A High-Resolution Freezing Precipitation dataset for the South-Central U.S

Freezing precipitation, including freezing rain, drizzle, and ice pellets, is a substantial winter weather hazard to surface and air transportation. While winter weather is infrequent in the southern United States, winter storms that impact this region often produce multiple precipitation phase-types, and in some cases, damaging ice storms. Most quantitative climatological length (> 30 years) datasets for freezing precipitation are obtained from in-situ instrumentation, such as National Weather Service (NWS) automated observing system stations (‘ASOS'), which are fairly widely spread, typically with the longest records at major city airports, leaving large areas without comprehensive spatial and temporal information. In recent years, observational and model–derived atmospheric reanalyses have provided spatially gridded long-temporal length datasets for multiple meteorological variables. Using one such reanalysis, the ‘North American Regional Reanalysis' (NARR), 1979-2014, with 32-km horizontal grid spacing and 3-hour temporal resolution, we develop a freezing precipitation dataset for the south central U.S. This region includes U.S Department of Transportation Region 6 states: Oklahoma, Texas, New Mexico, Louisiana, and Arkansas. This dataset was developed through use of multiple precipitation-type algorithms applied to the NARR vertical thermal profile, with an additional estimate from NARR categorical precipitation type variables. We calculate 3-hourly, daily, monthly and annual ice counts, liquid water equivalent, and associated meteorological conditions (e.g., temperature, precipitable water). Snowfall counts and liquid water are also assessed. NWS first-order stations sites were used to evaluate the performance of this dataset. We found that the statistical properties of the derived freezing precipitation climatology have excellent agreement with observations, including inter-annual variability and severe ice days/years, although light freezing drizzle events were poorly resolved. This dataset provides a gridded long-term record of freezing precipitation, and may be useful to transportation planners and researchers in understanding the regional climatology and vulnerability of their region to icing hazard. We demonstrate how these data can be aggregated for geographic and political sub-domains, and the types of statistics that will be made available to end-users.