Advances in Earth Science ›› 2020, Vol. 35 ›› Issue (12): 1256-1269. doi: 10.11867/j.issn.1001-8166.2020.100

Previous Articles     Next Articles

A Novel Bilateral Filter Hydrometeor Detection Method for Microwave Radar

Jinming Ge( ),Xiaoyu Hu,Chen Wang,Zixiang Dong,Jiajing Du   

  1. Key Laboratory for Semi-Arid Climate Change of the Ministry of Education and College of Atmospheric Sciences,Lanzhou University,Lanzhou 730000,China
  • Received:2020-09-20 Revised:2020-11-20 Online:2020-12-10 Published:2021-02-09
  • About author:Ge Jinming (1982-), male, Lanzhou City, Gansu Province, Professor. Research areas include atmospheric radiation and remote sensing. E-mail:
  • Supported by:
    the National Natural Science Foundation of China “A study of an improved cloud mask method for CloudSat based on bilateral filter noise reduction scheme”(41875028);The Science and Technology Projects of Gansu Province “A study of target detection with small radar cross section”(20JR5RA301)

Jinming Ge,Xiaoyu Hu,Chen Wang,Zixiang Dong,Jiajing Du. A Novel Bilateral Filter Hydrometeor Detection Method for Microwave Radar[J]. Advances in Earth Science, 2020, 35(12): 1256-1269.

Clouds play an import role in weather and climate change, and are one of the most principal sources of uncertainty in climate projection. Long-term accurate observations of clouds are vital to validating and constraining model simulations, and reducing the uncertainty caused by clouds in climate models. The millimeter-wavelength cloud radar is a powerful tool for cloud observation by directly detecting signal backscattered from cloud droplets, and thus can provide cloud three-dimensional features. In this paper, we presented in detail an improved cloud detection algorithm to distinguish real cloud echoes from radar background noise. The bilateral filter idea from image process was adopted into millimeter-wave cloud detection algorithm, which compressed the radar noise while preserving cloud edge, therefore being able to identify more real weak signal ignored by traditional cloud mask methods. We also used the Ka-band Zenith Radar (KAZR) in the Semi-Arid Climate and Environment Observatory of Lanzhou University (SACOL), and the W-band cloud profiling radar aboard CloudSat, along with the synchronized lidar measurements to demonstrate that the improved algorithm could significantly reduce false negative rate, and increase the cloud detection accuracy. This paper also discussed the advantages of this algorithm for microwave radar in other remote sensing applications, taking track detection as an example. It shows that the algorithm could be generally used for small radar cross section target recognition. We believe this method will enhance the active microwave radar target detection ability especially for the objects with small radar cross section.

No related articles found!
Full text