Dong-He Ma

A Hybrid Bi‐ LSTM ‐ AM Framework for Improving Extreme Rainfall Prediction and Bias Correction

Reliable forecasting of extreme rainfall remains a critical challenge for flood management in climate‐sensitive regions. This study develops and evaluates an attention‐augmented, bidirectional long short‐term memory framework (Bi‐LSTM‐AM) used for gauge‐based prediction and raster‐based bias correction for extreme ra...