Regional-Scale Analysis and Forecasting
Abstract
This final report completes the description, begun in the interim report, of the development of the regional-scale analysis and forecasting procedure (RAP). The RAP is based on multivariate optimum interpolation of observations selected rigorously by the forward stepwise regression procedure. The complete observation set undergoes a unique buddy check that carefully eliminates those observations that adversely affected the analysis. The scheme used regionally-developed univariate and bivariate forecast error correlations, which were a function of distance between observations. The models are stratified by region, season, standard atmospheric levels, and forecast lengths. The models are superior to the model described in the interim report, and the models are shown to be applicable to rawinsonde and satellite observations. Results of verification experiments with numerical analyses of real data showed that the RAP concept is viable. Observing system simulation experiments (OSSEs) further verified the quality of RAP and the validity of the developed correlation models. In addition, the OSSEs provided relative confidence estimates of the quality of RAP analyses. An interesting forecasting procedure was recommended that extended the RAP analyses into a short-range forecast.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1992
- Accession Number
- ADA259894
Entities
People
- Christopher Redder
- Isidore Halberstam
- John Burgeson
- Ralph Shapiro