Evaluation of Precipitation Type from Observational Datasets to Improve Monitoring of Hazardous Winter Weather at the Storm Prediction Center

The Storm Prediction Center (SPC) issues mesoscale discussions (MD) for hazardous, short-term winter weather events affecting the continental United States.  Since the precipitation type and intensity often vary greatly over a short duration and distance for most high-impact episodes of frozen precipitation, the forecast can often be complex and challenging.  As a result, the development of a winter weather MD often requires in-depth examination of a diverse collection of data sets (e.g., model guidance, radar, satellite, and observations).   Consequently, tools and products which summarize information provide considerable value for the SPC forecaster.  For this purpose, a simple objectively gridded, precipitation type system was developed recently and has shown some promise in internal testing. 

 The current work explores the value of using multiple observational datasets in the diagnosis of a dominant precipitation type for a particular region.   Historically, standard surface observations (i.e. METARs from ASOS/AWOS) have been the only data source available for monitoring precipitation type in winter weather events.   Additionally, SPC has been decoding winter-weather-related local storm reports (LSR) for a few years and has recently obtained the crowd-sourced reports from mPING (Meteorological Phenomena Identification Near the Ground).  The goal of this investigation will be to compare the reported precipitation type from these alternative sources against those identified in the METARs.  The validation will incorporate a check on both spatial and temporal continuity and consistency to identify potential biases and overall accuracy.  Ultimately, the results will hopefully guide future refinement and improvement of an observationally based, gridded precipitation-type system.