Model Simulation Outputs, Evaluation Metrics, and Figure Source Data for Regionalized AI-Based Runoff Simulation in Humid Southern China

This dataset contains model simulation outputs, catchment-level evaluation metrics, evapotranspiration simulation results, source data underlying the figures and tables, and postprocessing scripts associated with a regionalized artificial intelligence–based runoff simulation study across 298 catchments in humid southern China. The experiments compare multiple precipitation products, static catchment representations, and runoff modeling paradigms, including LSTM, DPLHBV, DPLXAJ, and DRUM. Original third-party model implementations are not redistributed and are available from their respective repositories. Observed daily streamflow records are not included because redistribution is restricted by the data provider.