Water management decisions are highly dependent on forecast information regarding the possible future evolution of precipitation. Quantitative Precipitation Forecasts (QPF) from the current generation of numerical weather prediction models such as the Global Forecast System (GFS) at the National Centers for Environmental Prediction (NCEP) are biased due to model deficiencies. Probabilistic QPF (PQPF) forecasts based on the global ensemble forecasts at NCEP are biased as well due to imperfections in model and ensemble formation. By calibrating each member of the ensemble based on verification statistics accumulated over the continental US, the bias in QPF forecasts ( first moment ) is practically eliminated, and the bias in PQPF forecasts (second moment) is substantially reduced. This study will facus on two problems: 1) Development of a new statistical calibration method to adjust the second moment of PQPF forecasts in order to reduce or eliminate the second moment bias while keeping the already bias-free first moment (mean) unchanged;. 2). Optimization of the temporal and spatial domains over which verificaiton statistics for the calibration method are accumulated. Unlike the raw ensemble, the bias corrected forecasts, and/or PQPF information derived from them will be ready for use in sophisticated probabilistic hydrological models as input describing the expected future evolution of precipitation on the scale of the NWP models.