Lattice implementation of “superresolving” methods for meteorological objects spectra estimation

Abstract

The paper addresses a practical implementation of “superresolving” spectral estimation (SE) methods for the tasks of meteorological objects energy spectra reproduction in pulse Doppler weather radars. We demonstrate that the universal adaptive lattice filters (ALF) are an effective tool to solve this task. A specificity of ALF-based formation of spectral functions of various “superresolving” SE methods is described. The recurrent algorithm of ALF adjustment is presented. On each step of recursion, it includes one-fold correction of ALF parameters by algorithms of K - rank modification instead of usually applied K - fold correction of ALF parameters by algorithms of one-rank modification. We demonstrate lesser computational complexity and higher numerical stability of the proposed algorithm as compared with known ones. The results of proposed algorithm investigation in existing weather radar are presented.