An advanced algorithm for recognizing wind shear using airborne Doppler weather radar

Abstract

Wind shear must be rapidly recognized during flight because it represents a safety threat. The conventional recognition algorithms rely mainly on the processing of radar signals. Consequently, their results are not two-dimensional and are difficult to visualize, preventing pilots from rapidly and accurately recognizing and predicting the position of shear lines. This study proposes an algorithm for recognizing regions of horizontal wind shear at different altitudes from airborne weather radar. A simulation analysis is used to analyze the radial velocity and spectral width data gathered by ground-based Doppler radar. The algorithm employs the Perona-Malik partial differential equation model to pre-treat Doppler radar base data to reduce noise, maintain fidelity, and remove isolated points. A region-growing algorithm, using radar spectral width and average radial velocity data, is used to rapidly identify regions of horizontal wind shear at different altitudes. By analyzing precipitation events in Fuyang, Chengdu, and Nanjing in China, this study verifies the feasibility of the proposed algorithm. As the simulation results show, the proposed algorithm has better accuracy, better speed, and better shear line continuity than conventional recognition approaches. The improved recognition speed could help pilots recognize and predict wind shear to guarantee flight safety.