This investigation examines the utility of RUC-2 model analysis soundings as a surrogate for observed soundings in supercell environments across the contiguous United States. These model soundings are available hourly at the Storm Prediction Center, interpolated from the full-resolution grid files. Their increased spatial and temporal availability, compared to standard rawinsonde launches twice daily, allows relatively rapid collection of a large number of close proximity soundings to represent various supercell environmental characteristics. A preliminary study by Thompson and Edwards (2000) is extended to a larger sample (roughly 150 cases) of RUC-2 analysis soundings. During the period from March 1999 through June 2001, supercells were identified via radar imagery across the contiguous United States. Observed regional proximity soundings were collected for each supercell case where the closest available sounding revealed CAPE, and there was no obvious contamination by recent precipitation. Spatially and temporally matching RUC-2 model soundings were also generated for each, along with a one hour forecast from the prior RUC-2 model run (e.g., one hour forecast from 23 UTC valid at 00 UTC).

This examination evaluates comparative statistical characteristics of the RUC-2 and observed soundings. Each sounding is interpolated to standardized levels with 25 mb vertical resolution. A Student's t test is used at each level to compare the analysis to the one hour forecast soundings and to available observed soundings. Error characteristics of severe storm parameters derived from the RUC-2 analysis soundings are provided, along with recommendations on the use of these model analysis soundings as the basis for supercell proximity sounding investigations such as Thompson and Edwards (2002, this volume).