Hybrid Method to Identify Second-trip Echoes Using Phase Modulation and Polarimetric Technology

• Published in: Advances in atmospheric sciences Vol. 38; no. 3; pp. 480 - 492
• Main Authors: Zhang, Shuai, Min, Jinzhong, Zhang, Chian, Huang, Xingyou, Liu, Jun, Wei, Kaihua
• Format: Journal Article
• Language: English
• Published: Heidelberg Science Press 01.03.2021; Springer Nature B.V; Key Laboratory of Meteorological Disaster of Ministry of Education, Nanjing University of Information Science and Technology, Nanjing 210044, China%Beijing Metstar Radar Co. Ltd., Beijing 100094, China%Taizhou Meteorological Administration, Taizhou 225400, China%Guangdong Meteorological Observatory, Guangzhou 510080, China
• Subjects: Algorithms; Atmospheric Sciences; Data analysis; Data processing; Doppler sonar; Earth and Environmental Science; Earth Sciences; Echoes; Geophysics/Geodesy; Identification; Identification methods; Information processing; Meteorology; Original Paper; Phase modulation; Polarimetry; Regions; Signal processing; Signal quality; Technology; Tornadoes; Turbulence; Wind; Wind shear; second-trip echo; 相位调制; 二次回波; 风切变; polarimetric radar; wind shear; phase modulation; 偏振雷达
• ISSN: 0256-1530; 1861-9533
• DOI: 10.1007/s00376-020-0223-3

For pulse Doppler radars, the widely used method for identifying second-trip echoes (STs) in the signal processing level yields significant misidentification in regions of high turbulence and severe wind shear. In the data processing level, although the novel algorithm for ST identification does not yield significant misidentification in specific regions, its overall identification performance is not ideal. Therefore, this paper proposes a hybrid method for the identification of STs using phase modulation (signal processing) and polarimetric technology (data processing). Through this approach, most of the STs are removed, whereas most of the first-trip echoes (FTs) remain untouched. Compared with the existing method using a signal quality index filter with an optimized threshold, the hybrid method exhibits superior performance (Heidke skill scores of 0.98 versus 0.88) on independent test datasets, especially in high-turbulence and severe-wind-shear regions, for which misidentification is significantly reduced.