Deep learning for monthly rainfall-runoff modelling: a comparison with classical rainfall-runoff modelling across Australia

Abstract. A deep learning model designed for time series predictions, the long short-term memory (LSTM) architecture is regularly producing reliable results in local and regional rainfall-runoff applications around the world. Recent large-sample-hydrology studies in North America and Europe have shown the LSTM to successfully match conceptual model performance at a daily timestep over hundreds of catchments. Here we investigate how these models perform in producing monthly runoff predictions in the relatively dry and variable conditions of the Australian continent. The monthly timestep matches historic data availability and is also important for future water resources planning, however it provides significantly smaller training data sets than daily time series. In this study, a continental-scale comparison of monthly deep learning (LSTM) predictions to conceptual rainfall-runoff model (WAPABA) predictions is performed on almost 500 catchments across Australia with performance results aggregated over a variety of catchment sizes, flow conditions, and hydrological record lengths. The study period covers a wet phase followed by a prolonged drought, introducing challenges for making predictions outside of known conditions - challenges that will intensify as climate change progresses. The results show that LSTMs matched or exceeded WAPABA prediction performance for more than two-thirds of the study catchments; the largest performance gains of LSTM versus WAPABA occurred in large catchments; the LSTM models struggled less to generalise than the WAPABA models (eg. making predictions under new conditions); and catchments with few training observations due to the monthly timestep did not demonstrate a clear benefit with either WAPABA or LSTM.