Downloads & Free Reading Options - Results

Thiebaux and Pedder (1987) have categorized types of objective analysis used in the atmospheric sciences into three classes: surface fitting, empirical linear interpolation, and statistical objective analysis. The recursive filter method described here falls in the second lot, and is thus something of a throwback from the currently fashionable third category. However, the recursive filter is faster and more economical and thus more practical for applications with small computers and limited time. For this reason, it has been implemented on the University of Wisconsin McIDAS to provide gridpoint analyses from the real time data bases available on that system. Why another empirical fitting technique when other excellent methods exist (e.g., Cressman, 1959 or Barnes, 1973)? We believe that the recursive filter technique provides rather unique capabilities for handling domain boundaries and background fields, and for locally variable scaling dictated by observation quantity and quality. Also, it provides gridpoint estimates of the reliability of the final analysis as a by-product (somewhat like the more elegant statistical schemes). This is useful for both qualitative and quantitative application of the analyses.

Thiebaux and Pedder (1987) have categorized types of objective analysis used in the atmospheric sciences into three classes: surface fitting, empirical linear interpolation, and statistical objective analysis. The recursive filter method described here falls in the second lot, and is thus something of a throwback from the currently fashionable third category. However, the recursive filter is faster and more economical and thus more practical for applications with small computers and limited time. For this reason, it has been implemented on the University of Wisconsin McIDAS to provide gridpoint analyses from the real time data bases available on that system. Why another empirical fitting technique when other excellent methods exist (e.g., Cressman, 1959 or Barnes, 1973)? We believe that the recursive filter technique provides rather unique capabilities for handling domain boundaries and background fields, and for locally variable scaling dictated by observation quantity and quality. Also, it provides gridpoint estimates of the reliability of the final analysis as a by-product (somewhat like the more elegant statistical schemes). This is useful for both qualitative and quantitative application of the analyses.