微波雷达双边滤波云检测新方法的研究

A Novel Bilateral Filter Hydrometeor Detection Method for Microwave Radar

云是影响天气、气候变化的重要因子,也是目前造成气候模拟的最大不确定因素之一,云的长期准确观测,对验证和约束模式结果,减少云对气候变化模拟预测造成的不确定性有重要意义。毫米波云雷达通过接受云滴粒子回波信号,可以获取云的三维结构特征,是云探测的有力工具。以兰州大学半干旱气候与环境观测站Ka波段云雷达和CloudSat星载W波段云雷达为例,详细介绍了一种基于双边滤波的噪声压缩思想,对地—空基云雷达信号回波和背景噪声区分识别的云检测改进算法。通过将图像平滑处理的双边滤波思想引入毫米波雷达云检测算法当中,在压缩雷达背景噪声的同时,保持了弱信号边缘的清楚完整性,从而识别出更多被以往算法忽略的真实信号,并与地面和星载激光雷达同步观测对比,证实了该算法能显著降低传统算法的漏检率,提高毫米波雷达云检测准确度。同时,还以飞行器航迹检测为例,探讨了该算法在微波雷达对目标物探测应用中的改进,说明该算法对增强雷达小散射截面目标物识别方面的普适性。由此认为该方法对微波主动雷达目标物检测,特别是弱信号物体识别提供了一种有效改进。

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 SemiArid 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.