An Algorithm for Automated Identification of Gust Fronts from Doppler Radar Data

Abstract

Gust fronts are weak narrow-band echoes of increased reflectivity at the background levels in the low-elevation fields of Doppler radar. An automated approach to gust front detection that relies on the image features of radar observations is presented in this paper. The algorithm is not sensitive to the variations in reflectivity values and gust front widths. The approach includes the following steps. First, a novel local binary with dual-template (LBDT) algorithm is designed as the fundamental algorithm to identify the potential areas of narrow-band echoes. Second, based on the disadvantages of the LBDT algorithm, several modifications are made, including splitting the intersecting lines, connecting the fragments, and filtering the edges and radial interference noise. Third, an optical flow method is used to determine whether a weak narrow-band echo is a gust front according to the prior knowledge that a gust front usually propagates in front of the associated generating storm. The results of experiments show that the proposed method can automatically identify gust fronts with a high probability of detection and a low false alarm rate. The automatic identification of gust fronts is potentially useful for accurate short-term weather forecasting, particularly in the forecasting of storm winds.