Evapotranspiration (ET) plays a key role in the water cycle and balance, and its estimation is of paramount importance in hydrological studies. This variable is also strongly influenced by the land use and land cover (LULC). This study use a synthesis approach to analyse the relationship between ET variation and LULC transformation in the period 1990–2020. The study area represents a major agricultural region in the Mekong Delta of Vietnam between the Mekong and Bassac rivers. The multivariate dataset was ingested into the hydrological model Soil and Water Assessment Tool to examine the multidecadal ET evolution across the study region. The ET evolution is investigated together with LULC changes to identify the primary drivers of the ET change. The results reveal an increase in landscape fragmentation together with a decrease in ET. The reduction in ET is associated with the decrease in water surfaces corresponding to upstream dam operations and with the construction of new urban areas. In the coastal area, a conversion from natural areas (i.e. mangroves cover) to aquaculture has been observed, leading to a slight increase in ET.

High-resolution air-temperature fields are essential for climate, hydrologic, and ecological applications in complex terrain, yet operational products often lack the spatial detail to resolve topographic effects. We develop an observation-driven reconstruction of daily air temperature fields for South Korea (2024) using ordinary kriging with lapse-rate correction (OKLR), integrating a dense network of over 500 stations from the Automatic Mountain Meteorology Observation System (AMOS) and the Automated Surface Observing System (ASOS). The OKLR framework systematically removes elevation-driven trends using a physically based fixed lapse rate (-6.5 degrees C km(-1)), performs kriging on detrended residuals, and reapplies Digital Elevation Model (DEM)-based corrections to generate high-fidelity daily fields at a 270 m grid spacing. Unlike numerical weather prediction (NWP) models that simulate atmospheric processes, this approach reconstructs spatially continuous fields directly from dense in situ observations, ensuring empirical grounding. Extensive daily spatial cross-validation (n = 37,813) demonstrates that OKLR (MAE = 0.656 degrees C) significantly outperforms elevation-unadjusted ordinary kriging by approximate to 37% and the operational 1.5 km LDAPS product (MAE = 0.895 degrees C) by 27%. This performance gain is particularly pronounced in high-elevation zones (>700 m) and natural surfaces (approximate to 73% of the study area), where topographic complexity is greatest. The final observation-constrained reconstruction attains a robust MAE of 0.462 degrees C with near-zero bias over 188,318 station-days. As the first nationwide daily temperature dataset for South Korea at 270 m resolution, this study provides a critical foundation for precision agriculture, ecosystem monitoring, and climate change adaptation in topographically diverse environments.

Estimating net ecosystem exchange (NEE) in agricultural ecosystems is essential for quantifying carbon fluxes and supporting carbon neutrality strategies. However, satellite-based NEE products developed at coarse spatial resolutions often exhibit limited applicability to fragmented agricultural landscapes, where mixed-pixel effects and scale mismatch obscure field-scale carbon exchange signals. This limitation is particularly pronounced in South Korea, where rice paddies are widely distributed within complex terrain and exhibit strong seasonal variability driven by intensive management practices. In this study, we developed a high-resolution, grid-based NEE estimation framework for rice paddy areas in South Korea with a daily temporal resolution and a spatial resolution of 500 m. The proposed approach integrates satellite-derived MODIS vegetation indices—normalized difference vegetation index (NDVI), fraction of photosynthetically active radiation (FPAR), and leaf area index (LAI)—with meteorological variables from the Local Data Assimilation and Prediction System (LDAPS) of the Korea Meteorological Administration. Machine learning models based on Extremely Randomized Trees (ERT) and Automated Machine Learning (AutoML) were constructed and evaluated using a leave-one-year-out (LOYO) cross-validation scheme. Both models exhibited robust and stable predictive performance, with AutoML showing slightly more consistent accuracy than ERT. To address systematic bias and the tendency of data-driven models to smooth extreme NEE values, an empirical cumulative distribution function (ECDF)–based correction was applied. Although this correction led to a marginal in-crease in pixel-wise error metrics, it substantially improved the representation of NEE distributional characteristics, resulting in more realistic spatiotemporal patterns of ecosystem carbon uptake and release. Site-specific evaluations demonstrated high ac-curacy at the Cheorwon and Naju flux-tower sites, while relatively larger errors at the Gimje site were primarily associated with periods of reduced observational data quality. Despite these limitations, the proposed framework consistently preserved realistic seasonal NEE dynamics. Comparison with the global SMAP L4C NEE product further revealed that coarse-resolution datasets are strongly affected by mixed-pixel effects in heterogeneous rice paddy landscapes. In contrast, the proposed high-resolution approach more effectively captured fine-scale spatiotemporal variability in agricultural carbon fluxes. Overall, this study demonstrates that high-resolution, distribution-aware satellite-based modeling is essential for overcoming scale mismatch and mixed-pixel limitations in agricultural NEE estimation, providing a practical pathway toward im-proved carbon flux monitoring in managed agroecosystems.

Wildfire smoke visualization using geostationary satellite imagery is essential for real-time monitoring and atmospheric analysis; however, inconsistencies in color tone across Geostationary Environment Monitoring Spectrometer (GEMS) images hinder reliable interpretation and model training. This study proposes a Standardized False Color Composite (SFCC) framework based on deep learning style transfer to enhance the visual consistency and interpretability of wildfire smoke scenes. Four tone-standardization methods were compared: the statistical Empirical Cumulative Distribution Function (ECDF) correction and three neural approaches—ReHistoGAN, StyTr2, and Style Injection Diffusion Model (SI-DM). Each model was evaluated visually and quantitatively using six metrics (SSIM, LPIPS, FID, histogram similarity, ArtFID, and LSCI) and validated on three major wildfire events in Korea (2022–2025). Among the tested models, SI-DM achieved the most balanced performance, preserving structural features while ensuring consistent color-tone alignment (ArtFID = 1.620; LSCI mean = 0.894). Qualitative assessments further confirmed that SI-DM effectively delineated smoke boundaries and maintained natural background tones under complex atmospheric conditions. Additional analysis using GEMS UVAI, VISAI, and CHOCHO demonstrated that the styled composites partially reflect the optical and chemical characteristics distinguishing wildfire smoke from dust aerosols. The proposed SFCC framework establishes a foundation for visually standardized satellite smoke imagery and provides potential for future aerosol-type classification and automated detection applications.

Agriculture is a major global source of methane (CH4), and accurate emission estimates are essential for refining national greenhouse gas inventories and supporting climate-resilient policies. This study develops a high-resolution estimation framework for CH4 emissions from Korean rice paddies by integrating multi-source datasets, including Moderate Resolution Imaging Spectroradiometer (MODIS) vegetation indices, European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis Version 5 (ERA5)-Land meteorological variables, and Harmonized World Soil Database (HWSD) soil properties. Using CH4 flux observations from four global rice ecosystems (Italy, Japan, South Korea, and USA), we constructed parallel daily and hourly machine learning models using an automated machine learning (AutoML) framework to compare their performance and process-level interpretability. The daily model demonstrated high predictive accuracy with correlation coefficients (CC) of 0.897 in 5-fold cross-validation and 0.819 in Leave-One-Year-Out (LOYO) cross-validation. Shapley Additive Explanations (SHAP) analysis revealed that while soil temperature is the dominant predictor for daily emissions (explaining ~50% of the variance), variable importance shifts significantly at finer resolutions. The hourly model exhibited a more complex multivariate structure. In this high-resolution context, although Normalized Difference Vegetation Index (NDVI) remains constant diurnally, its importance strengthens as a critical regulator of emission sensitivity, interacting with hourly meteorological fluctuations to capture short-term dynamics. The resulting 500 m daily gridded maps provide a robust foundation for national inventory refinement and spatially targeted mitigation planning. Our findings suggest that while the daily model offers optimal computational efficiency for long-term monitoring, the hourly model is superior for mechanistic understanding and detecting episodic emission events. This multi-resolution framework establishes an empirical basis for selecting appropriate temporal scales in operational greenhouse gas monitoring systems.

Aerosol optical depth (AOD) is essential for air quality monitoring and climate research. However, satellite-based retrievals suffer from cloud-related data gaps, and reanalysis products are limited by coarse spatial resolution and substantial production latency. This study develops a real-time, gap-free, high-resolution (1.5 km) AOD retrieval system for South Korea. The system integrates Copernicus Atmosphere Monitoring Service (CAMS) forecasts, high-resolution meteorological fields, and ground-based air quality observations within a machine learning framework. Three models with varying training periods were systematically evaluated using cross-validation and independent validation with 2024 Aerosol Robotic Network (AERONET) data. The optimal model, trained on 2015-2023 data, achieved a mean absolute error (MAE) of 0.075 and a correlation coefficient (R) of 0.841 during the 2024 independent validation, significantly outperforming the original CAMS forecast. The system demonstrated robust and consistent performance across varying land cover types, seasons, and AOD conditions, from clean to highly polluted. Empirical orthogonal function (EOF) analysis confirmed that the product successfully captures physically meaningful spatiotemporal patterns, including transboundary pollution transport, regional emission gradients, and topographic effects. Providing real-time, gap-free, 3-hourly daytime AOD, the proposed model overcomes the limitations of cloud-induced gaps in satellite data and the latency and coarseness of reanalysis products. This enables robust operational monitoring and aerosol research across the Korean Peninsula.

Traditionally, sea fog detection technologies have relied primarily on in situ observations. However, point-based observations suffer from limitations in extensive monitoring in marine environments due to the scarcity of observation stations and the limited nature of measurement data. Satellites effectively address these issues by covering vast areas and operating across multiple spectral channels, enabling precise detection and monitoring of sea fog. Despite the increasing adoption of deep learning in this field, achieving further improvements in accuracy and reliability necessitates the simultaneous use of multiple satellite datasets rather than relying on a single source. Therefore, this study aims to achieve higher accuracy and reliability in sea fog detection by employing a deep learning-based advanced co-registration technique for multi-satellite image fusion and autotuning-based optimization of State-of-the-Art (SOTA) semantic segmentation models. We utilized data from the Advanced Meteorological Imager (AMI) sensor on the Geostationary Korea Multi-Purpose Satellite 2A (GK2A) and the GOCI-II sensor on the Geostationary Korea Multi-Purpose Satellite 2B (GK2B). Swin Transformer, Mask2Former, and SegNeXt all demonstrated balanced and excellent performance across overall metrics such as IoU and F1-score. Specifically, Swin Transformer achieved an IoU of 77.24 and an F1-score of 87.16. Notably, multi-satellite fusion significantly improved the Recall score compared to the single AMI product, increasing from 88.78 to 92.01, thereby effectively mitigating the omission of disaster information. Ultimately, comparisons with the officially operational GK2A AMI Fog and GK2B GOCI-II Marine Fog (MF) products revealed that our deep learning approach was superior to both existing operational products.

Offshore wind farm projects are being promoted in the seas surrounding the Korean Peninsula to secure renewable energy. To support site selection, offshore wind resource maps were generated using deep neural networks trained on Sentinel-1 SAR imagery, numerical weather prediction data, offshore wind observations, sea surface temperature, and bathymetry. The deep neural network (DNN) framework consisted of six sub-models targeting eastward and northward wind components across three regions—the Yellow Sea, Korea Strait, and East Sea—to account for spatial heterogeneity. The proposed models outperformed existing approaches, achieving mean absolute errors (MAE) ranging from 1.31 to 1.69 m/s and correlation coefficients (CC) between 0.827 and 0.913. These DNN models were then applied to produce offshore wind energy maps at a 150 m resolution, effectively capturing seasonal and regional variability. The resulting high-resolution maps provide valuable insights for evaluating the suitability of existing wind farm sites and identifying potential new candidates.

In California (CA), the wildland-urban interface (WUI) faces escalating challenges due to surging population and real estate development. This study evaluates communities along CA’s WUI that have witnessed substantial population growth from 2010 to 2021, utilizing demographic data and the 2020 WUI boundaries by the University of Wisconsin-Madison SILVIS Lab. Employing the Mann-Kendall test, we analyze yearly population trends for each census tract along the CA WUI and assess their significance. House ownership, affordability, and wildfire risk are examined as potential drivers of this demographic shift. Our findings indicate that 12.7% of CA’s total population now resides in census tracts with significant population increases over the past decade, labeled as ‘high-growth tracts.’ The Bay Area and Southern California, encompassing 76% of all high-growth tracts in CA, witnessed the most substantial population increase along the WUI. Notably, Riverside County stands out with 29.2% of its residents (approximately 717,000 residents) located in high-growth tracts, exemplifying a significant population surge within CA’s WUI. Our analysis identifies a significant relationship between population increase in the WUI, house ownership, and affordability, where lower-priced homes come at the expense of heightened wildfire risk. However, the impact of house affordability on population growth within the WUI varies by region, playing a more prominent role in explaining population proportions in Southern California’s WUI, while in the universally low-affordability Bay Area, other motivations may drive residents to live within the WUI. Given the rapid growth and insufficient consideration of wildfire risk in the WUI, policymakers must take prompt action, ensuring adequate infrastructure and resources as more individuals relocate to areas with heightened wildfire risk.

Soil moisture is a critical parameter that significantly impacts the global energy balance, including the hydrologic cycle, land–atmosphere interactions, soil evaporation, and plant growth. Currently, soil moisture is typically measured by installing sensors in the ground or through satellite remote sensing, with data retrieval facilitated by reanalysis models such as the European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis 5 (ERA5) and the Global Land Data Assimilation System (GLDAS). However, the suitability of these methods for capturing local-scale variabilities is insufficiently validated, particularly in regions like South Korea, where land surfaces are highly complex and heterogeneous. In contrast, artificial intelligence (AI) approaches have shown promising potential for soil moisture retrieval at the local scale but have rarely demonstrated substantial products for spatially continuous grids. This paper presents the retrieval of daily soil moisture (SM) over a 500 m grid for croplands in South Korea using random forest (RF) and automated machine learning (AutoML) models, leveraging satellite images and meteorological data. In a blind test conducted for the years 2013–2019, the AutoML-based SM model demonstrated optimal performance, achieving a root mean square error of 2.713% and a correlation coefficient of 0.940. Furthermore, the performance of the AutoML model remained consistent across all the years and months, as well as under extreme weather conditions, indicating its reliability and stability. Comparing the soil moisture data derived from our AutoML model with the reanalysis data from sources such as the European Space Agency Climate Change Initiative (ESA CCI), GLDAS, the Local Data Assimilation and Prediction System (LDAPS), and ERA5 for the South Korea region reveals that our AutoML model provides a much better representation. These experiments confirm the feasibility of AutoML-based SM retrieval, particularly for local agrometeorological applications in regions with heterogeneous land surfaces like South Korea.

Given the complex spatiotemporal variability of aerosols, high-frequency satellite observations are essential for accurately mapping their distribution. However, optical remote sensing encounters difficulties in detecting Aerosol Optical Depth (AOD) over cloud-covered regions, creating data gaps that limit comprehensive environmental analysis. This study introduces a spatial gap-filling method for Himawari-8/Advanced Himawari Imager (AHI) hourly AOD data, using a Random Forest (RF) model that integrates meteorological variables and model-based AOD data. Developed and validated over South Korea from 1 January to 31 December 2019, the model effectively improved data coverage from 6% to 100%. The approach demonstrated high performance in blind tests, achieving a root mean square error (RMSE) of 0.064 and a correlation coefficient (CC) of 0.966. Meteorological analysis indicated optimal model performance under cold, dry conditions (RMSE: 0.047, CC: 0.956), compared to humid conditions (RMSE: 0.105, CC: 0.921). Validation against Aerosol Robotic Network (AERONET) ground observations showed that, while the original Himawari-8 data exhibited higher accuracy (RMSE: 0.189, CC: 0.815, n = 346), the gap-filled dataset maintained reasonable precision (RMSE: 0.208, CC: 0.711) and significantly increased the number of valid data points (n = 4149). Furthermore, the gap-filled dataset successfully captured seasonal AOD patterns, with values ranging from 0.245–0.300 in winter to 0.381–0.391 in summer, providing a comprehensive view of aerosol dynamics across South Korea.

The recent decades have seen an increasing academic interest in leveraging machine learning approaches to nowcast, or forecast in a highly short-term manner, precipitation at a high resolution, given the limitations of the traditional numerical weather prediction models on this task. To capture the spatiotemporal associations of data on input variables, a deep learning (DL) architecture with the combination of a convolutional neural network and a recurrent neural network can be an ideal design for nowcasting rainfall. In this study, a long short-term memory (LSTM) modeling structure is proposed with convolutional operations on input variables. To resolve the issue of underestimation of heavy rainfall that challenges most of the DL models, a pixelwise modeling approach is adopted to facilitate a stratified sampling process in generating training data points for calibrating models to predict rain rates at locations. The proposed pixelwise convolutional LSTM (CLSTM) models are applied to data on mesoscale convective systems during the warm seasons over the Korean Peninsula. Results show a significant and consistent improvement in prediction skill scores produced by the CLSTM models than a traditional rainfall nowcasting method, the McGill algorithm for precipitation nowcasting by Lagrangian extrapolation, across all considered lead times from 10 to 60 min. Future work needs to reduce the relatively large false positive rates produced by the CLSTM models and their blurring effect in mapping spatial distributions of rain rates, in particular for longer lead times.

The diagonal squall line that passed through the Korean Peninsula on the 18 May 2020 was examined using wind data retrieved from multiple Doppler radar synthesis focusing on its kinematic and dynamic aspects. The low-level jet, along with warm and moist air in the lower level, served as the primary source of moisture supply during the initiation and formation process. The presence of a cold pool accompanying the squall line played a role in retaining moisture at the surface. As the squall line approached the Korean Peninsula, the convective bands in the northern segment (NS) and southern segment (SS) of the squall line exhibited distinct evolutionary patterns. The vertical wind shear in the NS area was more pronounced compared to that in the SS. The ascending inflow associated with the tilted updraft in the NS reached an altitude of 7 km, whereas it was only up to 4 km in the SS. The difference was caused by the strong descending rear flow, which obstructed the ascending inflow and let to significant updraft in the SS.

SummaryWe examine the distributions of lightning strikes and wildfires across the Western United States (WUS).Lightning data from the Geostationary Lightning Mapper (GLM) are consistent with the ones from the National Lightning Detection Network (NLDN).We use 10-km spatial and 7-day temporal maximum distances to associate lightnings with wildfires.Only wildfires identified by the United States Forest Service (USFS) with a natural cause are considered for modeling.Considered factors of fire lightning include location, time, precipitation, and lightning area and energy.Location and time variables are significantly more important than the other factors.Albeit a good predictor, precipitation is usually nonzero along with a fire lightning.Data of years 2019–2021 are used for model training and data of 2022 for validation.Validation results show decent predictive performance regarding identification of fire lightning.

The nature of superconductivity in newly discovered kagome materials, AV(3)Sb(5) (A = K, Rb, Cs), has been a subject of intense debate. Recent experiments suggest the presence of orbital current order on top of the charge density wave (CDW) and superconductivity. Since the orbital current order breaks time-reversal symmetry, it may fundamentally af-fect possible superconducting states. In this work, we investigate the mutual influence between the orbital current order and superconductivity in kagome metal with characteristic van Hove singularity (vHS). By explicitly deriving the Landau-Ginzburg theory, we classify possible orbital current order and superconductivity. It turns out that distinct unconventional superconductivities are expected, depending on the orbital current ordering types. Thus, this information can be used to infer the superconducting order parameter when the orbital current order is identified and vice versa. We also discuss possible experiments that may distinguish such superconducting states coexisting with the orbital current order.

With the emergence of large amounts of historical records on adverse impacts of hazardous events, empirical predictive modeling has been revived as a foundational paradigm for quantifying disaster vulnerability of societal systems. This paradigm models societal vulnerability to hazardous events as a vulnerability curve indicating an expected loss rate of a societal system with respect to a possible spectrum of intensity measure (IM) of an event. Although the empirical predictive models (EPMs) of societal vulnerability are calibrated on historical data, they should not be experimentally tested with data derived from field experiments on any societal system. Alternatively, in this paper, we propose a Monte Carlo simulation-based approach to experimentally test EPMs of societal vulnerability. Our study applied an eigenvalue-based method to generate data on societal experiences of IM and pre-event vulnerability indicators. True models were designed to simulate event loss data. Supervised machine learning (ML) models were then trained on simulated data and were found to provide similar predictive performances as the true models. Our results suggested that the calibrated ML-EPMs could effectively quantify societal vulnerability given a normally experienced IM. To extrapolate a vulnerability curve for large IMs, however, simple models should be preferred.

Several studies have attempted to estimate particulate matter (PM) concentrations using aerosol optical depth (AOD), based on AOD and PM relationships. Owing to the limited availability of nighttime AOD data, PM estimation studies using AOD have focused on daytime. Recently, the Aerosol Robotic Network (AERONET) produced nighttime AOD, called lunar AOD, providing an opportunity to estimate nighttime PM. Nighttime AOD measurements are particularly important as they help fill gaps in our understanding of aerosol variability and its impact on the atmosphere, as there are significant variations in AOD between day and night. In this study, the relationship between lunar AOD and PM was investigated using data from AERONET station, meteorological station, and air pollution station in Seoul Metropolitan area from May 2016 to December 2019, and then PM estimation model was developed covering both daytime and nighttime using random forest machine learning techniques. We have found the differences in the importance of variables affecting the AOD-PM relationship between day and night from the random forest model. The AOD-PM relationship in the daytime was more affected by time-related variables, such as the day of the year among the variables. The new model was developed using additional lunar AOD data to estimate continuous PM concentrations. The results have shown that the model based on lunar AOD data estimated well PM 10 and PM 2.5 with similar performance of model using solar AOD. The results imply the possibility of seamless near-surface PM concentration data on a large scale once satellites produce nighttime AOD data.

Highly short-term forecasting, or nowcasting, of heavy rainfall due to rapidly evolving mesoscale convective systems (MCSs) is particularly challenging for traditional numerical weather prediction (NWP) models. To overcome such a challenge, a growing number of studies have shown significant advantages of using machine learning (ML) modeling techniques with remote sensing data, especially weather radar data, for high-resolution rainfall nowcasting. To improve ML model performance, it is essential first and foremost to quantify the importance of radar variables and identify pertinent predictors of rainfall that can also be associated with domain knowledge. In this study, a set of MCS types consisting of convective cell (CC), mesoscale CC, diagonal squall line (SLD), and parallel squall line (SLP), was adopted to categorize MCS storm cells, following the fuzzy logic algorithm for storm tracking (FAST), over the Korean Peninsula. The relationships between rain rates and over 15 variables derived from data products of dual-polarimetric weather radar were investigated and quantified via five ML regression methods and a permutation importance algorithm. As an applicational example, ML classification models were also developed to predict locations of storm cells. Recalibrated ML regression models with identified pertinent predictors were coupled with the ML classification models to provide early warnings of heavy rainfall. Results imply that future work needs to consider MCS type information to improve ML modeling for nowcasting and early warning of heavy rainfall.