Radar identification of clouds

Abstract

1. The algorithm of radar recognition of meteorological objects on the base of autoregressive model is proposed. 2. The measurement complex is developed on the base of the modernized incoherent pulse meteorological radar MRL-1 and experimental study of the signal reflected from various meteorological objects are performed, including high-altitude cumulus clouds, some worse for cumulonimbus clouds and not satisfactory for the “angel-echo” which are formed by Bragg reflections from inhomogeneities in surface layer of a clear atmosphere. 3. It is shown that the probability of correct recognition is extremely high >0,95 for high-altitude cumulus clouds even in the case of small AR model order and the probability is some worse for cumulonimbus clouds. 4. In the case of the “angel-echo” the probability of correct recognition is lowest because diffuse scattering from the “angel-echo” cannot be satisfactory described by a set of harmonic oscillations which are typical for the AR model.