31 December 2024 in GEO DATA
Due to the recent anomalous climatic conditions, the mean summer temperature in Korea for the current year reached 25.6°C, marking the highest temperature recorded since 1973. Concurrently, the incidence of large-scale green algal blooms has escalated beyond the typical frequency in the four principal river systems. The Ministry of Environment has instituted a water quality monitoring network aimed at gathering data pertinent to water quality assessment and regulatory measures; however, the reliance on discrete point data imposes constraints on comprehending the spatial distribution and proliferation of green algae. Techniques for remote monitoring of green algae utilizing satellite imagery may serve as an alternative or complementary resource to conventional point-based water quality assessments. In the present study, we devised and critically assessed an algorithm for the remote estimation of freshwater chlorophyll-a concentrations through the analysis of Sentinel-2 satellite data. Given the challenges associated with field data collection, the GLORIA dataset was employed, and three variations of the CatBoost algorithm were developed based on different input variables, with their effectiveness concerning Sentinel-2 satellite data subjected to evaluation. Among the three algorithms, the variant that was trained utilizing variables such as band ratio and spectral shape was contrasted with 76 field measurements of surface chlorophyll-a, demonstrating a commendable accuracy with an R value of 0.79, root mean square error of 10.1 mg/m3 when applied to the GLORIA dataset. The chlorophyll-a quantification algorithm predicated on Sentinel-2 satellite data, which was developed in this study, holds the potential to be utilized effectively for collaborative management and responsive measures regarding green algae by providing objective insights into the prevalence of green algae across extensive areas, temporal trends, and localized hotspots.
31 March 2025 in GEO DATA
In this study, we used a Bayesian mixture model (BMM) to monitor water surface areas and estimate water levels in Yeongcheon Dam through Sentinel-1 synthetic aperture radar (SAR) imagery. Reservoirs serve vital functions such as flood control, drought mitigation, and ecosystem support, highlighting the importance of precise monitoring of their water surface and level variations, especially in the context of climate change and increased human impact. The BMM method was employed to accurately delineate water boundaries, benefiting from SAR’s capability to capture data regardless of weather conditions. Regression analysis was conducted between the extracted water surface area and observed water levels to create a predictive model, yielding a highly accurate equation with an R2 core of 0.981 on the test set. This result indicates a strong correlation between water surface area and water level, affirming the model’s reliability in estimating water levels based solely on surface area data. One of the key findings of this study is that even with a 10 m spatial resolution, reliable water level inferences can be made using water surface area as a proxy. The mean absolute error values obtained validate the model’s capability to monitor water level fluctuations with a satisfactory degree of accuracy. Despite limitations in detecting narrow tributaries or other small-scale features due to SAR resolution, the model performs well overall in monitoring broad water bodies. These findings underscore the potential of Sentinel-1 SAR data for effective reservoir monitoring, especially where real-time water level data may be lacking. For future research, higher-resolution data or complementary algorithms may further enhance detection accuracy for smaller and more complex water features, contributing to more refined water resource management strategies.
31 December 2025 in GEO DATA
Satellite-derived digital elevation models (DEMs) are indispensable for hydrologic analysis, yet riverbed topography beneath water bodies is systematically missing because of water reflectance/dispersion and water-level variability between satellite revisits. To address this structural limitation, we propose an automatic dataset construction method for DEM inpainting. For each DEM patch, the empirical cumulative distribution function of elevations is compared with a linear baseline; the threshold is chosen at the knee point where their vertical deviation is maximized. Elevations at or below this threshold are masked as likely water-occupied terrain, while the original DEM serves as ground truth. Using the Copernicus DEM GLO-30, we sample patches from 30°S to 30°N, enforce a water-area ratio of 5-15%, and apply morphological post-processing. A patch size of 64×64 is adopted, yielding 10,000 input-ground-truth pairs in GeoTIFF format (EPSG:4326). Dataset validity is assessed with a lightweight multi-scale U-Net, evaluation is restricted to the masked (missing) regions. On the test set, the model achieves PSNR 13.44 dB, SSIM 0.78, and MAE 2.80 m. Qualitative analyses show stable reconstruction of topographic continuity near shorelines, with most vertical errors confined to within ±5 m around boundaries. The proposed approach enables standardized, automated generation of large training datasets that reflect local terrain characteristics without field-grade ground truth, providing a practical basis for hydrologically complete DEMs for riverbed restoration, water-level estimation, and large-scale hydrologic simulation. Future work will extend to mid- and high-latitude regions and incorporate hydrological-consistency constraints to jointly improve structural similarity and absolute accuracy.
31 December 2025 in GEO DATA
The demand for high-resolution satellite imagery is rapidly increasing in various fields such as national defense, disaster response, and environmental monitoring, drawing significant attention to super-resolution (SR) techniques in remote sensing. In this study, a high-quality high resolution (HR)-low resolution (LR) training dataset was generated using a single sub-meter satellite image acquired from WorldView-3. The enhanced correlation coefficient image registration method was employed to correct geometric distortions between the panchromatic and multispectral images, and the modulation transfer function-generalized Laplacian pyramid algorithm was applied to produce HR images with improved spatial resolution while preserving spectral fidelity. LR images were synthesized through sensor-specific degradation modeling, resulting in a total of 14,776 HR-LR patch pairs. The constructed dataset covers diverse land cover types, including urban areas, agricultural fields, and water bodies, enabling comprehensive evaluation of super-resolution models in both spatial detail reconstruction and spectral information preservation. Unlike conventional datasets that rely on cross-sensor fusion with complex preprocessing, this study simplifies the data generation pipeline using a single-sensor approach and ensures spatial-spectral consistency. The proposed dataset is expected to serve as a reliable benchmark for training and evaluating deep learning-based satellite image SR models.