BiasCast: learning and adjusting real time biases from meteorological forecasts to enhance runoff predictions

Abstract

Abstract. The use of deep learning models in hydrology is becoming an ever more prevalent application in operational flood forecasting. Such operational systems face performance degradation when transitioning from high quality reanalysis to meteorological forecast data with lower accuracy. This study investigates training strategies and Long Short-Term Memory network architectures to mitigate meteorological forecast-induced bias in maximum daily discharge predictions using the Extended LamaH- CE dataset and a subset of 451 basins. We systematically evaluated cross-domain generalization, transfer learning approaches, Encoder–Decoder LSTMs, Sequential Forecast LSTMs, and the role of input embeddings and integrating past discharge observations. The results show that domain shifts between reanalysis and forecast data lead to substantial skill loss, with median Nash–Sutcliffe Efficiency decreasing from 0.58 to 0.33. Among the tested strategies, the Sequential Forecast LSTM demonstrated the most stable improvements, achieving a median NSE of 0.63. Integrating recent discharge observations further enhanced performance, raising median NSE to 0.71 and surpassing even the reanalysis-driven baseline. In contrast, integrating archived forecasts or using more complex input embeddings did not yield consistent benefits and in some cases degraded model stability. Basin-level analysis reveals that forecast skill improvements compared to our baseline are not uniformly distributed across catchment types: the largest gains are concentrated in arid and precipitation-limited catchments, while alpine and snow-dominated catchments, despite experiencing the largest meteorological domain shift, show smaller improvements. This is likely because the LSTM cell state retains strong seasonal signals and thereby compensates for forecast input bias through its long-term memory mechanism, particularly in alpine and snow-dominated catchments where strong seasonal cycles dominate the hydrological response. These findings highlight the value of training strategies that allow models to directly learn bias correction during forecast transitions, emphasize the operational potential of combining sequential processing with near real-time discharge observations and identify physiographic catchment characteristics as key modulators of forecast skill improvement across diverse hydroclimatic settings.