Machine Learning for Postprocessing Medium-range Ensemble Streamflow Forecasts

Skillful streamflow forecasts can inform decisions in various areas of water policy and management. We integrate numerical weather prediction ensembles and a distributed hydrological model to generate ensemble streamflow forecasts at medium-range lead times (1 - 7 days). We demonstrate a case study for machine learning application in postprocessing ensemble streamflow forecasts in the Upper Susquehanna River basin in the eastern United States. For forecast verification, we use different metrics such as skill score and reliability diagram conditioned upon the lead time, flow threshold, and season. The verification results show that the machine learning postprocessor can improve streamflow forecasts relative to low complexity forecasts (e.g., climatological and temporal persistence) as well as deterministic and raw ensemble forecasts. As compared to the raw ensembles, relative gain in forecast skill from postprocessor is generally higher at medium-range timescales compared to shorter lead times; high flows compared to low-moderate flows, and warm-season compared to the cool ones. Overall, our results highlight the benefits of machine learning in many aspects for improving both the skill and reliability of streamflow forecasts.