Code for WRR paper "A novel strategy for automatic selection of training data to reduce local performance loss in regional models inspired by Multi-Task learning".
The code is used to model rainfall-runoff behaviors of the target basin (ID=07208500)