Integrating weather and climate prediction: Toward seamless hydrologic forecasting

Seamless 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.