80 temporally-deployed geophones since April 2023. We observe around 70 microearthquakes every month. The microseismicity is high in the Hongcheon area and coastal regions of the East Sea (Sea of Japan). This study presents numerous microearthquakes in regions characterized by low seismicity, demonstrating the effectiveness of the method. Subsurface blind fault structures are illuminated by the microseismicity.
IG04-A012
Improved Statistical Downscaling for Short-term Forecasting of Summer Air Temperatures Based on Deep Learning Approach
Dongjin CHO#+, Jungho IM, Sihun JUNG
Ulsan National Institute of Science and Technology, Korea, South
Reliable early forecasting of extreme summer air temperatures is essential for effectively preparing and mitigating the socioeconomic damage caused by thermal disasters. Numerical weather prediction models have become valuable tools for operational forecasting air temperature. However, they incur high computational costs, resulting in coarse spatial resolution and systematic bias owing to imperfect parameterization. To address these problems, we attempted to develop a novel statistical downscaling and bias correction method (named DeU-Net) for the maximum and minimum air temperature (Tmax and Tmin, respectively) forecasts obtained from the Global Data Assimilation and Prediction System with a spatial resolution of 10 km to 1.5 km over South Korea through the fusion of deep learning (i.e., U-Net) and spatial interpolation. In this study, we used a methodology to decompose statistically downscaled Tmax and Tmin forecasts into temporal dynamics over South Korea and spatial fluctuations by pixels. When comparing the proposed DeU-Net with the dynamical downscaling model (i.e., Local Data Assimilation and Prediction System) and support vector regression (SVR)-based statistical downscaling model at the seen and unseen stations for forecasting the next-day Tmax and Tmin, respectively, DeU-Net showed the highest spatial correlation and the lowest root mean square error in all cases. In a qualitative evaluation, DeU-Net successfully produced a detailed spatial distribution most similar to the observations. A further comparison extending the forecast lead time to seven days indicated that the proposed DeU-Net is a better downscaling approach than SVR, regardless of the forecast lead time. These results demonstrate that bias-corrected high spatial resolution air temperature forecasts with relatively long forecast lead times in summer can be effectively produced using the proposed model for operational forecasting.
IG04-A016
Characteristics of Soil Geochemical Anomalies and Prospecting Prediction in Baixingtu Area, Alxa Right Banner, Inner Mongolia
Wen GAO+, Xianrong LUO#, Jiewei LI
Guilin University of Technology, China
In order to preferably select the prospecting prediction target area in Baixingtu area of Alxa Right Banner, 1:10,000 soil geochemical measurements were carried out on the basis of identifying the geological conditions of mineralization in the area. Elemental variation coefficients, ILR-Robust transform and ILR-PCA were used to analyze seven elements of Au, Ag, Co, Cu, Pb, Zn and As in the measuring area. The results show that: (1) ore-forming elements are mainly located at the intersection between the fracture and mineralized alteration zone. Au and Ag show very strong differentiation and enrichment characteristics with the highest mineralization potential, while Pb, Zn, As, Co and Cu show low differentiation and enrichment characteristics, and low mineralization potential. (2) ILR-Robust transformation eliminates the closure effect of the original data and makes the elements show standard normal distribution; the first principal component of ILR-PCA can extract the comprehensive information of the elements of the original data to the maximum extent, i.e., the positive end-load of PC1 (Au-Ag-As) is the mineralization indicator, and the intensity and morphology of elemental anomalies are controlled by the fracture and mineralization alteration zone, and the negative end-load (Co-Cu-Pb-Zn) is caused by the enrichment of elements in multi-phase magmatic hydrothermal activity. (3) Engineering verification of the mineralization target area No. Ⅰ has identified several gold (silver) ore (mineralization) bodies, and the gold and silver contents have reached the minimum industrial grade or above, indicating that the Baxingtu area has a large potential for gold and silver prospecting.
IG04-A018
Long-term PM2.5 Concentration Prediction Under Climate Changes Using Geo-AI Based Machine Learning Model
Wei Lun HSU1+, Yu Ju LIN1, Pei-Yi WONG1, Yu-Ting ZENG1, Shih-Chun Candice LUNG2, Chih-Da WU1#
1National Cheng Kung University, Taiwan, 2Academia Sinica, Taiwan
Air pollution is a pressing issue, threatening human health and societal development. The likely increase in air pollution due to climate change underscores the intricate interplay between air quality and environmental challenges in the 21st century. The objective of this study is to assess the changing trend of PM2.5 levels in response to climate change. In order to predict PM2.5 concentrations in Taiwan's mainland, where air quality stations are limited, this research combines the principles of land use regression and harnesses the capabilities of machine learning algorithms. The machine learning model incorporates various factors, including land use elements and other pertinent variables, enabling the capture of nonlinear relationships. Data comprising daily average PM2.5 observations and meteorological information spanning from 1994 to 2019 were gathered, with additional consideration given to related variables such as land use inventory and landmark data. Five machine learning algorithms, namely eXtreme Gradient Boosting (XGBoost), Gradient Boosting Machine (GBM), Light Gradient Boosting Machine (LGBM), Categorical Boosting (CatBoost), and random forest, were utilized to predict the PM2.5 concentration. The models' performance was evaluated through multiple validation processes. To project long-term PM2.5 levels based on the climate scenarios outlined in the IPCC Sixth Assessment Report (AR6), data on future projected temperature and precipitation will be obtained from the Taiwan Climate Change Projection Information and Adaptation Knowledge Platform (TCCIP). The projections for long-term PM2.5 concentration under climate scenarios SSP3-7.0 were conducted using climate data simulated by global climate models (GCMs), including MIROC6 and MPI-ESM1-2-LR. The analysis revealed that PM2.5 concentrations did not exhibit a definitive correlation with various climate scenarios.
IG04-A019
A Ship Speed Prediction Model Based on Meteorological Ocean Element Data
Jin ZHUYU#+
Chinese Academy of Meteorological Sciences, China
In order to more accurately predict the speed loss of ships under various sea conditions, a real-time vessel speed prediction model based on ensemble learning algorithm was proposed. Eighteen trans-Pacific vessel AIS speed data from China to North America were selected, and the wind speed, effective wave height, sea surface velocity, wave period and other elements of ERA5 and HYCOM data were selected as training features. The random forest method was used to calculate the relative importance of many input features. The selected features were trained into the ensemble learning model: LightGBM, which is a lightly and higher efficiency machine learning model compare with others, and the AIS speed data of one voyage was selected as the test set to verify the model effect. The results show that compared with the traditional stall equation based on the empirical formula, the ensemble learning model based on data-driven has better prediction performance. The RMSE of the predicted speed and the real vessel speed is 0.87, and the RMSE of the empirical equation is 2.02, the RMSE improvement rate is 56.9%. The running time of the model can reach the minute response, which meets the requirements of real-time ship speed prediction. The vessel speed prediction model with high accuracy and strong real-time performance can provide a theoretical basis for ships to take reasonable measures to avoid severe weather such as strong wind and waves in the future, which is of great significance to the improvement of ship navigation safety and the operation efficiency of vessels and related ports.
IG04-A028
Development of a High-resolution Dataset Based on the C-LSAT 2.0 Land Surface Air Temperature Dataset
Sihao WEI#+, Qingxiang LI
Sun Yat-sen University, China
Land surface air temperature is a crucial indicator of climate change and a significant focus of global and regional climate research. Scientists have long been committed to researching the homogenization of station data and the development of grid datasets. Several benchmark grid datasets of global land surface air temperature have been available since the instrumental era, including CRUTemp, GHCN, GISTEMP, BEST, and C-LSAT. These datasets have been crucial in supporting research related to global climate change. However, due to their rough resolution, many details of regional climate change are difficult to reflect. Based on the global land surface air temperature dataset C-LSAT 2.0 independently developed by our team, we develop a global long-time series (1901-2022) with high spatial resolution (0.5° × 0.5°) land surface air temperature grid dataset C-LSAT HR. The C-LSAT 2.0 station data undergoes quality control before being decomposed into a climatology field (1971-2000) and an anomaly field. The climatology and anomaly fields are then interpolated into 0.5° × 0.5° gridded data using partial thin-plate spline (TPS) and Adjusted Inverse Distance Weight (AIDW), respectively. Finally, the two fields are merged into C-LSAT HR. In the field of climatology, the global average mean absolute error (MAE) and root mean square error (RMSE) of C-LSAT HR are 0.263 °C and 0.439 °C, respectively. Regarding the anomaly field, the global average annual MAE and RMSE are approximately 0.4-0.6 ℃ and 0.6-0.9 ℃, respectively. The global land surface air temperature anomaly series based on C-LSAT HR and C-LSAT 2.0 exhibit a high level of consistency. The C-LSAT HR indicates that the average warming trend of each season in the Northern Hemisphere is higher than that in the Southern Hemisphere. Additionally, the average warming trend in winter and spring is higher than that in summer and autumn.
IG04-A030
Improving Deep Learning-based Ground Deformation Detection and Extraction from InSAR Images with Noisy Interferograms Synthesized by Generative Adversarial Networks
Xuekai LIN#+, Caijun XU
Wuhan University, China
With the ever-increasing volume and accessibility of InSAR images, the automatic detection and extraction of ground deformation signals from various tectonic or geological processes (e.g., earthquakes, volcanos, and landslides) have become a pressing and valuable task. Supervised deep learning approaches have shown great promise in this regard. However, these geological events occur infrequently, making it impossible to obtain a sufficient number of interferograms containing these ground deformation signals from the real world for deep learning model training. Consequently, most previous studies resort to using synthetic interferograms, which are a combination of synthetic deformation and simulated noise signals, to train the models. Unfortunately, the application of these models to real-world data is not yet satisfactory due to the gap between the distribution of synthetic and real data. While the number of real interferograms (positive samples) containing effective deformation signals is scarce, interferograms containing only noise (negative samples) are readily available. Generative Adversarial Networks (GAN), where a generator and a discriminator are included, are known for their ability to learn the distribution from real samples and produce realistic synthetic samples. In this work, we constructed a conditional GAN model with real-world noisy interferograms. In the GAN model, the generator produces synthetic noisy interferograms based on randomly generated vectors and topographic data to confuse the discriminator. In this way, more realistic noise signals that have the potential to bridge the distribution gap between the synthetic and real-world samples can be generated. Subsequently, we train deep learning models for ground deformation detection and extraction using these synthetic noisy interferograms as well as simulated deformation signals. Performance improvement could be observed when applying the model trained with interferograms synthesized by GAN to detect and extract deformation signals from real-world InSAR observations.
IG06-A004
Geochemical and Thermal Characteristics of Hot Spring Waters in the Japanese Island Arc
Risako KIMURA#+, Hitoshi TOMARU
Chiba University, Japan
A large number of hot springs are widely distributed in Japan, closely associated with the development of a geothermal regime in the island arc system. These areas are geologically important for understanding the fluid-rock reactions, material circulation processes, and the utilization of geothermal resources. In this study, we analyzed the chemical and isotopic composition of hot spring waters collected from various locations across the Japanese island arc to characterize the geochemical features and discuss the origin and geothermal history of these waters. We have employed hexadiagrams and trilinear diagrams in the classification of hot spring waters. Additionally, geothermometers with Na-K, Na-K-Ca, and SiO2 concentrations were used to estimate the subsurface temperatures. Hot spring waters exhibiting higher estimated temperatures are generally of meteoric water origin and enriched in Na-HCO3, Ca-SO4, Ca-HCO3, and Na-SO4. These waters are largely influenced by volcanic gases, volcanic fluids, and the thermal energy derived from recent volcanic activities. In contrast, hot spring waters with relatively lower temperatures, enriched in Na-Cl, are considered to have interacted with seawater and fossil seawater, often sourced from non-volcanic areas like sedimentary basins and coastal regions. The geochemistry of hot spring waters primarily reflects their interactions with rocks, especially in high-temperature environments, leading to active material circulation. It serves as a significant indicator of subsurface processes essentially linked to the development of the island arc system.
IG06-A005
Advancing Groundwater Level Prediction for Irrigation by Using the XGBoost-based Approach
Sheng-Wei WANG#+, Yen-Yu CHEN, Li-Chiu CHANG
Tamkang University, Taiwan
Groundwater resources play a pivotal role in sustaining agriculture, and predicting groundwater levels is essential for effective water resource management. Agriculture faces significant impacts from climate change-induced alterations in rainfall patterns, especially during drought. The physical models and machine learning methods are the most popular approaches for groundwater level prediction, this study highlights the efficiency of the latter as an emerging technology, providing a viable alternative that demands less time and extensive hydrogeological data. Hence, this research employs the XGBoosting method to predict groundwater levels, incorporating pumping well electricity consumption, precipitation, and current groundwater levels as model features. The research zone around a groundwater observation well in central Taiwan encompasses 2,418 pumping wells and a rainfall observation station within a five-kilometer radius. In regions characterized by dense irrigation for agriculture, obtaining precise groundwater extraction data proves challenging. To address this, we utilize electricity consumption data from pumping motors as a quantifiable representation of pumping behavior. Monthly data from 2007 to 2019 and 2020 to 2021 were used for model training and testing, and predictions, respectively. R2 values of 0.99 and 0.60 are achieved for training and testing, with 0.72 for predictions, accompanied by an RMSE of 1.27. Notably, the model exhibits stability for up to 18 months. This study offers a valuable framework for proactive groundwater management, presenting a promising tool for sustainable agricultural water use. The findings underscore the importance of accurate groundwater predictions in fostering agricultural sustainability and resilience to climate change.
IG06-A008
Ground Subsidence and Hydrological Properties in Yangon, Myanmar
Saw MYAT MIN1#+, Yu WANG1,2, Nina LIN1, Win NAING3
1National Taiwan University, Taiwan, 2Nanyang Technological University, Singapore, 3University of Yangon, Myanmar
Yangon is the former capital city of Myanmar, hosting ~ 2.75 % of the total population (55 millions of people) in the country. The city is bounded by the Yangon River to the south and the Bago River to the east, with the Shwegondaing-Mingalardon anticline sandwiched in between. In the past century, the city of Yangon has gradually expanded from its core area above the anticlinal crest to the adjacent low-lying flood plain. Owing to the rapid population and economic growth in the past decade, the public water supply system failed to cover the whole metropolitan, leaving a large proportion of residents dependent on groundwater from private wells. The lack of proper groundwater regulations and the increased extraction of groundwater from the low-lying flood plain raise the risk of land subsidence and flood hazards in Yangon. This study aims to derive the long-term ground deformation of Yangon by using multi-temporal interferometric synthetic aperture radar (MTInSAR) techniques. We obtain displacement time series from the L-band ALOS data for the period of 2007-2009 and from the C-band Sentinel-1 data for the period of 2015 to 2022. Our result indicates higher subsidence rates (up to 2.8 cm/yr) in the eastern flood plain area of Yangon during 2007-2009, which further increased to more than 10 cm/yr during the period of 2015-2022. In the western flood plain area, the subsidence rate is a few millimeters per year during 2015-2022. For the downtown area, the ground subsidence rate is less prominent due to the Pleistocene formations exposed along the anticline. The subsidence record obtained in the western part of the city agrees with the long-term trend of the only groundwater head record in this area. Together with the aquifer system constructed from hydrological wells, we estimate the first geodesy-based hydrogeological properties for the Yangon area.
IG06-A009
Combining Convolutional Neural Network and Hydraulic Tomography for Hydraulic Heterogeneity Estimation
Zi-Yan CHOU+, Jui-Pin TSAI#
National Taiwan University, Taiwan
Characterizing the heterogeneity of hydraulic parameters (K or Ss) in the subsurface is crucial for contaminated site remediation. Hydraulic Tomography (HT) is a well-developed approach for estimating three-dimensional hydrogeological parameters fields by capturing variations of groundwater head stimulated by pumping/injection events. The head variations from HT are converted to three-dimensional parameter fields using successive linear estimation (SLE), a well-proven geostatistical inverse method in various scale problems. However, the computational efficiency decreases as the number of grids or head observations increases. To overcome this issue, we propose an HT-based convolutional neural network (HT-NN) to replace SLE for converting groundwater head variations into hydraulic heterogeneity. We generate 10000+ random fields and produce their corresponding head observations using a forward model as input and output data pairs for training HT-NN. The results show that HT-NN successfully converts the groundwater head variations into hydraulic parameters fields. The proposed HT-NN can be an efficient and alternative tool to depict the subsurface heterogeneity.
IG06-A010
An Impact Estimation of Irrigation and Groundwater Pumping on the Regional Hydro-climate Using an Earth System Model
Yusuke SATOH1#+, Yadu POKHREL2, Hyungjun KIM1,3, Tokuta YOKOHATA4
1Korea Advanced Institute of Science and Technology, Korea, South, 2Michigan State University, United States, 3The University of Tokyo, Japan, 4National Institute for Environmental Studies, Japan
Irrigation is an anthropogenic forcing to the Earth-system that alters the water and heat budgets at the land surface, leading to changes in regional hydro-climate conditions. It is imperative to better understand the nature, extent, and mechanisms through which irrigation affects the Earth's system. However, despite its increasing importance, irrigation remains a relatively nascent component in the Earth-system modeling community, necessitating advancements in modeling and a deepened understanding. Our research aims to improve the quantitative understanding of the impacts of irrigation and groundwater use as anthropogenic drivers on regional climate and environmental changes. To this end, we developed an improved Earth-system modeling framework that is based on MIROC-ES2L (Hajima et al 2020 GMD) coupled with hydrological human-activity modules (Yokohata et al. 2020 GMD), enabling the simulation of a coupled natural-human interaction. Employing this Earth-system model, we carried out a numerical experiment utilizing an AMIP set-up. Here, our ensemble simulation allows for statistical quantification of the irrigation impact differentiating them from the uncertainties arising due to natural variability. We have identified regions and seasons where irrigation exerts a discernible influence on regional hydro-climate. Notably, our results show substantial disparities—larger than or comparable to inter-annual variability—between simulations incorporating and excluding the irrigation process, particularly in heavily irrigated regions such as Pakistan and India. Our model demonstrates that the introduction of moisture into the soil through irrigation alters the hydrological balance of the land surface, consequently influencing the overlying atmosphere. Furthermore, our study delves into estimating regional variations in the contributions of groundwater and surface water use to these impacts. Emphasizing the importance of a more nuanced understanding of regional characteristics in irrigation impact assessments, our research underscores the significance of coupled earth system models in comprehending and predicting the intricate interplay between human activities and the Earth's climate system.
IG06-A012
Data-driven Investigations Into Land Subsidence Evolution and Its Impacts on Infrastructures in Choushui River Fluvial Plain
Thai Vinh Truong NGUYEN+, Chuen-Fa NI#, I-Hsien LEE, Gumilar Utamas NUGRAHA
National Central University, Taiwan
Severe land subsidence in Taiwan’s Choushui River Fluvial Plain (CRFP), primarily triggered by long-term groundwater extraction for agricultural irrigation, poses a major environmental threat. This study integrates data from an extensive range of monitoring systems, such as groundwater level monitoring wells, Global Positioning System (GPS) stations, survey leveling benchmarks, multi-level compaction monitoring wells, and Interferometry Synthetic Aperture Radar (InSAR), to investigate the complex patterns and causes of subsidence in this important agricultural region. The analysis demonstrates a significant correlation between the decline in groundwater levels and the rise in subsidence magnitudes. This relationship is particularly evident in areas with a high concentration of fine-grained sedimentary materials, where sediment compaction is most significant within the first 150 meters of depth. In addition, the study examines 292 Sentinel-1 SAR images using the SBAS-PSInSAR technique, revealing significant sinking in the southern part of CRFP, which is strongly associated with regions containing high percentages of fine-grained sediment. These findings emphasize the harmful impact of subsidence on essential infrastructures, particularly the high-speed railway, highlighting the necessity of ongoing monitoring and the integration of multiple monitoring approaches for efficient subsidence management. The study highlights the intricate nature of land subsidence in the CRFP and the ongoing need to expand and maintain land subsidence monitoring networks.
IG06-A013
Evaluation of Baseflow Dynamic Through Hydrological Signatures in Taiwan
Hsin-Yu CHEN#+, Hsin-Fu YEH
National Cheng Kung University, Taiwan
Baseflow comes from delayed sources associated with rainfall events and plays a critical role in sustaining streamflow and river ecosystems during non-rainy periods. Taiwan is facing increasingly severe water challenges, including droughts and floods due to extreme weather events. Evaluating baseflow is benefit for effective ecological and water resource management. However, baseflow cannot be directly observed and measured. To address this, this study employed baseflow separation methods combined with hydrological signatures to assess dynamic baseflow characteristics. This study utilized three separation streamflow approaches—UKIH, Lyne-Hollick, and Eckhardt algorithms—to quantify baseflow. Eleven baseflow dynamic signatures included baseflow index (BFI), frequency of high-baseflow and low-baseflow days, seasonality ratio (SR), representative baseflow percentiles (Qb5, Qb33, Qb50, Qb66 Qb95), concavity index (CI) and slope (SBDC) of baseflow duration curve, were used to characterize baseflow. This study investigated the spatial distribution of baseflow signatures and analyzed the factors influencing these signatures to get the following findings (1) a comparative analysis of different baseflow separation algorithms, (2) explanation of spatial regional patterns and upstream-downstream evolution of baseflow signatures, and (3) identification of significant influential factors on baseflow dynamics. These insights into baseflow processes and influencing factors are beneficial for enhancing environmental resilience and advancing water resource management in Taiwan catchments.
IG06-A015
Long-term Analysis of Hydrological Sensitivity and Memory Characteristics in Taiwan Catchments
Ting-Jui FANG#+, Hsin-Yu CHEN, Hsin-Fu YEH
National Cheng Kung University, Taiwan
In recent years, the world has been grappling with the impacts of abnormal and extreme weather events, and Taiwan is no exception. The unusual frequency of typhoons and droughts has resulted in uneven spatiotemporal distribution of precipitation, contributing to extreme climate disasters. The water resource in Taiwan is quite hard to store due to the rapid-changing river flow and the steep topography. Moreover, prevention of hydrological disasters such as floods and droughts become significant challenges. Notably, when considering the current climate factors, the inherent memory ability of the catchments itself is often neglected. The significance of streamflow within the catchment's memory reflects the importance of water storage and propagation in the hydrological system. Thus, understanding the degree of control over changes in streamflow becomes crucial. To better comprehend the behavior of Taiwan's catchments in response to external climate factors and internal memory characteristics, this study conducts an analysis across 67 catchments with data spanning over 30 years. In terms of the impact of climate factors, the study employs the relative streamflow elasticity coefficient to quantify the sensitivity of catchment streamflow to precipitation and potential evapotranspiration. This study understands how streamflow responds to external climate conditions. Regarding catchment memory, an analysis of climate parameters and streamflow is conducted to assess the significant influence of past hydrometeorological conditions on the streamflow in current year. The results indicate that the elasticity coefficient is effective in analyzing the sensitivity of catchments in Taiwan, revealing that certain catchments have memory capabilities. These findings contribute to the hydrological prediction and offer valuable insights for hydrological modeling, as well as water resource development and management.
IG06-A017
Optimization Strategies in Unstructured Mesh Generation for Micro-scale Porous Media
I-Hsien LEE#+, Chuen-Fa NI, Thai Vinh Truong NGUYEN
National Central University, Taiwan
To evaluate the chemical reactions, physical processes, and transport mechanisms of carbon dioxide in reservoir formations are essential. Particular emphasis should be placed on the changes in pore structure and rock properties. This study is to simulate the perfusion of carbon dioxide within the cores and explore fluid transport at the micro-scale. The main tasks of this project involve analyzing X-ray CT images, constructing pore networks based on image brightness, and comparing simulation results with image analysis using mesh generation module. The three-dimensional tetrahedral mesh of the pore network is generated using the unstructured mesh generation module, and a computable mesh file compatible with fluid mechanics software is created. Ultimately, the image analysis process will be integrated to provide the Institute of Nuclear Energy Research with a basis for subsequent refinement.
IG07-A002
Geotechnical Characterization and Shear Strength Analysis of Rain-induced Shallow Landslides in Mt. Diwata, Monkayo, Davao De Oro, Philippines
Kristine Mae CARNICER1,2#+, Joel MAQUILING1
1Ateneo de Manila University, Philippines, 2Ateneo de Davao University, Philippines
This study aims to identify the potential mechanism of the frequent rain-induced shallow landslides in Mount Diwata, Monkayo, Davao de Oro, Philippines. It will investigate how the shear strength parameters of the soil from various landslips relate to the moisture content and their physical and index properties. Analysis of the direct shear test results showed that the increase in the water content decreased the shear strength of the slopes in the landslide sites. This suggests that water saturation is the primary trigger of these slope failures since they occur after days of heavy rains. In addition, the study's findings also point to the significant influence of soil composition and grain size on the soil's shear strength behavior. While the friction angle generally decreases with increasing water content for both fine-grained and coarse-grained soils, a contrasting behavior for cohesion was observed for these two. The apparent cohesion rises with increasing water content for coarse-grained soils. However, for fine-grained soils, cohesion initially increases when the water content increases but decreases when a specific moisture content is attained. These findings suggest that the shear strength is primarily attributed to cohesion for fines and friction angle for the coarser particles. The data gathered in the study can provide valuable insights to identify the most appropriate engineering interventions and mitigating solutions for this recurring problem in Mount Diwata.
IG07-A004
Fundamental Study on Estimating the Location of Buried Objects Due to Landslides Using Particle Methods
Kodai MORI1#+, Satoru OISHI1,2
1Kobe University, Japan, 2RIKEN, Japan
Consider estimating the location of buried victims after the onset of a landslide involving people; it is necessary to detect the victims. This study uses the Smoothed Particle Hydrodynamics (SPH) method, a type of particle method for numerical analysis to estimate the burial position of an object assuming a victim. The Drucker-Prager model, a type of soil constitutive law, is used to analyze mudslides based on [Bui et al., 2008]. The Passively Moving Solid (PMS) model presented in [Koshizuka et al. 1998] has simplicity and high computational efficiency for the solid-liquid multiphase flow analysis method. In this method, a rigid body is modeled as a collection of particles. Like solid particles, object particles are calculated as particles with different densities. After that, the object particles are fixed as a rigid body. To ensure the validity of the analytical method, we refer to [Bui et al. 2020] and compare it with the soil self-weight collapse experiment in [Lube et al. 2014]. The PMS model was also validated by buoyant oscillations due to the density difference between the object and the fluid, as shown by [Wang et al., 2019]. Subsequently, as a fundamental stage of the research, a sensitivity analysis was conducted on the factors that could influence the final burial position under ideal conditions of only object particles, soil particles, and wall particles. Perturbations were applied to the slope angle, viscosity, and initial object particle distance as influencing factors, respectively, and the factors affecting the object burial position were considered.
IG07-A006
Development of Multi-spatial-temporal Fusion Technologies for Rock Slope Monitoring
Chih-Chung CHUNG#+, Bo-Chi CHEN, Chun-Cheng LIN, Yu-Zhi QIU, Te-Wei TSENG
National Central University, Taiwan
The rock slopes in Taiwan are hazardous due to the influence of climate and earthquakes. The rock slopes were analyzed for different trigger factors, simulating the impact range of rock slope sliding. However, in the early stage, the project is mainly based on laboratory experiments and numerical analysis. It still needs the application and verification of substantive monitoring technology. The main topic of this sub-project is the research and development of multi-spatio-temporal scale monitoring technologies integrated into the early warning of rock slopes. This task will be based on the survey results for different rock slope failure mechanisms and approach monitoring technologies at different time and space scales. First, the space scale to be monitored will be from long-range satellite or ground-based InSAR monitoring long-term LOS (Line of Sight) displacement, combined with developing a dual-band sacrificial GNSS displacement monitoring device in the local area for comparison and verification. Furthermore, the simultaneous development of long-term real-time photogrammetry of simple optical and thermal image modules will be applied to monitor a rock slope surface. In order to introduce and test distributed strain sensing by optical fiber, in conjunction with laboratory tests and evaluations to monitor slope deformation will be proceeded; extended to distributed vibration sensing (Distributed Acoustic Sensing, DAS) for vibration (earthquake or falling rocks) or even long-term seismic analysis, can further observe the change of the shear wave velocity in the rock slope with time; in addition, through the time domain reflection method (Time Domain Reflectometry, TDR ), which can be used to monitor rock nails or anchors. This Sub-project is the integrated base of the above-mentioned novel monitoring methods to feedback to calibrate and validate physical and numerical models.
IG07-A012
Rainfall Landslide Threshold Values for Shallow Landslides in Selected Areas of the Bicol Region, Philippines
Beth Zaida UGAT1#+, Decibel FAUSTINO-ESLAVA1, Jenielyn PADRONES1, Juan Miguel GUOTANA1, Yusuf SUCOL1, Jefferson RAPISURA1, Bianca Maria Laureanna PEDREZUELA1, Maria Regina REGALADO1, Rosemarie Laila AREGLADO1, Gabriel Angelo MAMARIL1, Loucel CUI1, Wei-Yu CHANG2
1University of the Philippines Los Baños, Philippines, 2National Central University, Taiwan
The Philippines is a gateway for extreme weather events in Asia. Every year, the country experiences several tropical cyclones and heavy rains that often result in disasters. One of the triggers for these disasters are shallow landslides that damage properties and cause great loss of lives. An emerging method to mitigate the effects of landslide disasters is the generation of rainfall landslide threshold values for susceptible areas. This study presents the results of an initiative to generate such thresholds for a volcanic region in the country. The work involved crafting a simplified data collection platform to build up a landslide database from which rainfall-landslide thresholds were generated using the empirical information. Correlation work to find associations between the thresholds and land cover types allows for exploring the use of threshold values to fortify ecosystems-based disaster risk reduction and management measures. For the study areas, Albay and Sorsogon, a total of 279 events from 1985 to 2022 were collected from various sources such as reports of mandated agencies, news articles, online reporting platforms, and even social media. Out of these, only 82 events qualified to be used for generating the threshold values. On average, 33mm/hr rainfall intensities trigger landslides within the study area. Depending on the land cover type, rainfall events can trigger landslides between 28 to 40mm/hr in agricultural areas, shrublands, and built-up areas. The threshold values per land cover type will be discussed, including implications on how the management of particular ecosystems can reduce risks associated with rainfall-induced landslides.
IG07-A013
Analyzing the Impact of Rainfall-induced Pore Pressure on Slope Stability: Perspectives from a Tank Model Using a 3-D Geological Model of a Dip Slope
Alvian R. YANUARDIAN#, Jia-Jyun DONG+, Chih-Hsiang YEH, Chia- Huei TU, Jia-Yi WU
National Central University, Taiwan
For any engineering project, a 3-D geological model is crucial for estimating geological conditions, especially those related to the hydrogeological and slope stability of the dip slope. This study proposes the tank model as a fast way to know the abnormal pore pressure changes caused by rainfall and geological conditions. Moreover, FLAC3D analysis is chosen to enhance our comprehension of the diverse factors influencing slope stability. We found that the bedding plane is not planar by LiDAR-based analysis, 16 borehole observations, and site observation. A fault in the West of the research area might cause it. Thus, polynomial surface analysis was carried out to identify the 3-D geological model in our research area. By analyzing the top of L2 from 6 boreholes, 38 corrected elevation points, and 8 LiDAR points, we successfully modeled the bedding plane controlled by fault where the structure contour at the West of the research area was relatively tilted to the Northwest. On the other hand, calibrating the tank model's parameters based on a hydraulic test was used to get the best fit simulated groundwater level’s result. Therefore, we adopted and re-calibrated the tank model’s parameters of the hydraulic test to ensure a fit for the rainfall case. Afterward, we made a prediction. As a result, the tank model's simulated groundwater level and the monitoring well's observed groundwater showed a good corresponding result with the RMSE close to zero. Moreover, our prediction of rainfall-pore pressure response also showed a good corresponding result. In conclusion, concerning different joint densities parallel to the slope surface, the findings indicate varying anisotropic distributions in pore water pressure and distinct shear strain areas corresponding to changes in joint density.
IG08-A005
The Link Between Atmospheric Persistence and European Heatwaves
Emma HOLMBERG1#+, Gabriele MESSORI1, Rodrigo CABALLERO2, Steffen TIETSCHE3, Davide FARANDA4
1Uppsala University, Sweden, 2Stockholm University, Sweden, 3European Centre for Medium-Range Weather Forecasts, Germany, 4University of Paris-Saclay, France
We investigate the link between European heatwaves and the persistence of large-scale atmospheric-circulation patterns. Leveraging concepts from dynamical systems theory, we assess atmospheric persistence, and reconcile this approach with the more conventional meteorological views of persistence. We find a weak yet significant link between the occurrence of heatwaves and anomalously persistent circulation patterns in the mid-troposphere, although few significant persistence anomalies when considering the surface circulation patterns. We thus argue that persistent atmospheric configurations are not a necessary requirement for heatwaves and that the results depend to a considerable extent on region and tropospheric level. Finally, we discuss the applicability of this metric to reforecast data, and the potential value of persistence as an additional forecast verification metric.
IG08-A008
Cluster Analysis of Tropical Cyclone Tracks Affecting South Korea Using a Self-organizing Map
Han-Kyoung KIM1+, Jong-Yeon PARK1#, Jun-Hyeok SON2
1Jeonbuk National University, Korea, South, 2IBS Center for Climate Physics, Korea, South
From 1982 to 2020, 144 tropical cyclones (TCs) that affect South Korea are classified into two clusters using a self-organizing map. TCs entering the emergency zone defined by the Korean Meteorological Administration (west of 132° E, north of 28° N) are considered as those affecting South Korea. These two clusters exhibit differences in the average genesis location, track length, and maximum wind speed of TCs. The first cluster has an average genesis location in the northwest part of the western North Pacific (WNP), with a relatively short track length and weaker maximum wind speed. From an interannual variability perspective, the frequency of TCs in the first cluster shows a high negative correlation with eastern Pacific sea surface temperature (SST) (corr = -0.43, P<0.01), interpreted as a result of variations in the Walker circulation caused by changes in eastern Pacific SST. In contrast, the second cluster has an average genesis location in the southeastern WNP, with a relatively long track length and stronger maximum wind speed. This cluster displays a strong positive correlation with central Pacific SST (corr = 0.54, P<0.01), and it is analyzed that the cause of the high correlation is related to the Gill-type response associated with central Pacific SST.
IG08-A014
Enhancing Climate Resilience Through Comprehensive Assessment of Slopeland Disaster Risk Under CMIP6 Scenarios
Yun-Ju CHEN#+, Yi-Hua HSIAO, Jun-Jih LIOU
National Science and Technology Center for Disaster Reduction, Taiwan
"The impacts of climate change are constantly increasing. It is crucial to understand disaster risks and implement climate risk assessments as outlined in Taiwan's "Climate Change Adaptation Act" (2023). This study aims to comprehensively assess climate change risks, specifically focusing on slopeland disaster risk under diverse CMIP6 warming scenarios. Utilizing climate model data after statistical downscaling, this study quantifies three critical indicators—hazard, vulnerability, and exposure—using the Quantile method. Additionally, geological disaster potential maps enhance the evaluation of terrain vulnerability. This study employs the aforementioned index-based approach to analyze slope disaster risk maps at different spatial scales, including 5km grids and township areas scales. Furthermore, for the convenience of users we have developed a Climate Change Disaster Risk Adaptation Platform (http://gg.gg/drancdr) , allowing online access to slopeland disaster risk maps. This interactive platform empowers users to query trends, assess risk levels for townships and cities, and explore climate indicators. The methodology employed in this research provides scientifically robust data, offering valuable insights for policymakers and stakeholders engaged in climate change adaptation and disaster risk management. This contribution aligns with ongoing efforts to fortify climate change resilience and foster sustainable development in the face of evolving climate challenges.
IG08-A015
Could “Hurricane” Pose Major Risk to Europe?
Kelvin NG#+, Gregor C. LECKEBUSCH
University of Birmingham, United Kingdom
Tropical cyclones (TCs) are recognized for their substantial socioeconomic impact on coastal cities. While some TCs remain confined to the tropics throughout their entire lifecycle, others migrate into the midlatitudes, retaining the physical characteristics of hurricanes. Some even undergo structural evolution, transforming into post-tropical cyclones (PTCs). We collectively refer to these phenomena as Cyclones of Tropical Origin (CTOs). CTOs can exhibit exceptional intensity, with hazards that distinguish them from typical extratropical cyclones. Recent occurrences of CTOs, such as remnants of Hurricane Ophelia (2017) and Hurricane Leslie (2018), have made landfall in Europe, resulting in significant damage. This prompts the question: are these isolated events? And will these events become more common in future climates? The current quantification of CTO risk is hindered by a lack of observations as well as lack of theoretical understanding of CTOs. In this presentation, we elucidate the fundamental principles of our methodology designed to address the challenge of insufficient observations, employing the UNSEEN approach. The UNSEEN approach enables the construction of a physically consistent event set, particularly suited for the study of extreme and rare events. Recently, this methodology has proven effective in typhoon hazard assessments in the Western North Pacific. Subsequently, we present preliminary results illustrating the genuine hazard posed by North Atlantic CTOs to Europe.
IG08-A018
Localizing Precipitation Parameters for Early Warning by Applying Machine Learning to Remotely Sensed Cloud Properties
Hwayon CHOI+, Yong-Sang CHOI#
Ewha Womans University, Korea, South
Early warning of precipitation is important globally to ensure people's safety and prevent economic losses from disasters related to rain. Since precipitation and cloud properties are closely related, we would like to try precipitation early warning using the properties of clouds. Additionally, these precipitation conditions are expected to vary depending on the region, so research is needed to determine independent conditions between regions. In this study, we aim to provide early warning individually for each region whether it will rain within the next 2 hours by using cloud brightness temperature, cloud optical thickness, and cloud effective radius among the cloud properties variables provided by geostationary satellites. Among the data used in this study, the cloud property variables were from the Korean geostationary satellite GEO-KOMPSAT-2A, and precipitation information was from Integrated Multi-Satellite Retrievals for Global Precipitation Measurement precipitation data. We used machine learning techniques to find precipitation conditions appropriate for each latitude and longitude, and measured the accuracy of the precipitation probability prediction algorithm using three indicators: Probability of Detection (POD), False Alarm Ratio (FAR), and Critical Success Index (CSI). As a result of this study, our method showed higher accuracy than previously known parameters, especially CSI. This result suggests that the missed alarm rate, the event of precipitation occurring despite not being forecast, has reduced. Additionally, precipitation conditions vary depending on the properties of clouds, and new regionalization is proposed by grouping areas with similar precipitation conditions. These results are expected to enable higher accuracy early warnings and improve understanding of precipitation conditions through localization using cloud properties information.
IG08-A019
Comparison of Different Downscaling Models for High-resolution Precipitation in Korea
Minjeong KONG+, Yong-Sang CHOI#
Ewha Womans University, Korea, South
Recently, extreme rainfall frequently occurring in South Korea has caused numerous casualties and property damage. To predict and manage such damage, high-resolution precipitation data is needed. However, there are limits to the data that can be obtained due to the lack of ground observatories. To overcome these limitations, we aim to produce and compare daily high-resolution (1 km) precipitation data using three models: linear regression, MK-PRISM (Modified Korea-Parameter elevation Relationships on Independent Slopes Model), and random forest model. In particular, the random forest model is a distance-based deep learning model newly proposed for producing high-resolution precipitation data. The daily precipitation data was based on ground observation station data from the Korea Meteorological Administration. The difference between estimated and observed values was calculated to measure the accuracy of the calculated high-resolution daily precipitation data. As a result, linear regression and random forest models tended to underestimate, while MK-PRISM tended to overestimate compared to observed values. Among the three models, the random forest showed the best performance. These findings could provide a universal daily precipitation estimation method based on deep learning that covers the entire South Korea. In addition, it will be helpful for estimating climate risk as basic data.
IG08-A020
Enhancing Quantitative Precipitation Estimation in the NWP Using Deep Learning Model
Haolin LIU#+, Zhenning LI, Jimmy Chi Hung FUNG, Alexis LAU
The Hong Kong University of Science and Technology, Hong Kong SAR
Precise precipitation quantification is vital for minimizing damage and safeguarding lives during extreme weather events, especially in a rapidly changing climate. Current quantitative precipitation forecasting (QPF) in numerical weather prediction (NWP) models heavily relies on parameterization schemes for microphysics, cumulus clouds, etc, introducing significant uncertainties due to our limited understanding of precipitation processes. To address this challenge, we propose a deep learning model based on the Vision-Transformer architecture. Our model directly utilizes fundamental meteorological variables computed by NWP models as inputs and quantitatively maps them to precipitation maps derived from satellite-merged data. We conducted Weather Research and Forecasting (WRF) model simulations with a 27km grid resolution over China and Southeast Asia from 2017 to 2021. We used simulation data from the wettest seasons of 2017-2019 for training and validated and tested the model on 2020 and 2021 data. The deep learning model aims to bypass uncertainties in physical parameterization schemes, driven by our incomplete understanding of physical processes, and reproduce high-resolution satellite rainfall observations (CMORPH data). Evaluation results on the test dataset demonstrate that our deep learning model effectively extracts meteorological features, leading to significant improvements in precipitation forecasting skill scores. Specifically, we achieved improvements of 21.7%, 60.5%, and 45.5% for light rain, moderate rain, and heavy rain, respectively, on an hourly basis. Two case studies under different synoptic conditions also exhibit promising results in estimating heavy precipitation during strong convective events. Overall, our proposed deep learning model offers vital insights into capturing precipitation-triggering mechanisms and enhancing precipitation forecasting accuracy. Additionally, we discuss the sensitivities of the fundamental meteorological variables used in our study, training strategies, and potential limitations. In conclusion, our approach addresses the limitations of current QPF methods, showing promise for more accurate precipitation forecasting, crucial for mitigating the impact of extreme weather events in a changing climate.
IG09-A003
Modern to Relict Sediment Continuum in the Taiwan Strait
Dominique VALDIVIA1#+, James LIU1, Rick YANG1, Aijun WANG2, Yonghang XU2, Yu-Min CHOU3, Chih-Chieh SU4, Yu-Shih LIN1
1National Sun Yat-sen University, Taiwan, 2Ministry of Natural Resources, China, 3Southern University of Science and Technology, China, 4National Taiwan University, Taiwan
The Taiwan Strait (TS) facilitates sediment exchange from Mainland China, encompassing large dispersal systems like the Zhujiang and Changjiang rivers, smaller rivers along the Zhejiang-Fujian coast, and those originating from Taiwan's mountainous rivers. A comprehensive multi-proxy investigation spanning the entire strait is essential to grasp the interplay between relict and modern sediments, enhancing comprehension of the modern-to-relict sedimentary continuum within the TS. Based on clay minerals, the primary fluvial source for sediments in the TS is sediments exported by the Changjiang River, which is enriched in Chlorite-Kaolinite. Another fluvial source, the Minjiang River, exported sediment enriched in Kaolinite. Sediments from Taiwanese rivers are enriched in Chlorite, which lacks Smectite. We identified two types of modern sediments: a) polymodal silt on the western side of the TS in the Zhe Min-Taiwan Strait mud belt, which is enriched with Kaolinite, b) polymodal silt on the eastern side of the TS is enriched with Chlorite-Smectite and exhibits a very low magnetic susceptibility. In the northeastern TS, in the Guanyin Depression, sediments exhibit a different type of bimodal silty sand enriched with Illite with very low magnetic susceptibility, which might correspond to a palimpsest type. In southern TS, TB is covered by well-sorted bioclastic coarse sand with high quartz content. Our findings suggest their sediments correspond to a transition from a relict type to a palimpsest type. At this preliminary stage, the information we collected shows the distribution and characteristics of relict and modern sediments in different parts of the TS. Our future work will be focused on understanding the interplay between modern and relict sediments and the hydrodynamic regime that might be responsible for the reworking, winnowing, and resuspension of particles.
IG09-A004
Mass-transport Complexes (MTCs) and Bathymetrical Control on Submarine Channels: A Case Study from the Taranaki Basin, NW New Zealand
Wenjing LI1#+, Nan WU2
1State Key Laboratory of Marine Geology, Tongji University, China, 2Tongji University, China
Mass-transport complexes (MTCs) are ubiquitous on continental margins worldwide. During emplacement, MTCs can evacuate tremendous volumes of sediment and dramatically modified seabed geomorphology. However, the role of MTCs, more specifically their top surfaces, in influencing subsequent sedimentation processes has been largely overlooked. Therefore, we adopt multibeam bathymetry and 3D seismic reflection data from Taranaki Basin, NW New Zealand, to investigate the interaction between buried MTCs and subsequent sedimentary processes. We interpret a shallow buried MTC that is c. 19km long and 16km wide, covering an area of c. 250km2, deposited near the upper slope of the study area. The existence of the headwall scarp and lateral margins has created a large ponding space on the top surface of the MTC. When subsequent channels debouch into the headwall scarp, their flow direction diverts from NNW to NNE, and the width and depth of the channel have increased from 640-1600m to 54-70m separately. The giant blocks (c.54-86m high, 370-600m wide) within MTC have created a set of local topography highs on the top surface, which have influenced the morphology of subsequent channels by regulating the direction of flow and defining the location of crevasses. Additionally, thrust-fault-like structures have been revealed on the western levee of the channel, indicating that lateral confining pressure generated by the flow energies are extremely high when the channel flows into these topographies. Thus, we suggest that the top surface of MTCs can cause huge accommodation spaces, forming a 'mini basin' like geometry that can divert the flow direction, enhance the erosional ability and flow energy, and control the distribution pattern of subsequent channelling processes. We indicate that this process can facilitate submarine sediment transportation and redistribution from shallow to deep marine settings, which is also essential for deciphering the pathways from sediment source to sink systems.
IG09-A013
A Hidden Submarine Giant: Observation, Interpretation and Implication
Nan WU#+
Tongji University, China
Submarine landslides (slides) are ubiquitous on continental margins worldwide, they can cause damaging tsunamis and destroy submarine infrastructures. In the Gippsland Basin and Bass Canyon regions, southeast offshore Australia, we observed a giant seabed slide (termed Bass Slide) extending more than 450 km and covering an area of c. 30,000 km2. We combine high-resolution multibeam bathymetry, 2D and 3D seismic reflection, and core data to investigate the kinematic indicators and preconditioning factors of the Bass Slide. We found at least five buried slides deposited below the headwall part of the Bass Slide. These buried slides exhibit a retrogressive pattern based on the spatial distribution of their basal shear surfaces. We demonstrate that the giant Bass Slide can be formed by repeated slope failure processes, and that landsliding is still active in shaping the system. This supports previous studies indicating that tsunamis have been a significant geomorphic process in the Gippsland Basin since the late Pleistocene. We therefore consider the study area to have a high tsunami risk, which requires modelling-based studies to determine if any coastal regions are at risk in the future.
IG09-A029
Various Amberground Marine Animals on Burmese Amber with Discussions on Its Age
Yingyan MAO+, Diying HUANG#
Chinese Academy of Sciences, China
Burmese amber represents the world’s most diverse biota in the Mesozoic. Previous studies have focused on the biodiversity of its inclusions, as well as pholadid borings. Here we report a variety of marine animals symbiotic with or adhere to Burmese amber or the amber deposits, including crinoid columns, corals and oysters. We propose that there is no distinct evidence indicating the secondary transportation of Burmese amber over long distances. The ancient sedimentary environment was likely located in the coastal area. The hardening time of the resin was not long after secretion. The resin has been mixed with fragments of marine organisms in the ancient sediments, and has been deposited for a longtime. The zircon age in the sediments surrounding amber approximately represents the age of Burmese amber, but due to limits of the method, the current zircon U-Pb SIMS age maybe younger. Therefore, as far as the situation is concerned, the age of Burmese amber may be close to the boundary between the Albian and Cenomanian, or even late Albian. We suggest that it is plausible to generally refer to the age of Burmese amber as mid-Cretaceous, and a precise age requires further biostratigraphic and chronological studies.
IG09-A030
The Spatiotemporal Evolution of the Early Cretaceous Jehol Biota in East Asia
Daran ZHENG1#+, Su-Chin CHANG2, Haichun ZHANG1
1Chinese Academy of Sciences, China, 2The University of Hong Kong, Hong Kong SAR
The modern terrestrial ecosystems greatly developed during the mid-Cretaceous, characterized by the dominance of angiosperms. The flourishment of the angiosperms drives elevated species richness by co-evolving with insects, vertebrates and fungi, strongly altering the Cretaceous climate and water cycles by increasing bedrock weathering. The Lower Cretaceous non-marine sediments in East Asia contain the famous Lagerstätte: the Jehol Biota. This Biota produced numerous exceptionally well-preserved fossils, including feathered dinosaurs, early birds, mammals, amphibians, pterosaurs, insects and early angiosperms, contributing to tracking the early evolution of modern terrestrial ecosystems. The Jehol Biota sensu stricto is a terrestrial fossil assemblage mainly distributed in northern Hebei Province, western Liaoning Province, and southeastern Inner Mongolia Region of the North China Craton, characterized by the 'Eosestheria-Ephemeropsis-Lycoptera' (EEL) assemblage. This assemblage has a very wide distribution in East Asia thus generating the view of the Jehol Biota sensu lato. The evolution of the Jehol Biota was generally divided into three evolving stages, i.e., Jehol Biota stages I to III (JBS I to III), represented by the fossil assemblages from the Dabeigou (~132–130 Ma), Yixian (~125–122 Ma) and Jiufotang (~122–115 Ma) formations, and their corresponding strata. The Jehol Biota begins with the appearance of elements like Ephemeropsis trisetalis (insect) and Peipiaosteus (fish) during the JBS I. It flourished during the JBS II, further developed during the JBS III, and finally transformed into the Fuxin Biota during the late Early Cretaceous. JBS Ⅱ was the evolutionary climax of the Jehol biota, and nearly all the typical vertebrates and angiosperms of this biota appeared. The biota’s survival of the middle Early Cretaceous breakup of Gondwana and the destruction of the North China Craton demonstrates an adaptive response to long-term environmental changes. This study widely investigated the Lower Cretaceous in East Asia, collecting abundant fossils to recover the Early Cretaceous terrestrial biodiversity and palaeogeography in East Asia. The spatiotemporal evolution of the Jehol Biota is evaluated after a comprehensive study from biostratigraphical, biogeographical, high-precision geochronological, and quantitative analyses.
IG09-A032
The Baiwan Biota from the Qinling Orogenic Belt Linked to the Early Cretaceous Biological Dispersal in Central China
Siyu SONG1#+, Daran ZHENG1, Xiao TENG1, Su-Chin CHANG2, Honghe XU1, Bo WANG1, Haichun ZHANG1
1Chinese Academy of Sciences, China, 2The University of Hong Kong, Hong Kong SAR
The Qinling Orogenic Belt (QOB) was uplifted during the Jurassic, forming the boundary zone of the climate division in China. However, it remains unclear whether this Belt has developed as the palaeogeographic isolation barrier for the biological communications between the North China Craton (NCC) and South China Craton (SCC) during the Mesozoic. We here report the Baiwan Biota (ca. 123.9 Myr ago), a new Early Cretaceous terrestrial fossil assemblage from the Baiwan Formation of the Baiwan Basin, southwestern Henan Province in the QOB. This biota contains abundant invertebrate fossils, as well as an angiosperm fossil, mostly belonging to the typical elements of the Jehol Biota, which was widely distributed in East Asia. Biostratigraphical and detrital zircon analyses reveal the absence of drainage systems between the southern NCC and QOB. Further quantitative analyses support that the QOB prohibited the biological dispersal between the southern NCC and QOB during the Early Cretaceous, and the Jehol Biota (especially its aquatic components) was probably expanded along the Shangdan Fault within the QOB. This study reveals the unique palaeogeographical position of the QOB for exploring the development of Cretaceous terrestrial ecosystems.
IG10-A004
Numerical Investigation of Run-up and Sediment Transport of Tsunami-like Solitary Waves on Fine-sand Beaches
Shuo LI1,2+, Huabin SHI1#
1University of Macau, Macau, 2Zhuhai UM Science & Technology Research Institute, China
A two-phase Smoothed Particle Hydrodynamics (SPH) model is adopted in this study to numerically investigate the hydrodynamics and sediment transport by tsunami-like solitary waves in the swash zone of fine-sand beaches. The model is carefully calibrated and validated regarding the run-up height and the sediment transport rate under solitary waves in laboratory experiments. The calibrated model is then employed to investigate the influential factors of the run-up and the sediment transport. Dimensionless relationships between the factors and the run-up height and the sediment transport pattern are proposed based on a large number of numerical experiments. The quantitative results help to assess the impact of tsunami-like solitary waves on sandy beaches.
IG10-A005
A New Tsunami Earthquake Model for the 1771 Meiwa Event and Implication for the Bathymetric Effect on the Tsunami Boulder’s Distribution
Koki NAKATA1#+, Kazuhisa GOTO1, Hideaki YANAGISAWA2
1The University of Tokyo, Japan, 2Tohoku Gakuin University, Japan
In 1771, the large tsunami (Meiwa event) struck the Sakishima islands, Japan. By this event, tsunami run up ~ 30 m in elevation and 12000 people died. Previous studies proposed several tsunami source models for this event but it is still debated. In addition, more reliable and accurate historical record data was developed in recent years and these data was not used in the previous works. These studies also used coarser topographic and bathymetric data for the tsunami calculations and they may be insufficient for the accurate tsunami modeling. In order to update the source model of the Meiwa tsunami, we adopted high-resolution topographic and bathymetric data with the latest historical record as a constraint. The results show that a very large slip (30 m) is required in the shallow and narrow area along the Ryukyu Trench to reproduce the Meiwa tsunami run-up height distribution. Indeed, the model can be classified as a tsunami earthquake. As well as large fault rupture, inelastic deformation of unconsolidated sediments and/or submarine landslides may be involved in this large slip. Our high-resolution modeling allowed us to understand the detail propagation and run-up processes. For example, it was known that tsunami boulders are deposited along the coasts but their distributions were not uniform. Our modeling results suggested that these boulders were concentrated in the area where the tsunami was higher than the surroundings. The wave ray analysis further revealed that tsunami propagation was strongly controlled by bathymetry and hence boulders’ distribution as well. Since these boulders are fragments of corals or coral reefs, our results in turn suggest that the impacts of tsunamis on corals and coral reefs are also heterogeneous: their damage could have been controlled by bathymetry.
IG10-A008
Formation Processes of Erosional Landforms by the 2011 Tohoku-oki Tsunami
Haruki IMURA#+, Kazuhisa GOTO
The University of Tokyo, Japan
The 2011 Tohoku-oki tsunami caused extensive geomorphological changes along the Pacific coast of Tohoku, Japan. Although extensive surveys of tsunami deposits were conducted using various methods, tsunami erosional features have not been studied well due to the limited locations where erosion occurred and their transient natures. Nonetheless, residential and agricultural areas have been significantly damaged by the tsunami erosion. Understanding formation processes of erosional landforms by modern tsunamis are essential to better understand paleotsunami processes. In this research, we aimed to elucidate the formation processes of erosional landforms by the 2011 Tohoku-oki tsunami in various locations along the Pacific coast of Tohoku to clarify the factors that determine where erosion occurs through the field survey as well as analyses of satellite imageries and numerical calculations. Our results revealed that tsunami erosion is often caused by prolonged return flow rather than by the strong but short-lasting impact of run-up flow. In fact, while the run-up flow can cause some minor erosion at the coast, the return flow, which concentrates in low-lying coastal areas, generates a strong current with high Froude Number. This current toward the sea persists for tens of minutes and forms localized but significant erosional landforms. Our study also found that the tsunami erosional features are largely controlled not by the wave height but rather by the topography (slope inclination) in the inundation area. While, the localized erosional landforms are strongly influenced by the meter-scale microtopography. Our discovery underscores that coastal erosion by the tsunami is closely linked to both macro- and micro-topography. This in turn indicate that there is a predictability at the sites of future significant tsunami erosion based on the geomorphological analysis and tsunami sediment transport modeling.
IG10-A009
A Proposed Approach Towards Estimating Tsunami Drift to Aquaculture Rafts Through the Development of a Drift Simulation Model
Kento TANAKA1#+, Anawat SUPPASRI1, Yoshinori SHIGIHARA2, Fumihiko IMAMURA1
1Tohoku University, Japan, 2National Defense Academy, Japan
Coastal aquaculture is widely practiced in Japan. Aquaculture rafts, which consist of culture of shellfish and other macroalgal species, are generally placed in shallow marine regions – making them extremely susceptible to wave impacts. Tsunamis frequently cause damage to aquaculture rafts in Japan, even when tsunami generation source is distant. Despite a relatively small wave height, tsunami waves could result in drift motion of a freely floating object and hence, significant damage to the object. However, our knowledge of drift motion due to tsunami and the relationship between tsunami drift motion and aquaculture raft damage is still limited. Additionally, there are only few attempts in current literature that have developed methods in predicting tsunami drift damage to aquaculture rafts. Therefore, the objective of this study is to develop a model for drift simulation of aquaculture rafts during a tsunami, to predict and reduce the drift damage Tsunami drift, quantified by the object’s (in this case, raft’s) position and orientation, can be estimated by solving the equations of motion through numerical simulation of tsunami water level and current velocity. Since the aquaculture facilities have anchors for mooring, in this study, the equations of motion are solved including the frictional resistance due to the concrete anchors. However, due to the complex structure of aquaculture rafts, it is challenging to calculate drag coefficient just from numerical simulation. We have therefore also created models of aquaculture rafts and performed laboratory experiments to measure their drag coefficient. The improved model was applied to the 2022 Tonga volcanic tsunami scenario, with Yamada Bay, Japan as our case-study area. When validated with observation data, our results show that rafts that were damaged/not damaged during the 2022 event were similar to those simulated.
IG10-A010
Suggestion on Developing Indonesia Tsunami Observation Networks and Forecasting System Based on Deep Learning
Muhammad Rizki PURNAMA#+, Anawat SUPPASRI, Kwanchai PAKOKSUNG, Constance Ting CHUA, Fumihiko IMAMURA
Tohoku University, Japan
Past events of the catastrophic tsunami in Indonesia, such as Flores tsunami 1992, Banyuwangi tsunami 1994, The Great Indian Ocean tsunami 2004, Pangandaran tsunami 2006, Palu tsunami 2018, and Krakatau tsunami 2018, have caused significant damage and losses in the coastal area. Most of the events are triggered by a submarine earthquake. One of which is from the Java subduction zone. Three major earthquakes with tsunamis have been caused by deformation along this subduction zone. Thus, this zone covers most of the populated as well as critical areas spanning from Java to Nusa Tenggara. Therefore, Indonesia needs a decent Tsunami Early Warning System. There are only six existing OBPGs (Ocean Bottom Pressure Gauges) to inspect tsunami wave propagation in Indonesia's deep-sea area to issue an early warning. In this research, we assessed the optimal deployment of tsunami buoys along The Southern part of Java Island to Nusa Tenggara and made a simple deep-learning-based model to forecast the tsunami waveform in the nearshore area. First, we create a set of synthetic tsunami multi scenarios using a stochastic-slip earthquake model for the Java Megathrust fault. Next, we apply the empirical orthogonal functions (EOF) based on the tsunami scenarios to determine the initial location for the deployment of each tsunameter. We also assessed the reliability of the spacing and bathymetry for each suggested sensors. Finally, we apply a series of deep learning models to forecast tsunami waveforms and further inundation area in the nearshore area.
IG10-A024
A Decade After the Onslaught: Tracking Down Super Typhoon Haiyan Deposits and Other Potential Strong Wave Events in Leyte Island, Philippines Using eDNA
Wenshu YAP1#, Janneli Lea SORIA2+, Ronald ILOREN3, Mischa HAAS4, Adonis GALLENTES5, Jodivine NAVAROSA6, Mark Russel GUATNO7, Clyde PELESCO8, Nathalie DUBOIS 9, Fernando SIRINGAN5, Adam SWITZER1
1Nanyang Technological University, Singapore, 2University of the Philippines Diliman, Philippines, 3ETH Zürich, Switzerland, 4Kanton Luzern, Dienststelle Umwelt und Energie, Switzerland, 5University of the Philippines, Philippines, 6Central Visayan Institute Foundation, Philippines, 7Guatno Surveying Services, Philippines, 8GeoPoint Land Surveying Services, Philippines, 9Eawag, Switzerland
Understanding the recurrence interval of extreme coastal events, such as super typhoons, is crucial for coastal zone management and resilient community development. While the sedimentary deposits of these extreme coastal events offer valuable records, their preservation potential beyond historical timescales presents a significant challenge. This study explores the viability of environmental DNA (eDNA) as a complementary tool for fingerprinting the overwash signatures of Super Typhoon Haiyan (ST Haiyan) in Leyte Island, Philippines, a decade after the event. In June 2023, we revisited and collected transect cores of the previous study conducted by Soria et al in 2017, which documented the thickness of the overwash deposits. The Nypa mangrove forest recorded a thickness of 5-7 cm, while inland, about 1.2 km from the coast, had a 1-3 cm thickness. The varying distances from the coast and deposit thicknesses allow us to capture potential spatial variations in deposit characteristics and eDNA preservation. Despite differences in deposit thickness, grain size, sorting, and composition between sites, eDNA analysis consistently differentiates ST Haiyan deposits from pre- and post-event soil samples. These findings demonstrate the remarkable persistence of eDNA within overwash deposits even under variable preservation conditions. This suggests the potential of eDNA as a valuable proxy for reconstructing past extreme events beyond traditional sedimentological approaches. This work presents a significant advancement in reconstructing historical coastal hazard events and their long-term impact on coastal communities. Leveraging eDNA’s resilience and discriminatory power opens a novel tool for enhancing the interpretability of sedimentary records of marine inundation events and extending typhoon history to centennial or millennial timescales, thereby informing coastal hazard mitigation strategies for a safer future.
IG11-A001
| Invited
Integration of Dynamical Downscaling and a Time-lagged Ensemble to Enhance Medium-range Forecasts for Agricultural Applications
Subin HA1, Eun-Soon IM1#+, Jina HUR2, Sera JO2, Kyo-Moon SHIM2
1The Hong Kong University of Science and Technology, Hong Kong SAR, 2National Institute of Agricultural Sciences, Korea, South
More accurate and reliable meteorological information in advance of up to a month is expected to have significant implications for impact sectors such as agricultural practices. With the increasing demand for subseasonal-to-seasonal weather forecasts, many organizations and institutions worldwide produce GCM-driven forecasts extending several months. However, the coarse resolution of these forecasts often hinders their applicability and usefulness for farmers and agricultural stakeholders, limiting their ability to make informed decisions based on accurate and localized meteorological information. In this regard, this study aims to produce fine-scale 1-month forecasts tailored in South Korea by dynamical downscaling of NCEP Climate Forecast System version 2 (CFSv2) operational forecasts initialized at different times. The optimal time-lagged ensemble technique, utilizing systematically selected members, is also employed to enhance predictability. By offering a range of potential meteorological conditions in advance, this approach is vital for facilitating the implementation of timely adaptation measures, thus assisting in mitigating the adverse impacts of natural hazards that are expected to worsen in a changing climate. Acknowledgments: This study was carried out with the support of “Research Program for Agricultural Science & Technology Development (Project No. PJ014882)”, National Institute of Agricultural Sciences, Rural Development Administration, Republic of Korea.
IG11-A002
Nowcasting of Landslides: A Geo-satellite Synthesis Approach for Real-time Hazard Mapping Across Asia and Oceania
Min-Jae KWON#+, Yong-Sang CHOI
Ewha Womans University, Korea, South
Landslides are one of the most widespread natural hazards on Earth, inflict huge losses of both life and property every year. In advance of landslide occurrences, people located in landslide hazard area need risk information to prepare for the disasters. According to landslide catalog, 75% of landslide cases are triggered by rainfalls which occurs in short time and needs near-real time monitoring. For continuous observation, geostationary satellite data are useful which cover global scale area and observe a short timescale. In this study, we use GEO-KOMPSAT-2A (GK2A), the latest geostationary satellite data (2 km and 10 minutes resolution). For a quick update of information on the landslide risk area, a hazard map is created based on the frequency ratio (FR). To evaluate the risk of landslides, five contributing factors for landslides are considered which are Enhanced Vegetation Index (EVI), land cover, soil moisture, elevation and slope gradient. They are respectively coupled with the NASA Global Landslide Catalog (GLC) data to calculate the FR, which includes landslide occurrence information from 1915 to 2022. The FR from each factor are combined to determine the final hazard area of landslides. Lastly, alarming area is defined as the high risky area where heavy rain is expected. In this study, the possibility of heavy rain is determined using GK2A L1B data to detect developing or mature clouds to give fast and near real-time information. In addition, it does not only analyze FR values uniformly, but furthermore, for areas where the correlation with factors are unusual, regional characteristics are identified through case analysis and reflected in the algorithm.
IG11-A004
Improvement of GK2A Forest Fire Detection Algorithm
Seoyoung CHAE#+, Yong-Sang CHOI
Ewha Womans University, Korea, South
Wildfire management is becoming more important around the world as disasters by wildfires increase due to climate change. Wildfires cause irreversible ecosystem destruction, loss of life, and property damage. It has been proven in many cases that early detection and rapid response to wildfires can reduce their impact. Detecting wildfires using satellite data has the advantage of being free from time and space constraints and can be cost-effective. Currently, GK2A is providing Level 2 wildfire detection data using its wildfire algorithm. The GK2A algorithm is based on the MODIS wildfire algorithm and contains temperature fluctuations due to the topographic lapse rate. The current algorithm uses the 3.8 µm and 10.5 µm channels for wildfire detection. The 3.8 µm channel is a combination of thermal and solar and has been known to detect high temperatures since the early days of satellite detection. However, the combination of the two sources inevitably leads to contamination from the larger energy source which is solar. If the other IR channels of GK2A can effectively observe surface temperature changes, it is reasonable to use other channels. In this study, we propose a more accurate algorithm for wildfire detection using channels other than the 3.8 µm channel. Using the Radiative Transfer Model to model brightness temperature changes caused by wildfires to found the reasons why 3.8 µm is difficult to use and the wildfire detection strengths of other channels. Based on the model results, improved the GK2A wildfire detection algorithm, and compared the GK2A algorithm and the renewal algorithm about the 2022 wildfire case.
IG11-A008
Independent Validation of Deep-learning-based Rainfall Induced Shallow Landslides Forecasting in Italy
Alessandro MONDINI1#, Massimo MELILLO1+, Fausto GUZZETTI1, Michele CALVELLO2, Gaetano PECORARO2
1National Research Council, Italy, 2University of Salerno, Italy
Geographical landslide early warning systems are difficult to evaluate. In this work, we assess the performance of a recently proposed deep-learning-based system for short-term forecasting of rainfall-induced shallow landslides in Italy. For our assessment, we use rainfall measurements from the same rain gauge network used to construct the forecasting system, and different and independent information on the timing and location of 163 rainfall-induced landslides that occurred in Italy in a period non covered by the data used to train the forecasting system. The new dataset is extracted from the FraneItalia catalogue (https://zenodo.org/records/7923683). Results confirm the good predictive performance of the forecasting system and reveal no geographical or temporal bias in the forecasts. The analysis shows that the forecasting system is effective at predicting multiple landslides in the same general area. Analysis of the false negatives shows that approximately one-third of the landslides were rockfalls, and for approximately another third there was uncertainty in the database about when or where landslides have occurred. We confirm that the deep-learning-based system analysed is well suited for short-term operational forecasting of rainfall-induced shallow landslides in Italy.
IG12-A007
Temporal and Spatial Studies Over a Scour Hole Downstream of a Grade‐control Structure
Tung Hsuan TSAI#+, Dong-Sin SHIH
National Yang Ming Chiao Tung University, Taiwan
In recent years, climate change has affected Taiwan, resulting in an increasing trend in rainfall intensity and frequency. The occurrence of intense rainfall events significantly impacts the erosion and accumulation of riverbeds. The erosion of these sediments can, in turn, cause damage to the foundations of in-river structures, thereby affecting the safety of these structures. Therefore, installing protective measures on the in-river structures is crucial to monitoring erosion. Common grade‐control structures are usually made of square or rectangular cement blocks. Hence, this study aims to utilize the different shapes of protection works to discuss their effects on downstream erosions. Laboratory flume experiments are conducted to study the impacts of scouring hold subjected to water flow. Moreover, the triaxial accelerometers are used to record vibration acceleration data of the protective structures. Vibration frequencies were obtained through the Fourier transformation of the acceleration data. Simultaneously, downstream erosion depth is recorded hourly using cameras, and Particle Image Velocimetry (PIV) was employed to calculate flow velocity and variations in the flow field within the scour hole. Theoretically, downstream erosion can decrease the foundation's bearing capacity, reducing structural stability and alterations in the overall structural vibration frequencies. This study explores the relationship between erosion depth and frequency under different flow rates and velocities as water flows over grade-control structures with different shapes.
IG12-A012
A Preliminary Study to Investigate the Relationship Between Urban Green Space and Mental Health
Jiyoon MOON#+, Kwangjae LEE
Korea Aerospace Research Institute, Korea, South
As the problem of urbanization intensifies around the world, the issue of human quality of life is also increasing. In particular, as part of improving the quality of life of urban residents, interest in urban green areas has steadily increased in the past, and there are studies showing that urban green areas provide various utility and functions to humans, thereby helping to improve physical and mental health. In this study, we tried to qualitatively investigate the relationship between urban green areas and human mental health, and the study area was targeted at Seoul, the capital of South Korea. NDVI acquired with Landsat 8 images was used to determine the distribution of urban green areas, and the number of depressive patients among mental diseases were used together to analyze the relationship between urban green space and mental health. As a result of the study, NDVI and depression were derived as Pearson correlation -0.084, and the probability of significance was 0.689. Kendall and Spearman's correlation coefficients were -0.040 and -0.064, and the probabilities of significance were 0.779 and 0.762, respectively. The results of this study did not show a high probability of significance of more than 95%, but recorded a correlation of about 70% or more, and it was confirmed that there was a negative correlation. Through this, it was confirmed that the higher the NDVI index, which means the more green spaces and the higher the vitality, the less patients with depression. In the future, it is necessary to continue to conduct related research to examine the relationship with depression as well as other mental diseases, and these findings are expected to be used as basic data for planning or expanding urban green areas in consideration of the number of patients with various diseases.
IG12-A015
Analysis of Urban Heat Island Through MODIS Land Surface Temperature Downscaling in Daegu Metropolitan City, Republic of Korea
Youngseok KIM#+, Siwoo LEE, Dongjin CHO, Jungho IM
Ulsan National Institute of Science and Technology, Korea, South
Investigating the spatiotemporal patterns of urban heat island (UHI) plays a crucial role in understanding urban thermal environments and in developing sustainable cities. Moderate Resolution Imaging Spectroradiometer (MODIS), offering land surface temperature (LST) with frequent revisits (i.e., four times a day) over global, has been widely used in urban climate studies. However, the coarse spatial resolution of MODIS LST (1 km) presents challenges in capturing detailed thermal distributions across heterogeneous urban areas. So, this study conducted the spatial downscaling of MODIS 1 km LST to 250 m and explored the spatiotemporal patterns of surface UHI (SUHI) by administrative district-level (dong) in Daegu, South Korea for the three years 2018, 2021, and 2023. For spatial downscaling, a kernel-driven method was used to combine the MODIS LST with several auxiliary variables related to the LST. Random forest (RF) and convolutional neural network (CNN) were utilized to build spatial downscaling models. To validate the downscaling performance, we aggregated the downscaled LSTs to 1 km and evaluated the aggregated 1 km LSTs with original 1 km MODIS LSTs. RF showed the superior performance than CNN showing a coefficient of determination of 0.86 with a root mean square error of 0.75 °C. As results of spatiotemporal patterns using SUHI intensity (SUHII) extracted from RF, Goseong-dong, Yucheon-dong, and Wondae-dong showed significant variations in SUHII over the study periods with -1.44 °C, -1.41 °C, and -1.38 °C, respectively. The SUHII was high in the order of summer, spring, autumn and winter all years. During spring, summer, and autumn, higher SUHII was observed during the daytime compared to nighttime, while the opposite pattern was evident during winter. These findings on spatiotemporal variation of SUHII can help understand patterns of UHI and formulate policies for UHI mitigation by administrative district-level in Daegu city.
IG12-A016
Hydrogeological Structure Identification Using the Transient Electromagnetic Method in the Northern Area of Choushui River Alluvial Fan, Central Taiwan
Agung Nugroho RAMADHAN+, Ping-Yu CHANG#, Jordi Mahardika PUNTU, Jun-Ru ZENG, Lingerew Nebere KASSIE
National Central University, Taiwan
This study aims to identify the hydrogeological structure in the northern area of the Choshui River Alluvial Fan, Changhua County, Central Taiwan, by using the Transient Electromagnetic (TEM) method. It is necessary to conduct this study since hydrogeological structure information is crucial for several reasons, such as groundwater exploration, groundwater management, land subsidence prevention, water supply planning, and environmental impact assessment. We deployed 54 TEM measurement sites to cover the Changhua area, using the FASTSNAP system with a 50 × 50 m transmitter loop configuration. Subsequently, the TEM data were analyzed based on available geological condition information and other geophysical measurements, such as existing borehole data and Vertical Electrical Sounding (VES) conducted previously. The results reveal that the eastern area near the Bagua Tableland mostly consists of a gravel layer, while the western area near the coast mostly consists of a clay layer. Overall, this study provides new insights into how to identify hydrogeological structure using the Transient Electromagnetic (TEM) approach.
IG12-A018
Investigation of Permafrost by Using Electrical Resistivity Imaging and Ground Penetrating Radar: A Case Study in Svalbard, Norway
Yin-Long CHEN+, Ping-Yu CHANG#, Jordi Mahardika PUNTU, Ding-Jiun LIN
National Central University, Taiwan
Over the years, climate change has significantly impacted the global ecological environment, with the polar regions being among the first to experience changes. Not only the environment above the surface has been affected, but also the subsurface. This study aims to characterize and find the interaction between the permafrost and active layer, by implementing Electrical Resistivity Imaging (ERI), and Ground Penetrating Radar (GPR). Along with the Nicolaus Copernicus University team, this study is eager to learn about this phenomenon in the Svalbard Archipelago, Norway. We collected the data by employing a Wenner array configuration using 40 electrodes with 1m spacing for the ERI, and a 100 MHz antenna frequency for the GPR. The profiling line is about 140 meters from the station and extends forward to the coast. The data were collected in 2022 and 2023, in order to observe the variation and behavior of the subsurface. Furthermore, the ERI and GPR results were integrated with groundwater level and temperature distribution data to enhance the result. Thus, a comprehensive understanding of the subsurface changes due to climate changes can be obtained. In summary, our geophysical approaches have proven to be a powerful tool for this purpose.
IG12-A022
Seismic Anisotropy of the Crust and Upper Mantle Beneath Rifts in Southern Tibet
Changhui JU#+
China Earthquake Administration, China
There are several large-scale N-S trending rift systems in southern Tibet, nearly perpendicular to the Indus-Yarlung Tsangpo Suture (IYS). Served as a focal point for Late Cenozoic magmatic activity, large to super-large metallic ore deposits have developed in southern Tibet and the formation of these deposits is closely related to the evolution of rift zones. We calculate the shear wave splitting parameters (Fast Polarization Direction and Delay Time, abbreviated as FPD and DT) by minimum energy method. North of the IYS, the predominant anisotropy is oriented from NNE to NE directions. South of the IYS, the distribution of anisotropy parameters is more complex, featuring various directions and intensity of anisotropy, and can be bounded in 87°E. In the west of NTR, subparallel to the seismic anisotropy to the north of IYS, NE oriented FPDs were observed, and can be interpreted as a coherent lithospheric deformation. Between NTR and XDR, subparallel to the strike of the orogen, E-W directed FPDs were calculated, with relatively small DTs. One possible explanation is the mantle convection caused by the sub-vertical subducting Indian lithosphere. In the east of NTR, orthogonal to the APM direction, NNW trending FPDs were obtained and can be attributed to the lithospheric deformation soured from the compressional deformation in the collisional front of the Himalayas.
IG12-A024
Monitoring Horizontal Mean Ocean Currents Under Typhoon Conditions Using Optical-fiber Distributed Acoustic Sensing
Sunke FANG+, Jianmin LIN#
Zhejiang University, China
The emerging distributed acoustic sensing (DAS) technology provides a potential way to monitor the sea state with pre-existing submarine optical-fiber (OF) cables, even in severe weather situations. Here, we present in situ observations of the microseismic noise during the passage of severe typhoon Muifa (2022), using a ~17 km DAS-instrumented OF cable located offshore from Daishan Island in the Zhoushan Archipelago, eastern China. The microseismic noise induced by ocean surface gravity waves (OSGWs) is mainly in the ~0.08 to 0.38 Hz frequency band. And the high-frequency (>0.3 Hz) component is tide-modulated and only observed during low-tide periods in areas of shallow water (<10 m). The noise component around 0.20 Hz at relatively deeper cable channels is observed strongly related to the sea-surface winds at the nearby ocean buoy (with a correlation coefficient of ~0.89), as well as the typhoon track. Furthermore, the OSGW propagation along the entire cable is estimated via frequency-wavenumber analysis and found to be significantly reversed during the typhoon passage, consistent with the observed sea-surface winds. Additionally, we propose a simple and effective method to estimate both the speeds and directions of horizontal mean ocean currents. It is based on the current-induced Doppler shifts of DAS-recorded OSGW dispersions. The measured currents are consistent with the tide-induced sea-level fluctuations and sea-surface winds. These observations demonstrate the potential feasibility of monitoring sea states under typhoon conditions using DAS-instrumented cables.
IG12-A025
Hydrous Mantle Transition Zone in the Western Himalayan Syntaxis
Guohui LI1#+, Yuanze ZHOU2, Yuan GAO1
1China Earthquake Administration, China, 2University of Chinese Academy of Sciences, China
Water in the Earth’s interior can significantly reduce mantle viscosity and melting temperature of rocks, which plays a very important role in the Earth's dynamic and chemical evolution, and habitability. Here, we image the P- and SH-wave velocities of the mantle transition zone by matching synthetic and observed triplicated seismic waveforms in the western Himalayan syntaxis. With respect to the IASP91 model, the depth of the 660 km discontinuity descends by 19~29 km. The P-wave velocity of mantle transition zone is normal, but the SH-wave velocity reduces a lot. These observations indicate that the mantle transition zone is rich in water. In addition, the existence of a partially melted low-velocity layer atop the mantle transition zone also suggests a water-rich mantle transition zone.
IG16-A003
Assessing Snowmelt and Forest Expansion in the Himalayan Region: Climate and Vegetation Dynamics
Jyoti SHARMA#+, Disha SACHAN, Aaquib JAVED, Pankaj KUMAR
Indian Institute of Science Education and Research Bhopal, India
Within the world's mountainous regions, Asia boasts the largest and most populous mountain system - the Himalayas. This distinctive terrain showcases a blend of glaciers and diverse vegetation, with the former crucially feeding perennial rivers in neighboring nations and the latter offering a spectrum of ecosystem services pivotal for supporting the livelihoods of local communities. It is endowed with an overwhelming richness of biodiversity, recognized among 36 global biodiversity hotspots. Since climate change is expected to bring alterations to all ecosystems, including those in the mountains, past studies have suggested higher vulnerability of the Himalayas to climate change, as warming in this region is greater than the global average rate. Consequently, most glaciers in the Himalayas are experiencing mass loss through melting and calving. This alteration in the availability of land surface water and the exposure of bare soil due to warming scenarios could potentially drive forest expansion. Such transformation from snow-covered areas to forests, including shifts in snowline altitude and treeline, might influence changes in surface energy fluxes by altering snow-albedo and triggering various vegetation feedback. However, the key factors steering vegetation change and the resulting impact on cooling or warming trends remain inadequately understood in the Himalayan region. The present study focuses on comprehending the patterns of snowmelt and forest expansion in the Himalayan region over the past two decades using satellite (MODIS) and re-analysis (ERA-5) datasets. Furthermore, the research aims to delve into the contribution of climatic and soil-related factors to the expansion of forests in the Himalayas. Preliminary results witnessed the greening trend in Sub-Tropical, Montane Temperate and Sub-Alpine Forests except Montane Wet Temperate Forest in the Himalayan region. Detailed findings will be discussed during the presentation.
IG16-A004
Characteristics of Western Disturbances Impacting Karakoram-Himalayan Region
Aaquib JAVED#, Pankaj KUMAR+
Indian Institute of Science Education and Research Bhopal, India
Extreme precipitation over the Himalayas is largely influenced by Western disturbances (WDs). In recent decades, there has been a notable strengthening of WD impact over the Karakoram (KR) region, playing a pivotal role in sustaining the “Karakoram Anomaly.” Utilizing the WD catalog derived from ERA5 and MERRA2 reanalysis datasets, our study reveals a statistically significant shift of approximately 9.7 degrees eastward in the core genesis zone for KR WDs. This shift indicates a migration towards more favorable conditions for cyclogenesis. We propose a novel parameter for identifying regions conducive to extratropical cyclogenesis. The observed shift in the core genesis is linked to increased genesis potential, convergence, and higher moisture availability along the WD path. Composite analysis further demonstrates a significant increase in moisture availability over the shifted zone. These findings underscore the crucial synoptic influence on the anomalous regional mass-balance phenomenon in the KR region.
IG16-A007
New Permafrost Maps in the Northern Hemisphere
Youhua RAN#+, Xin LI, Guodong CHENG
Chinese Academy of Sciences, China
Monitoring the thermal state of permafrost is important in many environmental science and engineering applications. However, such data are generally unavailable, mainly due to the lack of ground observations and the uncertainty of traditional physical models. This study produces novel permafrost datasets for the Northern Hemisphere (NH), including predictions of the mean annual ground temperature (MAGT) at the depth of zero annual amplitude (approximately 3 m to 25 m) and active layer thickness (ALT) with 1-km resolution for the period of 2000–2016, as well as estimates of the probability of permafrost occurrence and permafrost zonation based on hydrothermal conditions. These datasets integrate unprecedentedly large amounts of field data (1,002 boreholes for MAGT and 452 sites for ALT) and multisource geospatial data, especially remote sensing data, using statistical learning modelling with an ensemble strategy. Thus, the resulting data are more accurate than those of previous circumpolar maps. The datasets suggest that the areal extent of permafrost (MAGT≤0 °C) in the NH, excluding glaciers and lakes, is approximately 14.77 (13.60–18.97) millions km2 and that the areal extent of permafrost regions (permafrost probability>0) is approximately 19.82 millions km2. A biophysical permafrost zonation and hydrothermal condition-based permafrost zonation were also proposed. These new datasets based on the most comprehensive field data to date contribute to an updated understanding of the thermal state and zonation of permafrost in the NH. The datasets are potentially useful for various fields, such as climatology, hydrology, ecology, agriculture, public health, and engineering planning. All of the datasets are published through the National Tibetan Plateau Data Center, and the link is https://doi.org/10.11888/Geocry.tpdc.271190.
IG16-A014
Mapping the Temporal Evolution of Glaciers in Uttarakashi, Central Himalayas Using High Resolution Satellite Imagery
Iti SHRIVAS+, Supratim GUHA, Reet Kamal TIWARI#, Ashutosh Laxman TARAL
Indian Institute of Technology Ropar, India
Glaciers are crucial indicators of climate change and hold immense importance in sustaining freshwater resources essential for various ecosystems and human societies. The present study focuses on identifying temporal changes in glaciers with respect to area from 2000 to 2023 for a sample of 19 glaciers in the Uttarkashi region of Uttarakhand in the Central Himalayas, India. The study utilizes data from Landsat 5 (TM), Landsat 7 (ETM+), and Landsat 8 (OLI) with 30 meters of spatial resolution and the PlanetScope satellite constellation with a high resolution of 3 meters to digitize glacier boundaries. The results underscore a reduction in glacier area by 2.89% throughout the study period. Significant variations in this trend are determined, with Glacier G1 exhibiting the highest reduction rate at 0.8831% y-1, while Glaciers G6 and G8 display negligible changes. Friedman test implies an identical rate of area change in different time intervals. The 95% confidence intervals, ranging from 0.0957% to 0.3233% % y-1 (2000-2011), 0.1757% to 0.5063 % y-1 (2011-2017), and 0.2060% to 0.6857% y-1 (2017-2023), further contribute to the understanding of the evolving dynamics of glacier response to changing climate conditions. Emphasizing the pivotal role of high-resolution satellite imagery in precise mapping and monitoring, this study provides valuable insights into the hydrological implications associated with the noticeable recession of glaciers in the Central Himalayas.
IG21-A004
Monitoring the Ionospheric Signatures of Very Low-frequency Transmitters with Micro-/nano-satellites
Jaeheung PARK1#+, Hosung CHOI2, Magnus IVARSEN3
1Korea Astronomy and Space Science Institute, Korea, South, 2Republic of Korea Army, Korea, South, 3University of Oslo, Norway
As the frequency of electromagnetic waves becomes lower, the signal can penetrate deeper into the sea, while the installation and operation costs get higher. That is why only some countries operate Very-Low-Frequency (VLF) transmitters that can be used to communicate with submersible vehicles. At the same time, during their normal operation, VLF transmitters produce a unintended signature in the ionosphere above. This presentation will give a brief introduction to the effects of VLF transmitter signals on ionospheric cold plasma and energetic particles originating in the magnetosphere. We also show examples of the signatures captured by Low-Earth Orbit (LEO) spacecraft, such as the Swarm constellation of the European Space Agency (ESA) and the Korean NextSat-1. Finally, we discuss the possible contribution to the topic of the recently launched Korean CubeSat fleet, SNIPE.
IG21-A007
Total Ionizing Dose and Displacement Damage Dose Effects on On-board Computer Candidate for a Super Low Earth Orbit Optical Satellite
Jongdae SOHN1#+, Junga HWANG1, Hojin LEE1, Jaeyoung KWAK1,2, Hyosang YOON3
1Korea Astronomy and Space Science Institute, Korea, South, 2Korea National University of Science and Technology, Korea, South, 3Korea Advanced Institute of Science and Technology, Korea, South
At this time, we present the effects of total ionizing dose and displacement damage on On-Board Computer (OBC) candidates for Super Low Earth Orbit (SLEO) optical satellite. The Optical Satellite aims to operate for more than two years with the mission of taking high-resolution images with a resolution of 50 cm at an altitude of 300 km or less and. This satellite is a small Earth observation satellite weighing about 100 kg, suitable for super low earth orbit. Modules tested include the Arduino modules such as MKR1010, YUN, UNO, PICO, DUE, and NANO, and the Raspberry Pi modules such as Zero W, 4B, 3+ compute Lite, 3+ compute, 3 compute, and 4 compute. The modules' displacement damage dose and total ionizing dose performance were studied using 100 MeV protons, 10 MeV electrons, and 60 Co irradiation. We performed Total Ionizing Dose (TID) and Displacement Damage (DD) experiments on Arduino modules and Raspberry Pi modules to select OBC candidates for ultra-low altitude optical satellites.
IG21-A015
Lessons Learned from PEARL-1H CubeSat Operation for Power Consumption and Management
Wei-Rong HUANG#+, Mei-Hua HSU, Chi-Kuang CHAO, Yun-Ru CHEN
National Central University, Taiwan
PEARL-1H (Propagation Experiment using kurz-Above-band radio in Low earth orbit) CubeSats integrated by National Central University (NCU) and Hon Hai Precision Industry Co., Ltd. (Foxconn) for educational training/scientific research was launched into a sun-synchronous orbit at 520 km altitude around 1030 local time sector by SpaceX Transporter-9 rideshare mission from Vandenberg Space Force Base on 11 November 2023. A Communication PayLoad (CPL) developed by Tron Future was installed on PEARL-1H for broadband communication experiment with beam-steering phase array antenna. PEARL-1H is a dual-EPS CubeSat designed to support CPL, requiring high-voltage power for its operation. There are two power distribution modules and two battery packs in PEARL-1H. Except for the battery raw power, all power sources providing the same voltage are shared and each BP is managed individually through each PDM. In this study, expected system operation under such a system is discussed along with a discussion of on-orbit data.
IG21-A016
Lessons Learned from PEARL-1C CubeSat Operation During ADCS Commissioning Phase
Mei-Hua HSU#+, Chi-Kuang CHAO, Wei-Rong HUANG, Yun-Ru CHEN
National Central University, Taiwan
PEARL-1C, a 6U XL CubeSat developed by National Central University (NCU) for educational training and scientific research, was successfully launched into a sun-synchronous orbit at an altitude of 520 km around 1030 local time sector by SpaceX Transporter-9 rideshare mission from Vandenberg Space Force Base on 11 November 2023. Onboard PEARL-1C, two payloads were included: a Ka-band communication payload (KCP) developed by Rapidtek Technologies and NCU for a broadband communication experiment, and a Compact Ionospheric Probe (CIP) developed by NCU for ionospheric plasma measurement. Prior to entering the normal operation phase, commissioning phases were conducted to assess the condition of each system. The Attitude Determination and Control System (ADCS) played a crucial role in controlling the satellite's attitude to achieve the required pointing for the payloads. During the ADCS commissioning phase, a sequential activation and checkout of sensors and actuators took place to ensure that the control operations would perform as expected. This poster aims to present the lessons learned from the PEARL-1C CubeSat operation during the ADCS commissioning phase.
IG21-A019
The 2nd-generation Compact Ionospheric Probe on CubeSat for Space Weather Mission
Yun-Ru CHEN#+, Chi-Kuang CHAO, Wei-Rong HUANG
National Central University, Taiwan
Compact ionospheric Probe (CIP) is a miniature version of Advanced Ionospheric Probe (AIP), a successful science payload on FORMOSAT-5 satellite to operate for more than 6 years for global ionospheric space weather and distributions of plasma density irregularities occurrence rates at low latitudes, developed by National Central University (NCU). It is an all-in-one in-situ ion sensor, which can measure ion concentrations, velocity, and temperature to monitor ionospheric space weather conditions to detect potential disruptions in space communication and navigation systems. The 2nd-generation CIP occupied 0.8U form vector, 776g of mass, and 3.5W of power consumption. Like its predecessor, the 1st-generation CIP, it has a capability to operate as a standalone payload or installed in a sensor package with multiple roles (a Retarding Potential Analyzer, an Ion Drift Meter, an Ion Trap, etc.), and sampling rates (128 Hz and 1,024 Hz). It is a slight increase in both weight and size to its predecessor, but improves its heat dissipation issues with its aluminum alloy shell, and adds an RS-485 communication interface to enhance compatibility with most commercial Cubesat. The 1st-generation CIP has been launched with IDEASSat (2021), INSPIRESat-1 (2022), and ARCADE (2023) missions. The 2nd-generation CIP had a maiden flight on Pearl-1C (2023) in a 520 km sun-synchronous orbit in the 1030 local time sector by SpaceX. In the future, NCU will deliver three 2nd-CIPs for ELITE (anticipating a very low earth equatorial orbit in 2025), PEAL-1A and 1B (500 km sun-synchronous orbit in 2025Q4), etc. It is welcome for space scientists to consider using CIPs for their upcoming space weather mission.
IG21-A021
Lessons Learned from SPATIUM-II Technology Demonstration Mission for Ionospheric TEC Measurements
Makiko KISHIMOTO1#, Necmi Cihan ORGER1+, Tharindu DAYARTHNA2, Meng CHO1, Hoda A. ELMEGHARBEL3, Chee Lap CHOW4, Li King Ho HOLDEN4, Man Siu TSE4
1Kyushu Institute of Technology, Japan, 2Arthur C Clarke Institute for Modern Technologies, Sri Lanka, 3Misr International University, Egypt, 4Nanyang Technological University, Singapore
KITSUNE/SPATIUM-II was developed as a dual satellite system in the same structure, and both systems were operated in low-Earth orbit for approximately a year between March 2022 and March 2023. SPATIUM-II is the second mission of SPATIUM program, and it was developed in collaboration between Kyushu Institute of Technology and Nanyang Technological University. SPATIUM program utilizes chip-scale atomic clock onboard while aiming to perform 3D ionosphere mapping, and there are several capabilities to demonstrate prior to scientific-level TEC measurements. SPATIUM-II payload received spread spectrum BPSK signal at 450 MHz transmitted from the ground station, and it was demodulated and processed onboard to detect propagation time delay between the ground station and the satellite. In addition, it used chip-scale atomic clock for the receiver software-defined radio clock source, and 1-PPS from onboard GPS receiver is used for time synchronization between the ground systems and the satellite receiver. In this study, lessons learned from TEC technology demonstration mission will be discussed based on on-orbit results together with how to improve the mission to measure ionospheric TEC with a small satellite.
IG22-A008
Process of High School Students' Data Interpretation in Earthquake Education Program
Chaeeun YOON#+
Ewha Womans University, Korea, South
In this study, we developed a scientific inquiry education program using earthquake big data for high school students, and understand the characteristics of each stage in the integrated inquiry process. Big data is one of the core information technologies of the future society, and the 2022 revised curriculum in Korea also emphasizes it as part of digital education. The activity of presenting data in the form of tables or graphs and interpreting their meaning is closely related to the data interpretation skill, which is the major inquiry in earth science education. The big data used in the program is landslide data caused by earthquakes around the world provided by the United States Geological Survey, and includes a total of 12 variables including region, date, size, fault type, and number of casualties. The educational program was conducted for 120 minutes for 37 first and second year high school students in Seoul, Korea. Using actual big data, students experienced a total of six stages of research: making research questions, developing hypotheses, selecting variables, transforming data, interpreting data, and drawing conclusions. In the ‘transforming data’ stage, the CODAP (Common Online Data Analysis Platform) program, a platform that allows students to easily process data, was used. The data used in the study were graphs and inquiry journals written by students. The characteristics of high school students' inquiry stages were analyzed focusing on the data interpretation process.
IG22-A012
Elementary and Middle School Students’ Learning Characteristics in Polar Education Program : Focusing on the Framework of the Polar-relatedness
Soyoung MUN#+
Ewha Womans University, Korea, South
The purpose of the study is to identify the characteristics of students by applying the polar-Relatedness analysis criteria developed to reduce students' psychological distancing toward the poles, and to provide implications for polar education in terms of science education. Polar, one of the unexplored areas on Earth, is an important object in the development of modern science and is a emerging research topic. They are highly valued for their research due to their geographical and climatic characteristics. However, it was noted that the general public still does not know about the poles and feels psychological distance. Accordingly, it is necessary to strategically develop a polar education program that eliminates obstacles that cause psychologically distant feeling, identify the characteristics that appear therein, and derive educational implications. In the analysis by subject, stage, and student, students showed different patterns. First, as a result of the subject analysis, it was the biology topic that increased interest in the poles. Next, results of the step-by-step analysis, it was found that the main class stage in narrowing the psychological distance to the pole was the ‘Local’ stage, which became a leap point to approach the personal range from the global range. Finally, results of each student's analysis, after the application of the polar education program, the cognitive aspect of ‘perspective’ was relatively poorly revealed, and it was found that it is a key factor to have a new perspective on the polar and various perspectives on the polar to narrow the psychological distance to the polar. This study is meaningful in that it drew attention to the subject of polar, which had not been noted in science education, and led to class contents. In addition, it is meaningful to develop teaching and learning methods that fit the characteristics of polar regions that are far away.
IG27-A023
Development of Maritime Issues Detection Technology and Service System Based on a Second Geostationary Ocean Color Imager (GOCI-II)
Hee-Jeong HAN1+, Kwang Seok KIM1, Ki-Beom AHN2, Suk YOON1, Hyun YANG3, Young-Je PARK1#
1Korea Institute of Ocean Science & Technology, Korea, South, 2Sirius K, Korea, South, 3Korea Maritime & Ocean University, Korea, South
The Korean geostationary ocean color satellite, GOCI-II, is a valuable tool for monitoring the waters surrounding the Korean Peninsula. With its ability to provide 12-band datasets at a spatial resolution of 250 meters, it is ideal for observing marine phenomena. The high temporal resolution of GOCI-II also ensures stable observational performance, making it an effective tool for detecting maritime issues such as red tide, brown tide (Sargassum horneri), sea ice, and chlorophyll bloom. By utilizing GOCI-II data, a maritime issue monitoring service system has been developed and compared with field or satellite data from 2022 to 2023. Through the customization of the system to meet the practical needs of the industry and the application of AI technology, a highly effective system can be established to respond to maritime issues.
IG27-A024
Accurate Estimation of Forest Phenology with Snow-free Vegetation Index Derived from Himawari/AHI: An Application of Third Generation Satellites for Ecosystem Monitoring
Taiga SASAGAWA1#+, Wei LI2, Yuhei YAMAMOTO2, Kazuhito ICHII2, Kenlo NASAHARA1
1University of Tsukuba, Japan, 2Chiba University, Japan
Vegetation phenology plays a vital role in terrestrial ecosystems, and many researchers have tried to estimate the start of the season (SOS) and end of the season (EOS) with vegetation indices, such as NDVI and EVI, derived from satellite data. However, these vegetation indices are strongly affected by the changes in the background, especially snow melting, and it causes significant noises for SOS and EOS estimation. Currently, a new snow-free vegetation index named NDGI has been developed, and it showed better performance for the estimation of SOS and EOS with MODIS data. Furthermore, some research showed that hypertemporal datasets derived from geostationary satellites enable more accurate estimation of vegetation phenology than polar orbit satellites, such as MODIS. Therefore, in this study, we aimed to combine the geostationary satellite dataset and NDGI for accurate SOS and EOS estimation of deciduous forests around East Asia. We calculated NDGI from our land surface reflectance dataset of Himawari/AHI and conducted curve-fitting against the time series of NDGI. Then, we estimate SOS and EOS and validate them with an in-situ observation dataset provided by the Phenological Eyes Network (PEN) in some sites. Our results show that the Himawari/AHI and NDGI combination more accurately estimated SOS and EOS than previous methods. It anticipates that the combination of geostationary satellite and NDGI can be an effective way for monitoring phenology, and we can apply this method to other geostationary satellites, such as GeoKompsat-2A/AMI, MTG-I/FCI.
IG27-A025
GeoNEXtool: A Tool of GeoNEX Geostationary Satellite Data for Carbon and Water Flux Modeling at Fluxnet Sites
Hirofumi HASHIMOTO1#+, Weile WANG2, Taejin PARK2, Andrew MICHAELIS2, Ian BROSNAN2
1California State University Monterey Bay, United States, 2NASA Ames Research Center, United States
We introduce GeoNEXtool to the science communities who uses the geostationary satellite data and Fluxnet data to analyze hourly vegetation response to the climate in the continental scale. The importance of the new generation geostationary satellite has been well recognized for the Fluxnet users because only the geostationary satellites can monitor the vegetation in such a high frequency (10 minutes frequency for full disk images). However, the number of research using the geostationary satellite data for ecosystem modeling is still small. One of the biggest reasons of this poor usage of the data is that the volume of the geostationary satellite data is too large for most of the scientists. We, NEX (NASA Earth Exchange) group, are generating the GeoNEX TOA (L1G dataset) and surface (L2 dataset) reflectance dataset in the geographic projection from the geostationary satellite data. To promote the data to the Fluxnet community, we created the GeoNEX subset datasets for each fixed network site, and developed the GeoNEXtool to make it easy to process the subset dataset in the same way the MODIStool can handle MODIS subset dataset. The selection of network sites and file formats of GeoNEX subset are same with MODIS subset dataset. The subset data includes the timeseries of 9x9 1-km pixel data around the sites. Using GeoNEXtool, we demonstrated simple usage of the GeoNEX subset data with Ameriflux subset data. The estimated GPP using GeoNEX subset data were compared with the GPP at Ameriflux sites. The GeoNEXtool and GeoNEX subset data can contribute to the research of the Fluxnet community by facilitating the use of the geostationary satellite data.
IG30-A002
Evaluation of the Potential of Using Machine Learning, the Moving Average and the Savitzky–Golay Filter to Simulate the Daily Soil Temperature in Gully Regions of the Chinese Loess Plateau
Dengfeng LIU1#+, Wei DENG1, Fengnian GUO1, Lan MA1, Qiang HUANG1, Qiang LI2, Guanghui MING3
1Xi'an University of Technology, China, 2Northwest A&F University, China, 3Yellow River Engineering Consulting Co., Ltd., China
Soil temperature is an important variable in earth science research. Machine learning provides a new approach to simulating soil temperatures to compare the land surface’s energy balance with traditional simulations of physical processes. The long short-term memory (LSTM) was used to explore its potential in simulations of soil temperature at five soil depths, and the simulated soil temperature was postprocessed by the moving average (MA) and the Savitzky–Golay filter (SG). The models were forced by various combinations of environmental factors, including daily air temperature (Tair), daily net radiation (Rn), daily shortwave radiation (Rs) and the soil temperature (Ts) of the upper layer, which were observed in the Hejiashan watershed on the Loess Plateau in China. The optimal combination of input was identified as Ts-last+Tair+Rn+Rs and this served as input to the model. The other three models, (i.e., multiple linear regression (MLR), multilayer perceptron (MLP) and bidirectional long short-term memory (Bi-LSTM)) were also applied to simulate the daily soil temperatures at the five depths to compare these with the results of LSTM and LSTM-MA-SG. The simulated results were evaluated by the coefficient of determination (R2), the mean absolute error (MAE), the root mean square error (RMSE) and the Kling–Gupta efficiency coefficient (KGE). The results showed that the performance of LSTM-MA-SG in simulating Ts was better than the other four models. The evaluation of the results of LSTM-MA-SG showed that the range of R2 at different depths was 0.993-0.999, the range of MAE was 0.104-0.387 °C, the range of RMSE was 0.132-0.498 °C, and the range of KGE was 0.982-0.991. In the simulation of all depths, the performance of LSTM-MA-SG improved compared with LSTM. The study provides a critical reference for simulations of soil temperature.
IG30-A003
Evaluation of Water Balance and Water Use Efficiency with the Development of Water-saving Irrigation in the Yanqi Basin Irrigation District of China
Huan CHENG1+, Dengfeng LIU1#, Guanghui MING2, Fiaz HUSSAIN3, Lan MA1, Qiang HUANG1, Mengxian MENG4
1Xi'an University of Technology, China, 2Yellow River Engineering Consulting Co., Ltd., China, 3PMAS-Arid Agriculture University Rawalpindi, Pakistan, 4China University of Geosciences, China
With the continuous expansion of cultivated areas, there is an increasing demand for irrigation water, resulting in an irrigation efficiency paradox. In this study, the water balance method and the improved IWMI water balance method were used with remote sensing and statistical data from 1980 to 2020 to analyze the changes in the irrigation water supply, consumption, and loss for improvement in irrigation water use efficiency in the Yanqi Basin. The arable land area in the irrigation district increased from 1672 km2 in 1980 to 2494 km2 in 2020. The traditional water use efficiency showed an increasing trend. The lowest value for the field water-use coefficient was exceeded to 0.81 from 2009 to 2020. The canal water-use coefficient was increased from 0.54 in 2009 to 0.82 in 2020. The irrigation water-use coefficient increased from 0.35 in 1998 to 0.68 in 2020, with a general upward trend. The water consumption ratio indicator DFg, determined using the improved water balance method, increased from 0.8390 in1980 to 0.8562 in 2020, and the average was 0.8436. Cultivated land’s actual irrigation water consumption per unit area reached the highest value of 8.41×106 m3/hm2/a in 2011 and the minimum value of 4.01×106 m3/hm2/a in 2020, while the total water diversion showed an increasing trend due to the continuous expansion of arable land. From 1980 to 2020, water diversion into the irrigation district changed from 1.214 km3 to 1.000 km3, and it reached a maximum of 1.593 km3 in 2000; water diversion into the irrigation district showed an overall upward trend. The findings showed an increase in IWUE in the Yanqi Basin irrigation district. These results provide a theoretical basis for breaking the paradox of irrigation efficiency, which can be used in the water resource management of irrigation districts.
IG30-A006
Spatial-temporal Adaptive Planning of Flood Managed Aquifer Recharge Guided by Deep Reinforcement Learning
Meilian LI+, Xiaogang HE#
National University of Singapore, Singapore
Flood-managed aquifer recharge (Flood-MAR) is a crucial yet untapped solution that can simultaneously mitigate flood and drought risks and boost groundwater supply. The key constraint for this multi-benefit adaptation is how much water and land are available for recharge, which requires effective planning from spatial and temporal aspects to repurpose existing agricultural landscapes. However, future evolution of such landscapes is subject to uncertainties in changing human-nature systems, which is further challenged by a warming climate that brings more floods and droughts. Traditional spatial planning approaches often assume perfect future predictions and rely on static policies, which cannot well adapt to dynamic and uncertain environments. While some studies have adopted artificial intelligence to facilitate urban planning and conservation planning, most of them do not consider the synergistic benefits of multi-sector planning across time and space. To address these limitations, we propose a multi-objective spatially explicit planning framework leveraging recent advances in deep reinforcement learning (DRL) to develop adaptive policies for joint agriculture and water management, with a focus on Flood-MAR. DRL Agents dynamically allocate lands for different Flood-MAR related actions (e.g., land retirement, on-farm-based, or wetland-based recharge) over time. The environment in DRL responds to these actions through changes in land use and groundwater storage. Rewards in DRL are designed based on agricultural revenues, action costs, and penalties for not meeting groundwater targets. To tackle the challenge of the vast solution space, we use the proximal policy optimization algorithm to train a policy network to estimate pixel-level action probabilities and a value network to evaluate the current policy. Through a case study in California’s San Joaquin Valley, this study underscores the potential of Flood-MAR to mitigate flood and drought risks in uncertain environments and the efficacy of DRL in guiding adaptive Flood-MAR policies.
IG30-A009
Assessment of Groundwater Contamination with PFAS and Possible Remedial Measures
Harish BHANDARY#+, Adnan AKBAR, Chidambaram SABARITHINAM, Yogeesha JAYARAMU
Kuwait Institute for Scientific Research, Kuwait
Per and Ployfluoroalkyl substances (PFAS) are class of chemicals belonging to a group of emerging contaminants (ECs), previously thought to be of little significance only to be discovered recently that they can adversely affect aquatic ecosystems and human health. Being the only source of natural freshwater, groundwater is of strategic importance for Kuwait. The new discoveries of different emerging contaminants have given rise to the notion that this resource has been contaminated by these pollutants, especially by the PFAS which are linked to oil field activities and firefighting foams that were used to combat the fire blazes in the immediate aftermath of the 1991 Gulf War. This study is aimed at evaluating the levels of contamination by PFAS of the fresh groundwater fields in Kuwait and recommending possible remediation measures. This paper presents the initial activities and results of the study to evaluate PFAS contamination of fresh groundwater. The study involved collection and analyses of groundwater samples from Al-Raudhatain and Umm Al-Aish fresh groundwater fields to determine the PFAS levels in them. The collected samples were analyzed for PFAS using standard USEPA method No. 8327 using liquid chromatography/tandem mass spectrometry (LC-MS/MS). Initial results indicate the contamination of groundwater with PFAS. Preliminary data of this ongoing study and future plans is discussed in this paper. This research will employ GIS and statistical tools to map pollutant distribution, identifying potential sources. Various treatment techniques, including granular activated carbon, ion exchange, and nanofiltration, will be explored for their efficacy in PFAS removal within a pump-and-treat system. The collected data and information will be assembled to evaluate the degree of contamination and effective remedial methods. The study is expected to provide the information necessary for the development of a comprehensive groundwater management and protection plan to address risks associated with such contaminants.
IG30-A010
Semi-analytical Rate-limited Sorption Model for Sequential Degradation Products in Two-dimensional Multispecies Transport Under Arbitrary Time-dependent Inlet Boundary Conditions
Thu-Uyen NGUYEN#+
National Central University, Taiwan
Many semi-analytical and analytical models for multispecies transport traditionally rely on solving coupled advection-dispersion equations (ADEs) in which sorption is commonly assumed to be equilibrium-controlled. Nevertheless, recent demonstrations indicate that more accurate predictions of contaminant transport in groundwater can be achieved by incorporating a rate-limited sorption process instead of relying on an equilibrium sorption assumption. This study is specifically designed to formulate semi-analytical models for the two-dimensional multispecies transport of a chemical mixture, including a parent compound and its degradation-daughter products. This transport is influenced by rate-limited sorption subject to arbitrary time-dependent inlet boundary conditions. The simulation results obtained from our newly developed semi-analytical model closely align with those generated using a numerical model based on the Laplace transform finite difference (LTFD) method, affirming the validity of the new approach. The impact of rate-limited sorption on contaminant plume migration is investigated under various time-dependent inlet boundary conditions. Sorption rate values vary from low to high, specifically 0.05, 0.5, 5, and 50 years⁻¹. The outcomes reveal that the predicted concentrations of all contaminants within the decay chain decrease as the sorption rate constant increases, irrespective of whether the boundary sources are constant or exponentially time-dependent. However, under pulse loading boundary conditions, concentrations of later degradation products tend to increase with the sorption rate constant. The semi-analytical models developed in this study accommodate different inlet boundary conditions, making them versatile tools for simulating the transport of sequentially degrading reaction products.
IG30-A011
The Presence, Ecotoxicological Implications, and Associated Risk Assessment of Organochlorine Pesticides (OCPs) in Rural and Semi-urban Areas of West Bengal, India
Suchitra MITRA#+, Sujata RAY
Indian Institute of Science Education and Research Kolkata, India
Despite the global prohibition of persistent organochlorine pesticides, recent evidence suggests continued usage in India for vector-borne disease eradication programs and agricultural practices. This study assesses the presence, ecotoxicological implications, and associated risks of 20 banned and restricted organochlorine pesticides (OCPs) in the surface water and groundwater of semi-urban (Nadia district) and rural areas (Birbhum district) across the winter, premonsoon, monsoon, and postmonsoon seasons. The OCP concentrations in the collected water samples were observed in the 1.53 to 106.90 ng/L range and were higher in the winter than in other seasons. The detection frequencies of OCPs ranged from 85% to 100% in all seasons, with a few exceptions. The highest mean concentrations were observed for ΣHCHs, ΣEndosulfans, and chlordane during the pre-monsoon and monsoon, except for the absence of chlordane in groundwater during the pre-monsoon season. In the winter season, ΣHCHs concentration in groundwater increased in agricultural areas. In semi-urban areas, OCPs like HCH, endosulfan sulfate, and heptachlor were in higher concentrations in the water. Concentrations of β-HCH, γ-HCH, p,p’-DDD, p,p’-DDT, and ΣEndosulfans in both groundwater and surface water did not significantly differ between pre-monsoon and monsoon seasons, except for elevated concentrations of aldrin and β-HCH in the monsoon season and ΣEndosulfans in the post-monsoon season in groundwater. OCPs like Aldrin, heptachlor, and trans-chlordane were in higher concentrations in the water of rural areas. The study highlights the poor and delayed implementation of the ban on OCPs in India, emphasizing the urgent need for stricter adherence to the National Implementation Plan (NIP) on pesticides to mitigate associated environmental risks.
IG30-A015
Biomonitoring of Heavy Metals in Blood, Urine and Hair of Children (5-18 Years) in Communities of Atoyac Basin, Mexico
Estefania MARTINEZ-TAVERA1#+, Carlos David YEVERINO-MARTINEZ1, S.B. SUJITHA2
1Universidad Popular Autónoma del Estado de Puebla, Mexico, 2Instituto Politécnico Nacional, Mexico
In central Mexico, the Atoyac basin encompasses an active volcano and the second most polluted basin. Henceforth, the study aims to determine the prevalence of heavy metals in the biological matrices of children aged between 5 to 18 whom reside in the Alto Atoyac watershed. Blood, urine, and hair samples were collected from 78 minors in two populations and the concentrations of six heavy metals (As, Cd, Cr, Pb, Hg, and Zn) in all three matrices were determines using a ICP MS. Clinical tests were also done on blood and urine samples to examine the health state of the study population. The findings revealed that nearly 96% of the individuals had at least one parameter in the clinical tests that was above the reference levels. The analysis of heavy metals in blood samples revealed that 100% of participants in both communities had levels that exceeded the reference values for arsenic, cadmium, chromium, and lead, indicating a potential health risk, except for mercury and zinc, where 88% and 98.6%, respectively, had levels that exceeded the reference. In terms of urine samples, HQ index values for arsenic showed no health risk, but chromium and lead values surpassed the reference for 100% of participants in both areas, while cadmium, mercury, and zinc showed a potential danger in 52%, 34%, and 52% of participants, respectively. Arsenic, cadmium, chromium, lead, and zinc levels in hair samples above were above the reference in 86%, 98%, 100%, 17%, and 21% of participants, respectively, whereas mercury posed no health risk in any of the samples. Correlations were found between different heavy metals in the study and their amounts and aberrant clinical indicators, revealing that these pollutants have a negative impact on participant’s; health and that many contaminants coexist in the region.
IG31-A003
Vulnerability to Extreme Heat Risks in Southeast Asia: The Role of Adaptation Finance
Rainbow Yi Hung LAM#+, Laurence DELINA
The Hong Kong University of Science and Technology, Hong Kong SAR
Extreme heat, often referred to as the “silent killer,” can have devastating impacts, particularly in regions like Southeast Asia, where urbanization and population density are on the rise. Vulnerable populations in this world region are at a higher risk of heat-related mortality and illness, as well as reduced efficiency of energy systems. Additionally, limited financial resources hinder cities' adaptation efforts, making finance a key determinant of vulnerability to extreme heat risks. While there is a growing awareness of the importance of financing climate adaptation and mitigation through the SENDAI framework for Disaster Risk Reduction (SFDRR) and the recent Conferences of Parties (COPs) to the United Nations Framework Convention on Climate Change, there is still a lack of holistic understanding of adaptation or disaster risk management efforts for urban extreme heat and a dearth of adaptation finance. This presentation reviews the literature on adaptation efforts addressing extreme heat in Southeast Asia and maps adaptation finance according to SENDAI priorities, particularly Priority 3. The review highlights the prominence of extreme heat adaptation projects, particularly green roofs or spaces, in Southeast Asia, which tend to emphasize the financing aspects more. However, we also demonstrate the need for more financial investment and research on extreme heat adaptation solutions for this region. Future research, thus, may explore extreme heat research projects in Southeast Asia and their corresponding funding organizations for further insight into adaptation finance for extreme heat.
IG31-A007
CMIP6 Dynamical Downscaling for Future Climate Projections: Enhancing Urban Resilience Against Heatwaves and Extreme Precipitation in the Pearl River Delta
Ziping ZUO#+, Zhenning LI, Mau Fung WONG, Jimmy Chi Hung FUNG, Alexis LAU
The Hong Kong University of Science and Technology, Hong Kong SAR
In the past year of 2023, recorded as Earth’s hottest year, various extreme weather events, including flash floods, record-shattering heatwaves, and wildfires, ravaged many regions. Increasing greenhouse gas emissions may pressurize the expanding cities, causing elevated morbidity and mortality, power and water resources shortages, infrastructure breakdowns, economic impact, and more. Urban areas need to adapt and enhance resilience to withstand the challenges posed by worsening future weather conditions, leveraging scientific insights. In this study, by applying a dynamical downscaling method on the multi-model ensemble from CMIP6, we present model projections of heatwaves and extreme precipitation in the Pearl River Delta (PRD), one of the most densely populated and urbanized regions in the world. We considered three Shared Socioeconomic Pathways (SSP1-2.6, SSP2-4.5, and SSP5-8.5) in the mid-21st century (2040-2049) and at the end of the century (2090-2099). Our results suggest heatwaves are expected to be more frequent, intense, extensive, and longer-lasting for SSP2-4.5 and SSP5-8.5 until the end of the 21st century. Under the worst-case scenario of SSP5-8.5, extreme heat events that occurred once every 10 years in the 2010s are expected to occur once each month from June to September. While heatwaves pose significant challenges, they can also contribute to altered atmospheric conditions, potentially amplifying the likelihood of extreme precipitation events. Extreme hourly precipitation shows intensification under SSP2-4.5 and SSP5-8.5 in the 2040s and 2090s. Specifically, in the 2040s, 55.6% of the land area experiences an increase in extreme hourly precipitation under SSP2-4.5, while 69.5% under SSP5-8.5. By the 2090s, these percentages will rise to 75.6% and 88.6%, respectively. Hong Kong as a key megacity within the PRD, will experience the highest hourly precipitation of 233 mm/h in the 2040s and 295.1 mm/h in the 2090s, a significant rise from 158.1 mm/h in the 2010s.
IG31-A010
Impacts of Concurrent Hot and Dry Extremes on Hydropower Demand and Supply in the Pearl River Delta Realized by Convection-permitting Projections
Zixuan ZHOU1+, Eun-Soon IM1#, Lin ZHANG2
1The Hong Kong University of Science and Technology, Hong Kong SAR, 2City University of Hong Kong, Hong Kong SAR
The Pearl River Delta (PRD) is a low-lying area surrounding the Pearl River estuary, comprising several most densely populated megacities in Southeast Asia. In the summer of 2022, PRD experienced a prolonged period of record-breaking high temperatures, straining the power grid due to increased air-cooling demand. Concurrently, droughts in major hydropower generating regions exacerbated the power supply-demand gap. Considering China's goal of achieving carbon neutrality by 2060, ensuring energy security becomes crucial in the context of worsening climate extremes and the growing reliance on hydropower. Therefore, it is essential to understand how concurrent hot and dry extremes (CHDE) will affect the hydropower sector to facilitate climate adaptation decisions. Most existing studies have focused on univariate analysis of single extremes using coarse-grid global climate models (GCMs), which may not fully capture the region-specific climate impacts of global warming. To address this issue, this study will integrate convection-permitting (CP) regional climate modeling to assess the impacts of CDHE on hydropower demand and supply. Specifically, the Weather Research and Forecasting (WRF) model will be employed to downscale the bias-corrected CMIP6 GCM projections under the SSP5-8.5 scenario over southeastern China. The downscaled high-resolution and high-quality regional dataset will provide an enhanced representation of regional hydroclimate conditions. Subsequently, the climate data which translates into cooling demand and water scarcity, will be combined with plant-level electricity generation data, installed capacity, and other non-climatic data at the provincial level, including investment and electricity price. By using regression models, this combined dataset will enable empirical estimation of the relationship between CDHE and hydropower demand and supply. Overall, the project would be a meaningful showcase of the synergy of interdisciplinary collaborations in climate change adaptation. [Acknowledgements] This research was supported by the Theme-based Research Scheme, T31-603/21-N, which was funded by the Research Grants Council (RGC) of Hong Kong.
IG31-A011
Evaluation of Spatial Extent of Urban Heat Island Over Major Cities in Punjab, India
Vinayak BHANAGE#+, Han Soo LEE
Hiroshima University, Japan
The Urban Heat Island (UHI) is the most evident example of microclimatic modification induced by the process of urbanization. In the past, many researchers have shown interest in studying the UHI phenomenon. However, these studies, especially over the Indian region, rarely examined the UHI effect in terms of its spatial extent. In this study, with the aid of land surface temperature data obtained from the MODIS satellite (2003-2018), the seasonal and diurnal variations in the spatial extent of UHI were analysed over the seven cities of Punjab, India. In this process, a single exponential decay model was adopted to determine the spatial extent of the UHI effect. Results show that the UHI effect's spatial extent varies substantially concerning the cities and time. During the daytime of the summer (winter) season, the spatial extent of the UHI effect ranges from 4.1 to 6.97 (4.63 to 5.95) times of the urban area, while in the night period of summer (winter) season, this extent spans between 3.92 to 9.83 (2.49 to 7.88) times of the urban area. The results of this study depict that UHI affects not only the urban areas but also the microclimate of the surrounding rural area.
IG32-A002
Evaluate the Effectiveness of Machine Learning in Training Urban Canopy Temperature in China
Yingchao GUO#+, Ning ZHANG
Nanjing University, China
Urban heat waves can intensify the urban heat island effect, which has profound implications for public health and the climate system. However, current numerical models of the Earth lack accurate representations of urban areas and involve intricate parametric physical processes, resulting in significant uncertainties in forecasting daily canopy temperature. Therefore, this study aims to assess the efficacy of machine learning methods in training canopy temperature. In this paper, we employ the xgboost method to establish a training emulator. The atmospheric forcing field data, encompassing downward long-wave radiation, downward short-wave radiation, temperature, humidity, wind speed, air pressure, and precipitation, that drives the Community Land Model (CLM) in the Community Earth System Model (CESM), is utilized as the training set. Through this training process, we derive the maximum daily canopy temperature. The results demonstrate the following:(1)he machine learning method yields a root mean square error (RMSE) for canopy temperature and a simulated canopy temperature from the CESM model, both below 1.4℃, with an R2 above 95%;(2)Significant overestimations are observed in most parts of the Beijing-Tianjin-Hebei region, southern region, and Fen-Wei Plains region, while noticeable underestimates occur in Tianjin and the Yangtze River Delta region;(3)Additionally, by substituting canopy temperature with ten land cover zone (LCZ) conditions, we observe that the RMSE of LCZ1 is the smallest, while LCZ5 exhibits the largest values. This indirectly indicates variations in machine learning methods across different LCZ cases;(4)Furthermore, by utilizing machine learning to expedite numerical model calculations and unveil underlying physical processes, we contribute to addressing the interpretability issue of machine learning to some extent.
IG32-A005
Optimizing Climate Data Analysis Workflows: Strategies and Lessons Learned from Two Case Studies
Alexander GOODMAN1#+, Colin RAYMOND2, Peter KALMUS1
1California Institute of Technology, United States, 2University of California Los Angeles, United States
Computational requirements for climate data analysis workflows have been dramatically increasing in recent years. This has occurred in proportion to increases in both spatial and temporal resolution and particularly in the context of climate models, the number of datasets being used in the analysis as well. Much emphasis on the tooling that has been developed to meet this need has been in the form of big data frameworks and parallel computing to perform analyses at scale. However, the use of such tools also has some important drawbacks such as creating additional complexity in the analysis codes (and therefore more prone to bugs) as well as potentially being very expensive. In practice, we have found that in many cases, climate data analysis codes are poorly optimized and so greater benefits can be gained simply from writing more performant code without immediately resorting to parallelism. Here we will describe examples of how to optimize publicly available codes for two algorithms that are highly relevant in recent climate modeling studies. These will include calculations for wet bulb temperature (Davies-Jones 2008) and extended heat index (Lu and Romps, 2022). We will demonstrate how our optimizations were able to improve the running times of both use-cases by two orders of magnitude.
IG32-A013
Evolving the Coupled Model Intercomparison Project (CMIP) to Better Support the Climate Community and Future Climate Assessments
Helene HEWITT1, John DUNNE2, Julie ARBLASTER3, Olivier BOUCHER4, Paul DURACK5, Tomoki MIYAKAWA6+, Matthew MIZIELINSKI1, Robert PINCUS7, Sasha AMES5, David HASSEL8, Birgit HASSEL9, Forrest HOFFMAN10,11, Martin JUCKES12, Guillaume LEVAVASSEUR4, Chloe MACKALLAH13, Vaishali NAIK2,14, Atef BEN NASSER4, Benjamin SANDERSON15, Isla SIMPSON16, Martina STOCKHAUSE17, Karl TAYLOR5, Beth DINGLEY18#, Daniel ELLIS18, Eleanor O'ROURKE18, Briony TURNER18
1Met Office, United Kingdom, 2NOAA Geophysical Fluid Dynamics Laboratory, United States, 3Monash University, Australia, 4Institut Pierre-Simon Laplace, Sorbonne Université / National Centre for Scientific Research, France, 5Lawrence Livermore National Laboratory, United States, 6The University of Tokyo, Japan, 7Columbia University, United States, 8National Centre for Atmospheric Science / University of Reading, United Kingdom, 9Deutsches Zentrum für Luft- und Raumfahrt, Germany, 10Oak Ridge National Laboratory, United States, 11University of Tennessee, Knoxville, United States, 12UK Research and Innovation, United Kingdom, 13Commonwealth Scientific and Industrial Research Organisation, Australia, 14NOAA Oceanic and Atmospheric Research, United States, 15Centre for International Climate and Environmental Research, Norway, 16National Center for Atmospheric Research, United States, 17German Climate Computing Center, Germany, 18CMIP International Project Office, United Kingdom
Over four decades, CMIP has driven massive improvements in the modelled representation of the Earth system, whilst also seeing huge growth in its scope and complexity. In its most recent phase, CMIP6, a broad spectrum of questions continues to be answered across twenty-four individual model intercomparison projects (MIPs). This science improves process understanding and assesses the climate’s response to forcing, systematic biases, variability, and predictability in line with WCRP Scientific Objectives. CMIP and its associated data infrastructure have become essential to the Intergovernmental Panel on Climate Change (IPCC) and other international and national climate assessments, increasingly including the downstream mitigation, impacts, and adaptation communities. However, despite the invaluable science produced from CMIP6 data, many challenges were still faced by the model data providers, the data delivery infrastructure, and users, which need to be addressed moving forwards. A specific challenge in CMIP6 was the burden placed on the modelling centres, in part due to the large number of requested experiments and delays in the preparation of the CMIP6 forcing datasets and climate data request. The CMIP structure is evolving into a continuous, community-based climate modelling programme to tackle key and timely climate science questions and facilitate delivery of relevant multi-model simulations. This activity will be supported by the design of experimental protocols, an infrastructure that supports data publication and access, and quasi-operational extension of historical forcings. A subset of experiments is proposed to be fast-tracked to deliver climate information for national and international climate assessments and informing policy and decision making. The CMIP governing panels are coordinating community activities to reduce the burden placed on modelling centres, continue to enhance novel and innovative scientific activities, and maximise computational efficiencies, whilst continuing to deliver impactful climate model data.
IG32-A014
Hydroclimate Changes in Sri Lanka Over the Past Millennium
Liangcheng TAN1#+, Xiqian WANG2, Jin ZHANG2, Mathara SAMANMALI3, Le MA2, Qiang LI2, Gang XUE4, Jingjie ZANG2, Hai CHENG5, Ashish SINHA6
1Institute of Earth Environment, Chinese Academy of Sciences, China, 2Chinese Academy of Sciences, China, 3University of Colombo, Sri Lanka, 4Northwest University, China, 5Xi'an Jiaotong University, China, 6California State University Dominguez Hills, United States
Understanding the decadal variability of hydroclimate in Sri Lanka during historical time is of great significance for disaster prevention, water resource management, and agrarian economy in this area under a global warming scenario. However, the evolution patterns and driving mechanisms of decadal hydroclimate variability in Sri Lanka remain controversial due to lack of long-term and high-resolution records. Here we present the first temporally highly resolved stable isotope and trace element stalagmite records in Sri Lanka, covering the past millennium. Our records agree well with other hydroclimate records over Indian subcontinent and Southeast Asia, showing two notable humid intervals during the early medieval period and the late Little Ice Age, and prolonged droughts from 1200 to 1600 CE. By using this record, we further explore the driving mechanisms of hydroclimate variations in Sri Lanka on centennial to decadal scales and its possible impacts on societal changes.
IG32-A015
Temperature Variability Associated to Climate Feedback Using an Energy Balance Model
Hyoji KANG+, Minjeong CHO, Yong-Sang CHOI#
Ewha Womans University, Korea, South
Understanding the temperature variability is essential as the incidence of climate extremes changes potentially, even with a small shift of mean climate. However, the fundamental mechanism of temperature variability behind changes in mean climate remain largely unknown. Especially, the change in temperature variability might be subject to equilibrium climate sensitivity (ECS), as ECS implies the extent of Earth’s temperature response to increasing atmospheric greenhouse gases. In this facet, this study aims to explore the characteristic of temperature variability in response to the random forcing by different ECS. We assessed the change in temperature variability using the standard deviation of simulated temperatures for 100 years. For the sake of looking into this effect, we utilized the energy balance model, which enables to manipulate climate feedback and heat capacity for simulation of temperature in simple but conceptually physical framework. We obtained the relationship between the ECS and temperature variability, so that the future study can estimate the ECS by using temperature variability from observational data.
IG32-A020
Machine Learning-based Wildfire Detection Considering Environmental Context and Structural Similarity
Taejun SUNG#+, Yoojin KANG, Jungho IM
Ulsan National Institute of Science and Technology, Korea, South
The frequency and intensity of wildfires are increasing annually due to climate change, resulting in diverse and escalating damages. Geostationary satellites provide real-time acquisition of extensive wildfire information, facilitating efficient wildfire monitoring. Traditional satellite-based wildfire detection algorithms utilize the brightness temperature difference between the central pixel and its surrounding pixels to detect thermal anomalies. However, performance may decrease when environmental conditions vary between the central and surrounding areas. Additionally, traditional algorithms only employ vector-based statistics, neglecting the structural information inherent in satellite imagery. In this study, to address the limitations of traditional satellite-based wildfire detection algorithms, we developed a wildfire detection model based on light gradient boosting machine (LGBM) that incorporates environmental context correction and structural similarity comparison. The GK-2A (GEO-KOMPSAT-2A) satellite imagery was used as the main input variable, and the Korea Forest Service wildfire occurrence data was used as a reference. The detection performance of the LGBM model (precision: 98%, recall: 95.2%, F1 score: 96.6%) showed improvement after the application of environmental context correction and structural similarity comparison (precision: 98.5%, recall: 96.2%, F1 score: 97.3%). The LGBM model detected small wildfires that the GK-2A wildfire detection algorithm failed to identify and demonstrated higher detection rates. Environmental context correction and structural similarity comparison improved detection performance for extreme environmental conditions and low intensity wildfires, respectively. This study analysed error factors in traditional satellite-based wildfire detection algorithms and proposed a novel approach to improve them, suggesting the potential for a generalized wildfire detection model applicable in various environments.
IG33-A001
Spatial Characteristics of Daily Max/min Urban Surface Temperatures During Heat Waves: A Case Study of the Tokyo Metropolitan Area
Moena FUKATSU#+, Yuhei YAMAMOTO, Kazuhito ICHII
Chiba University, Japan
In response to climate change, Japan is facing an increasing frequency and intensity of heatwaves. The synergistic effect of urban heat island and extreme heat poses serious societal impacts, including heightened risk of heatstroke and increased energy consumption. This study investigates the relationship between land surface temperature (LST) and urban land use during heatwave conditions, under the assumption that the primary factor shaping the thermal environment during extreme heat lies in the urban surface structure. Focusing on the heatwave events of 2018 and 2023, we used Himawari-8/9 LST product to estimate the spatial distribution of daily maximum and minimum temperatures (Tmax and Tmin, respectively) in the urban areas surrounding Tokyo. The Himawari-8/9, Japan's geostationary satellite, LST product offers a spatial resolution of 0.02° and a temporal resolution of 10 minutes. The results showed that the Tmax during heatwaves exhibited a positive anomaly of 10–20°C in croplands and paddy fields and 5–10°C in urban areas compared to normal years (2015–2022). This suggests significant reflections of changes in surface water balance. On the other hand, the Tmin during heatwaves exhibited a higher positive anomaly in urban areas, indicating a strong influence of heat storage effects. Interestingly, in suburban inland areas, there were regions where high positive anomalies of both Tmax and Tmin overlapped, suggesting a compound effect of heat storage and local surface water balance.
IG33-A006
Optimization of Electric Scooter Placement Stations Based on Distance Tolerance
Yingqiu LONG1+, Jianwei YUE1#, Shaohua WANG2
1Beijing Normal University, China, 2Chinese Academy of Sciences, China
The proliferation of shared electric scooters(E-scooter) has facilitated urban mobility, yet it has also brought challenges such as disorderly parking and supply-demand imbalances. Given the current inadequacies in the quantity and spatial layout of shared E-scooter deployments, as well as the insufficient research on deployment stations, this study extensively analyzed the travel characteristics of E-scooters in Chicago. Employing a multi-criteria decision-making approach to assess factors influencing demand, the study aimed to maximize user coverage and minimize the total distance from users to deployment stations. To achieve this optimization goal, a distance tolerance-based maximum coverage placement model was developed, and a Deep Reinforcement Learning(Deep RL) was designed to solve the model. Case experiments confirmed the model's effectiveness in optimizing the layout of E-scooter deployment stations. The study also validated the feasibility and effectiveness of applying the Deep RL algorithm in facility placement by comparing its results with those obtained using the Gurobi solver and a genetic algorithm. This study provides valuable reference for the selection of facility sites for urban shared transportation, and the application of Deep RL offers a efficient perspective and approach to facility placement problems.
IG33-A007
Spatial Sensitivity Analysis of Evapotranspiration Over Korea Peninsula
Chanyoung KIM+, Kijin PARK, Jongmin PARK#
Korea National University of Transportation, Korea, South
Global climate change and extreme climate conditions intensifies the regional imbalance of water cycle components and water security, which leads to enhancing necessity of developing efficient and sustainable water management system. In Korea, Evapotranspiration (ET), a sum of soil evaporation and plant transpiration, accounts for 43% of annual precipitation, and thus, plays a critical role in understanding regional land-atmospheric components. ET is sensitive to various hydrometeorological variables such as relative humidity, air and surface temperature, vapor pressure, wind speed, and solar radiation. However, the sensitivity of each variable to ET may different based on the various conditions including land cover and land use, climate characteristics, and topographical aspects. Accordingly, this study explored spatial sensitivity of ET towards various hydrometeorological variables over Korea Peninsula. Specifically, this study implemented hydrometeorological variables from ECMWF Reanalysis version 5-land (ERA5-Land) from 2000 to 2022. In addition, various spatial sensitivity analysis tools (e.g., Regional Sensitivity Analysis [RSA], Variance-Based Sensitivity Analysis [VBSA], the Fourier Amplitude Sensitivity Test [FAST]) were employed to quantify the sensitivity of various hydrometeorological components to ET at grid scale. Acknowledgement: This work was supported by Korea National University of Transportation 2024.
IG33-A013
Consumption of Building Energy Under Global Warming
Soohyun AHN#+, Woosok MOON
Pukyong National University, Korea, South
Since the 1980s, an overabundance of urban land development projects has led to the proliferation of satellite cities around major metropolitan areas, resulting in heightened population density and increased energy demand. This surge in urbanization has amplified concerns regarding energy supply, highlighted by the widespread blackout incident on September 15, 2011, when reserve power dropped to 0 kW. A late-summer heatwave during this period was identified as a significant contributor to the escalating power demand, underscoring the necessity of accurate energy forecasting in the context of climate change. Consequently, precise prediction of building energy consumption has emerged as a critical issue. Among the tools gaining prominence for building energy simulation is EnergyPlus, developed by the U.S. Department of Energy (DOE). EnergyPlus simulates various aspects of building energy use, including heating, cooling, lighting, and ventilation. Operating fundamentally on weather data and building information, EnergyPlus underscores that weather is crucial for building energy simulation and energy-efficient retrofitting, given the strong dependency of building energy use on weather conditions. This model can serve as a significant indicator for power fluctuations due to climatic changes. The current study employs EnergyPlus to analyze trends in power variations in response to climatic shifts, aiming to enhance the predictability of electricity demand. This research supports the development of sustainable urban energy plans, taking into account the increasing variability and unpredictability associated with climate change. By focusing on building energy simulations, this study contributes to a deeper understanding of the relationship between urbanization, climate change, and energy demands, providing valuable insights for future urban planning and energy policy.
IG33-A014
Evolving Urban Heat Islands: A Comprehensive Study of Temporal and Developmental Impacts in South Korean Cities
Mijeong JEON#+, Woosok MOON
Pukyong National University, Korea, South
In the context of climate change induced by global warming, the urban heat island (UHI) effect has emerged as a significant concern in urban areas. As cities continue to expand, artificial structures such as buildings and roads increasingly absorb solar energy, leading to a rise in energy consumption. Recent research has predominantly focused on temperature differences between urban and rural regions, examining trends in urban temperature changes over time and daily temperature variations. However, it is worth noting that comprehensive studies on the UHI effect remain limited. Specifically, there has been a lack of research that simultaneously considers variables such as year, hour, and temperature in the context of daily cycle variations. Moreover, in the case of South Korea, there is still no research that utilizes data from all observation stations. With this backdrop, our study aims to investigate how the temperature daily cycle has evolved over time, from the past to the present, as a result of the UHI effect. We conduct an analysis using all data across South Korea and develop a simplified surface energy balance model for qualitative understanding. Additionally, we conduct a comparative analysis between major cities and newly developed urban areas to see the time-evolution of the UHI effect during the growth of a new city. The results of study align with previous research, indicating a more pronounced UHI effect during nighttime compared to daytime. Seasonally, this effect is most evident in autumn. We find that the UHI phenomenon is more noticeable in accordance with urban development. This distinction becomes particularly clear when comparing established major cities with newer urban regions. These findings suggest that the degree of urban development plays a significant role in the manifestation of the UHI effect, which has important implications for urban planning and sustainability in varying urban environments.