Probabilities of Pressure Heights Forecast by Graphical and Numerical Methods

Abstract

A statistical theory developed previously is applied to two prediction techniques: graphical integration of Estoque's baroclinic model and numerical integration of a barotropic model. In both instances, synoptic observations are regarded as samplings of the actual state at observation time. Probabilities of forecast height values to occur in specified height intervals at 500 and 1000 mb are computed with graphical aids. Analogous probabilities are obtained at 500 mb with aid of numerical methods. These are limiting probabilities whose values depend in part on network density and initial synoptic analysis. Forecast procedures are described. Examples of probability forecasts and verifications are presented. It is found that theoretical limiting probabilities compared successfully with empirical probabilities for several synoptic cases, showing closest agreement in the numerical technique. This and other results are discussed.