Global high-resolution total water storage anomalies from self-supervised data assimilation using deep learning algorithms

Abstract

Total water storage anomalies (TWSAs) describe the variations of the terrestrial water cycle, which is essential for better understanding our climate system. Global TWSAs can be simulated by hydrological models with high spatial resolution but limited accuracy or measured by the Gravity Recovery and Climate Experiment (GRACE) satellite mission with opposite characteristics. In this study, we designed a self-supervised data assimilation algorithm with a novel loss function, which combines the advantages of different TWSA sources. The determined global TWSAs have the same spatial resolution as the hydrological simulations (0.5°) and preserve the large-scale accuracy inherited from the GRACE measurements. They contribute to monitoring natural hazards locally and show potential for better understanding the impacts of natural and anthropogenic activities on the water cycle. We anticipate our approach to be generally applicable to other TWSA data sources and the resulting products to be valuable for the geoscience community and society.

Global high-resolution total water storage anomalies from self-supervised data assimilation using deep learning algorithms

“…The downscaled TWSA product generated in this study is publicly available at https://doi.org/10.3929/ethz-b-000648738 (ref. 77). The raw data used in this study are publicly available.…”