Statistical Detection of Anomalous Propagation in Radar Reflectivity Patterns

Abstract

The problem of anomalous propagation (AP) echoes in weather radar observations has become especially important now that rainfall data from fully automatic radar systems are sometimes applied in operational hydrology. Reliable automatic detection and suppression of AP echoes is one of the crucial problems in this area. This study presents characteristics of AP patterns and the initial results of applying a statistical pattern classification method for recognition and rejection of such echoes. A classical radar (MRL-5) station operates in central Poland performing volume scanning every 10 min. Two months of hourly data (June and September of 1991) were chosen to create learning and verification samples for the AP detection algorithm. Each observation was thoroughly analyzed by an experienced radar meteorologist. The features taken into account were visually estimated local texture and overall morphology of echo pattern, vertical echo structure, time evolution (using animation), and the general synoptic information. For each 4 km ? 4 km pixel of 933 observations the human classification was recorded resulting in a sample of 631 166 points with recognized echo type, 14.6% of them being AP echoes. The unsuppressed AP echo impact on monthly accumulated precipitation was 59% of the actual sum for the month of June and as much as 97% for September. Three Bayesian discrimination functions were investigated. They differ in selection of the feature vector. This vector consisted of various local radar echo parameters: for example, maximum reflectivity, echo top, and horizontal gradients. The coefficients of the functions were calibrated using the June sample. The AP echo recognition error was about 6% for the best-performing function, when applied to an independent (September) sample, which would make the method acceptable for operational use.