Time Scales of Systematic Errors in a Numerical Weather Forecast.
Abstract
Many numerical model verification systems are handicapped by their inability to separate random errors, defined as scales of motion smaller than the concerned scale, and systematic errors. Random errors may be eliminated by time averaging over a minimum allowable data period. Statistical methods are implemented to determine a minimum data period needed to constitute a time scale within which numerical forecasts errors are truly systematic and not smoothed fields of rapidly varying random errors for a synoptic scale event. Long term or model climatological error patterns of a specific event are compared with randomly determined shorter term error patterns through statistical tests and measures to provide results which indicate a minimum case number needed to sufficiently smooth out unwanted random errors.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1982
- Accession Number
- ADA113374
Entities
People
- Patrick A. Harr
Organizations
- Control Data Corporation