Assessing the Impact of Radar-Rainfall Uncertainty in Streamflow Prediction

This study investigates the challenges in understanding the contribution of radar rainfall to flood forecasting uncertainty. Previous research focused on different hydrological aspects separately, ignoring the relevance of existing errors in rainfall estimation. We introduced a systematic approach using a hydrological model in Iowa forced with three radar rainfall products. Results show that accurately predicting rainfall is crucial for flood forecasting, with variations based on location, rainfall amount, and watershed size. Seasonal trends also indicate underestimations during the cold season. Our results emphasize the importance of considering radar and rainfall estimation methods to improve flood forecasting accuracy.

You can read the full preprint here