Probabilistic Forecasts of Precipitation Type

Precipitation-type forecasting is the determination of when and where particular types of precipitation (e.g., snow, rain, ice pellets, freezing rain) will occur during a forecast period. Although much is already known about the physical processes that determine the type of precipitation that reaches the ground, these forecasts are very challenging for most forecasters because of inadequate atmospheric data sampling and limited access to high resolution model data. In this study, we examine the quality of six precipitation-type algorithms using Eta and RUC model data. We also analyze the quality of the probabilistic forecasts that were created from a combination of the algorithm outputs. Since a early examination of the algorithms using rawinsonde data showed that there was not one algorithm that accurately diagnosed the correct precipitation type for all types of precipitation, we combined the algorithms to provide a measure of forecast uncertainty. Data used in this study was created during the Precipitation-type Algorithm Experiment (PTAX), which occurred during the winter of 2000–2001 and involved meteorologists at the University of Oklahoma, the NOAA/Hydrometeorological Prediction Center, and the NOAA/Storm Prediction Center.