Forecasting Snowmelt-Induced Flooding Using GRACE Satellite Data: A Case Study for the Red River Watershed

. Flood forecasting of the spring freshet for cold-region watersheds where the discharge is predominately governed by snowpack accumulation and melting remains a challenge. A cold-region flood forecasting model is developed, using data from the Gravity Recovery and Climate Experiment (GRACE) satellite mission. The model forecasts flood by simulating peak surface runoff from snowmelt and the corresponding baseflow from groundwater discharge. Surface runoff is predicted from snowmelt, using a temperature index model. Baseflow is predicted, using a first order differential equation model. Streamflow measurement is used for model calibration. The model was applied to the Red River watershed, a USA–Canada transboundary basin located in central North America. The predicted flood compares well with the observed values at a downstream hydrometric station (r = 0.95). The result also reveals a 2-week hysteresis between the maximum snowmelt and the peak streamflow observed at the station. The model is relatively simple and needs only GRACE and temperature inputs for flood forecasting. It can be readily applied to other cold-region basins after simple calibration and could be particularly useful in regions with minimal data. For potential flood warning, the model also has the advantage of a much longer lead time than most traditional flood forecasting approaches.