A Comparison of Two Objective Analysis Techniques for Profiler Time-Height Data

Abstract

Two methods for objective analysis of wind profiler data in time-height space are proposed and compared. One is a straightforward adaptation of a procedure developed by Doswell for introducing time continuity into a sequence of spatial analyses. The second technique, named the correlation method, introduces a new rationale for selection of the Barnes filter parameter that is based on knowledge of the statistical structure of wind profiler data. The advantages and disadvantages of each method are discussed. It is noted that the correlation method, in principle, allows the deduction of consistent sampling intervals in time and space for the most dominant phenomena resolved by the data provided by a given atmospheric observing system. It is recommended that an objective analysis of wind profiler data be performed before single- or multiprofiler kinematic calculations are made. In addition, it is shown that the positions of extrema in kinematic quantities computed from profiler triangles are relatively insensitive to the number of passes used in the analysis procedures. In fact, it is demonstrated that multipass Barnes-type schemes can overfit the original data, suggesting that a one-pass method may be preferable provided that the filter parameter is selected properly.