Combining machine learning algorithms for bridging gaps in GRACE and GRACE Follow-On missions using ERA5-Land reanalysis

This study addresses the gap between GRACE and GRACE Follow-On (GFO) missions in monitoring terrestrial water storage anomalies (TWSA). By employing a combination of machine learning models (Random Forest, Support Vector Machine, eXtreme Gradient Boosting, Deep Neural Network, and Stacked Long-Short Term Memory), the research effectively bridges this gap and reconstructs global TWSA at a 0.5° grid resolution. The models were trained using six hydroclimatic variables and evaluated based on performance metrics (Nash-Sutcliffe Efficiency, Pearson's Correlation Coefficient, and Root Mean Square Error). Results show superior accuracy, outperforming previous methods, and demonstrating the model's potential for filling data gaps globally, with applications in flood/drought events and sea-level rise predictions.