In their recent publication in Nature Water, D-BAUG researchers Junyang Gou and Professor Benedikt Soja introduced a finely resolved model of terrestrial water storage using a novel deep learning approach. By integrating satellite observations with hydrological models, their method achieves remarkable accuracy even in smaller basins.
This model promises significant benefits across various domains, including hydrology, climate science, sustainable water management, and hazard prediction.
"We developed a self-supervised learning model to downscale GRACE(-FO) measurements and provide a global product with an effective spatial resolution of about 50 km. The specifically designed loss function allows the model to be optimized in the absence of high-resolution ground truth. The quality of the downscaled product has been proven by investigating different aspects, including closing the water balance equation in basins smaller the effective resolution of GRACE(-FO) missions and validating against altimetry-measured water levels. We also discuss the potential downstream applications, including monitoring environmental extremes at the local scale. The downscaled product should be beneficial for the geoscience community and society, especially in the fields of hydrology, climate science, sustainable water management and hazard prediction."
