基于多普勒天气雷达数据的中层径向辐合自动识别及其与强对流天气的相关性研究

本文提出了一种从多普勒天气雷达单仰角径向速度图中的正-负速度区域对入手自动识别一类强对流天气的中层径向辐合的方法.首先,根据雷达探测机理,解释了由气流形成的辐合场一定会在径向速度图中形成局部最大正速度区域和局部最大负速度区域的特点,然后构建正-负速度区域对的识别算法,通过来自多个单仰角径向速度图的区域对配准和相关信息,判断雷达径向速度图中是否存在着中层径向辐合,推算辐合的强度、延伸厚度等重要参数,确定得到最优垂直剖面图的剖切方位.经对客观存在显著中层径向辐合的384例样本和不存在明显中层径向辐合的356例样本的测试,本文方法对显著中层径向辐合的识别率达到100%、误识率为0,与人工方法相比,本文方法把对中层径向辐合的识别时间从分钟级缩短到秒级,同时自动给出其强度、高度、厚度、位置等丰富的定量信息和恰当的可视化垂直剖面图.利用给出的参数信息进行了与强对流天气的相关性研究,验证了中层径向辐合与对流行地面大风的较强相关性,同时发现,最强中层径向辐合的高度对区分强冰雹与强降水、中层径向辐合的强度对估计冰雹尺寸具有良好的指示性.

Automatic recognition of mid-altitude radial convergence and study on the relationship between the convergence and strong convective weather based on Doppler weather radar data

To identify the mid-altitude radial convergence of a strong convective weather automatically, we propose a method based on recognition of ‘positive-negative velocity region-pairs’ (region-pairs)in a single elevation angle of the Doppler radar radial velocity image. First of all, according to the principle of the radar detection, this paper explains the phenomenon that the convergence field formed by the airflow must produce a local maximum in positive or negative velocity region in the radial velocity image.The algorithms for recognizing these regions and matching the positive-negative pair are then devised.By searching a set of region-pairs with longitudinal extension, which are obtained from the multiple single elevation radial velocity images, we can judge whether there is a mid-altitude radial convergence in the convective storm, and estimate important parameters, such as the strength and extended thickness of the mid-altitude radial convergence.Finally, we determine the position of optimal section and present the cross-sectional view of the mid-altitude radial convergence. We have tested 384 samples with obvious mid-altitude radial convergence and 365 heavy rainfall samples without obvious mid-altitude radial convergence. Experimental results show that the recognition rate of obvious mid-altitude radial convergence is 100% and the false alarm rate is 0.Compared with the manual way by means of the cross-sectional view, the proposed method in this paper can more rapidly recognize the mid-altitude radial convergence (and reduce the recognition time from minutes to seconds). At the same time, it can present a great deal of quantitative information, including the strength, height, thickness, and position of the mid-altitude radial convergence.Furthermore, it shows the cross-sectional view automatically.We can obtain good results from the comparison between the mid-altitude radial convergence and strong convective weather by using the given parameters.We test and verify the strong correlation between the mid-altitude radial convergence and severe surface wind.Moreover, the height of the strongest mid-altitude radial convergence plays an important role in discrimination of strong hail and torrential rain. Also the strength of the mid-altitude radial convergence can be used to estimate the maximum dimensions of the hail.