Abstract：Rainfall-runoff (RR) modelling is a challenging task in hydrology, especially at the regional scale. This work presents an approach to simultaneously predict daily streamflow in 86 catchments across the US using a sequential CNN-LSTM deep learning architecture. The model effectively incorporates both spatial and temporal information, leveraging the CNN to encode spatial patterns and the LSTM to learn their temporal relations. For training, a year-long spatially distributed input with precipitation, maximum temperature, and minimum temperature for each day was used to predict one-day streamflow. The trained CNN-LSTM model was further fine-tuned for three local sub-clusters of the 86 stations, assessing the significance of fine-tuning in model performance. The CNN-LSTM model, post fine-tuning, exhibited strong predictive capabilities with a median Nash-Sutcliffe efficiency (NSE) of 0.62 over the test period. Remarkably, 65% of the 86 stations achieved NSE values greater than 0.6. The performance of the model was also compared to different deep learning models trained using a similar setup (CNN, LSTM, ANN). An LSTM model was also developed and trained individually to predict for each of the stations using local data. The CNN-LSTM model outperformed all the models which was trained regionally, and achieved a comparable performance to the local LSTM model. Fine-tuning improved the performance of all models during the test period. The results highlight the potential of the CNN-LSTM approach for regional RR modelling by effectively capturing complex spatiotemporal patterns inherent in the RR process.
KeyWord：Regional modelling; Deep learning; CNN; LSTM; CAMELS; Rainfall-runoff;

相关文献：
1.The responsiveness of renewable energy production to geopolitical risks, oil market instability and economic policy uncertainty: Evidence from United States
2.Evaluating the employment effects of environmental regulation without abatement cost data: A nonparametric cost function approach
3.Urban climate adaptability and green total-factor productivity: Evidence from double dual machine learning and differences-in-differences techniques
4.Abating ammonia emission from poultry manure by Pt/TiO2 modified corn straw
5.Editorial Board
6.To what extent can decommissioning options for marine artificial structures move us toward environmental targets?
7.Impact of political conflict on foreign direct investments in the mining sector: Evidence from the event study and spatial estimation
8.Adsorption on activated carbon combined with ozonation for the removal of contaminants of emerging concern in drinking water
9.Carbon conundrums: Geopolitical clashes and market mayhem in the race for sustainability
10.Technology-enhanced community forest management in tropical regions: A state of the art