A Case Study of the Persistence of Weather Forecast Model Errors

Report No. ARL-TR-3418
Authors: Barbara Sauter
Date/Pages: January 2005; 48 pages
Abstract: Decision makers could frequently benefit from information about the amount of uncertainty associated with a specific weather forecast. Automated numerical weather prediction models provide deterministic weather forecast values with no estimate of the likely error. This case study examines the day-to-day persistence of forecast errors of basic surface weather parameters for four sites in northern Utah. Although exceptionally low or high forecast errors on one day are more likely to be associated with a similar quality forecast the following day, the relationship is not considered strong enough to provide beneficial guidance to users without meteorological expertise. Days resulting in average forecast errors showed no persistence in the quality of the subsequent day?s forecast. More sophisticated methods are needed to generate and portray weather forecast uncertainty information.
Distribution: Approved for public release
  Download Report ( 0.892 MBytes )
If you are visually impaired or need a physical copy of this report, please visit and contact DTIC.

Last Update / Reviewed: January 1, 2005