[A-HW23] 水循環・水環境

[AHW23-10] Investigation of Point Level Data Between Remote Sensing Precipitation Data with Ground Precipitation Data to Support Water Resource Management Application in Humid Climate

*Aliya Binti Mhd Zahir -¹、Hiroaki SOMURA¹ (1.Okayama University )

Understanding the hydrological environment through long-term observation is crucial for water resource management studies. Nowadays, with the open-source earth observation platforms from satellites, it provides an opportunity to improve data acquisition of meteorological data such as precipitation. Acquiring precipitation data from in-situ measurement sometimes comes with challenges such as inadequate locations of the stations or the rain gauges too far from the research area that will contribute to inaccurate data. Therefore, this study aims to examine the applicability of earth observation precipitation data from the remote sensing for local water resources management in humid climates by comparing the ground measurement precipitation data at the same period. The data acquisition of remotely sensed precipitation was performed in Google Earth Engine (GEE) cloud computing platform. Two years (2019-2020) Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS), The Global Precipitation Measurement (GPM), and Precipitation Estimation from Remotely Sensed Information Using Artificial Neural Networks (PERSIANN) datasets were specified at the same coordinate of the rainfall station. Results show that CHIRPS and PERSIANN datasets perform poorly at daily time steps while daily scale dataset were not available for GPM. In the monthly temporal resolution, remote sensing datasets exhibit well to moderately error with in-situ measurement except for a significant large discrepancy for December 2019 to January 2020 period. Despite a variable pattern of error on a monthly scale, a good correlation can be seen in annual time steps in both years. In 2019, R² values are 0.80, 0.86, 0.68, while in 2020, R² values are 0.69, 0.84, 0.70 respectively for CHIRPS, GPM and PERSIANN. From this finding, the GPM dataset shows the best performance compared to other remote sensing datasets with the corresponding RMSE 50.45mm/month in 2019 and 61.42mm/months in 2020. This study will improve water resource management application, especially in poorly gauged areas.