Raw precipitation forecasts from numerical weather prediction models often exhibit systematic biases and spatial inconsistencies that limit their direct use in hydrological forecasting applications.