A Comparison of Objective and Subjective Means of Weather Typing: An Example from West Texas

Abstract

An objective weather typing scheme first proposed by Christensen and Bryson (1966) was applied to surface and upper air variables for the period April?September of 1973?76 at Midland, Texas. Principal components analysis showed that moisture and temperature, which were represented by the first and second components, respectively, are most important in distinguishing day-to-day weather, while synoptic variables such as wind and pressure are relatively unimportant. The days of the study period were then assigned to weather types after applying multiple regression analysis and an objective grouping method. The principal disadvantages of the latter procedure are the large number of untyped days and the relatively few days assigned to types after the second. Suggestions for improving this grouping method are offered. Surface and 500 mb charts for the same period were examined and each day was typed according to a method specifying surface and upper air synoptic features. The two methods were then compared. The correspondence is very general because the variables manifest in synoptic representation, pressure and wind, are of only secondary significance in the principal components. Air mass changes are therefore more important in distinguishing day-to-day weather than are synoptic controls; this applies less so at the beginning and end of the convective season than in the middle of it. In a specific application of both methods to convective activity levels as inferred from the number of initial echoes per day, the greater distinction was made by the objective method, with highest levels occurring in the warmest and most moist types.