Objective analysis of the structure of three-dimensional atmospheric fronts

A computational method for diagnosing three-dimensional atmospheric fronts from temperature, wind, and geopotential fields on a three-dimensional regular grid is proposed. The criterion, which serves for the diagnosis of atmospheric fronts, is discussed. The weights of the input information about the mentioned fields are optimized based on the maximal difference between the correlation functions for (a) pairs of particles separated by the front and (b) pairs from one synoptic mass. These weights were different for different baric levels. The correlation functions and the optimization of weights were estimated on the basis of the archive of fields of the NCEP objective analysis on the half-degree latitude-longitude grid and data from aerological observations. The results of numerical experiments on the construction of atmospheric fronts are presented. Applying the described method to fields predicted for a term of up to 36 h showed that errors in the prognostic models introduce a relatively weak distortion into the geometry of atmospheric fronts.