Uncertainty Estimation with Deep Learning for Rainfall-Runoff Modelling

Deep Learning is becoming an increasingly important way to produce accurate hydrological predictions across awide range of spatial and temporal scales. Uncertainty estimations are critical for actionable hydrological forecasting, andwhile standardized community benchmarks are becoming an increasingly important part of hydrological model developmentand research, similar tools for benchmarking uncertainty estimation are lacking. We establish an uncertainty estimation bench-5marking procedure and present four Deep Learning baselines, out of which three are based on Mixture Density Networks andone is based on Monte Carlo dropout. Additionally, we provide a post-hoc model analysis to put forward some qualitativeunderstanding of the resulting models. Most importantly however, we show that accurate, precise, and reliable uncertaintyestimation can be achieved with Deep Learning.