베이지안 혼합 모델을 이용한 Sentinel-1 GRD 데이터 기반 수체검출 및 저수지 수위 예측 (Waterbody Detection and Reservoir Water Level Prediction Using Bayesian Mixture Models with Sentinel-1 GRD Data)

서지정보

· 발행기관 : (사)지오에이아이데이터학회
· 수록지 정보 : GEO DATA / 7권 / 1호 / 18 ~ 26페이지
· 저자명 : 윤동현, 유하은, 황의호, 강기묵, 남기범, 김진겸

초록

In this study, we used a Bayesian mixture model (BMM) to monitor water surfaceareas and estimate water levels in Yeongcheon Dam through Sentinel-1 syntheticaperture radar (SAR) imagery. Reservoirs serve vital functions such as flood control,drought mitigation, and ecosystem support, highlighting the importance of precisemonitoring of their water surface and level variations, especially in the context ofclimate change and increased human impact. The BMM method was employed toaccurately delineate water boundaries, benefiting from SAR’s capability to capturedata regardless of weather conditions. Regression analysis was conducted betweenthe extracted water surface area and observed water levels to create a predictivemodel, 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 surfacearea data. One of the key findings of this study is that even with a 10 m spatialresolution, reliable water level inferences can be made using water surface area asa proxy. The mean absolute error values obtained validate the model’s capabilityto monitor water level fluctuations with a satisfactory degree of accuracy. Despitelimitations in detecting narrow tributaries or other small-scale features due to SARresolution, the model performs well overall in monitoring broad water bodies. Thesefindings underscore the potential of Sentinel-1 SAR data for effective reservoirmonitoring, especially where real-time water level data may be lacking. For futureresearch, higher-resolution data or complementary algorithms may further enhancedetection accuracy for smaller and more complex water features, contributing tomore refined water resource management strategies.

영어초록

In this study, we used a Bayesian mixture model (BMM) to monitor water surfaceareas and estimate water levels in Yeongcheon Dam through Sentinel-1 syntheticaperture radar (SAR) imagery. Reservoirs serve vital functions such as flood control,drought mitigation, and ecosystem support, highlighting the importance of precisemonitoring of their water surface and level variations, especially in the context ofclimate change and increased human impact. The BMM method was employed toaccurately delineate water boundaries, benefiting from SAR’s capability to capturedata regardless of weather conditions. Regression analysis was conducted betweenthe extracted water surface area and observed water levels to create a predictivemodel, 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 surfacearea data. One of the key findings of this study is that even with a 10 m spatialresolution, reliable water level inferences can be made using water surface area asa proxy. The mean absolute error values obtained validate the model’s capabilityto monitor water level fluctuations with a satisfactory degree of accuracy. Despitelimitations in detecting narrow tributaries or other small-scale features due to SARresolution, the model performs well overall in monitoring broad water bodies. Thesefindings underscore the potential of Sentinel-1 SAR data for effective reservoirmonitoring, especially where real-time water level data may be lacking. For futureresearch, higher-resolution data or complementary algorithms may further enhancedetection accuracy for smaller and more complex water features, contributing tomore refined water resource management strategies.

참고자료

· 없음