Development of a Regional Gridded Runoff Dataset Using Long Short-Term Memory (LSTM) Networks

Gridded datasets provide spatially and temporally consistent runoff estimates that serve as reliable sources for assessing water resources from regional to global scales. This study presents LSTM-REG, a gridded dataset northwest Russia based on Long Short-Term Memory (LSTM) networks. LSTM-REG covers the period 1980 2016 at 0.5° spatial daily temporal resolution. has been extensively validated benchmarked against GR4J-REG, parsimonious regionalization scheme GR4J hydrological model. While both with prediction efficiency, outperforms GR4J-REG most basins in independent evaluation set. Thus, results demonstrate higher generalization capacity of than which can be attributed efficiency proposed LSTM-based scheme. The developed are freely available open repositories foster further hydrology research Russia.