Cluster Analysis Techniques to Separate Air Motion and Hydrometeors in Vertical Incident Profiler Observations

Abstract

Profilers operating in the UHF range are sensitive to both Bragg scattering from radio refractive index structure and to Rayleigh scattering from small point targets. Identification of the scattering process is critical for proper interpretation of these observations, especially the data collected from the vertical incident beam. This study evaluates the performance of Doppler velocity thresholds as a means to separate air motions from hydrometeor motions in vertical incident profiler observations. This evaluation consists of three different steps. First, using two collocated profilers operating at different frequencies, the observations are unambiguously identified as Bragg or Rayleigh scattering processes. Second, the observations are separated into either air or hydrometeor motion using only the data from one profiler. The third step quantitatively evaluates the performance of the single profiler separation techniques by counting the number of correct classifications and adjusting the count by the number of incorrect classifications. Constant Doppler velocity threshold methods are acceptable methods to separate air motions from hydrometeor motions only after the correct threshold is determined. This study presents a cluster analysis method that robustly and objectively separates air from hydrometeor motions. The introduced cluster analysis produces two thresholds. The first threshold is a Doppler velocity threshold that is a function of reflectivity. The second threshold is the maximum reflectivity in which the Doppler velocity threshold divides the observations into two statistical distributions using the Kolmogorov?Smirnov statistical test. The cluster analysis method quantitatively performs better than constant Doppler velocity threshold methods, and is a repeatable, self-adapting, statistically based procedure.