Accounting for Uncertainty in Short Term Deterministic Precipitation Forecasts

National Weather Service Advanced Hydrologic Prediction Services (AHPS) will improve existing services by extending lead times, improving accuracy and accounting for uncertainty. Although new ensemble-based methods to account for uncertainty in short term quantitative precipitation forecasts (QPF's) are being developed, methods to account for uncertainty in existing short term deterministic QPF's are needed for immediate application in AHPS. Existing ensemble streamflow prediction (ESP) software being used for AHPS includes a technique to blend historical observed precipitation data with deterministic QPF's to create short term precipitation ensemble forecasts. The existing ESP technique includes parameters that depend on the inherent uncertainty in the deterministic QPF. The objective of this study is to provide guidance for estimating these parameters and to assess the strengths and limitations of the blending procedure.