Integrating weather and climate prediction: Toward seamless hydrologic forecasting

Abstract

AbstractSeamless hydrologic forecasting is explored through integrating medium‐range weather forecasts from NOAA's Global Ensemble Forecast System Reforecast version 2 (GEFSRv2) and seasonal climate predictions from the Climate Forecast System version 2 (CFSv2). A set of 25 year hydrologic reforecasts over the Ohio basin shows that incorporating GEFSRv2 14 day forecasts into the Ensemble Streamflow Prediction (ESP) and CFSv2‐based seasonal forecast systems improves efficiency scores for month‐1 streamflow by up to 32.6% and 11.2%, respectively. For the second biweekly forecast, the combination of GEFSRv2 and CFSv2 is superior to that of GEFSRv2 and ESP by increasing efficiency score up to 17.2%, suggesting that the climate prediction usefully extends the medium‐range hydrologic forecast. As compared with ESP, incorporation of either weather or climate prediction improves the month‐1 soil moisture drought prediction significantly. The potential of seamless hydrologic forecast should be further investigated from the operational service perspective and improved understanding of underlying physical processes.