王 萍，顾 恺，侯谨毅
AuthorsHTML:王 萍，顾 恺，侯谨毅
AuthorsListE:Wang Ping，Gu Kai，Hou Jinyi
AuthorsHTMLE:Wang Ping，Gu Kai，Hou Jinyi
Unit:天津大学电气自动化与信息工程学院，天津 300072
Unit_EngLish:School of Electrical and Information Engineering，Tianjin University，Tianjin 300072，China
Abstract_Chinese:强对流天气是由空气的剧烈垂直运动导致的，以上升气流和下沉气流为主，它突发性强、破坏力大，对国民经济与民众安全有重大危害，一直是天气预报与分析中的难点和重点．如果可以简捷地描述出包含上升气流和下沉气流的三维流场，将有助于强对流天气的分析．本文基于多普勒天气雷达反射率因子数据，在单体分割后通过双线性插值法获得三维格点插值点云，假设相邻体扫时刻同一数据点的反射率因子基本不变，设定反射率因子阈值，对相邻体扫时刻下反射率因子大于等于阈值的点云通过Marching Cubes方法构建三维曲面．然后，对于这对曲面设计出一种非刚性的曲面匹配方法来描述强对流天气的三维流场：设计包含曲面法向量和曲率信息的曲面各点形态特征向量，进行匹配区域限制，设置附加条件解决一点对多点匹配的不确定性问题．通过该曲面匹配方法获得三维流场中各数据点的位移向量，从而近似描述出强对流天气的三维流场．最后，通过由各种三维空间变换得到的模拟曲面数据进行曲面匹配方法的效果测试，综合平均匹配正确率达到87.0%．
Abstract_English:Severe convective weather is caused by the violent vertical movement of air in which updrafts and downdrafts predominate. This weather is sudden and destructive，causes great harm to national economies and human life，and is difficult to forecast and analyze. A concise description of the three-dimensional airflow field including the updrafts and downdrafts would facilitate the analysis of severe convective weather. Based on reflectivity factor data obtained by the Doppler weather radar，we obtained the three-dimensional lattice-point-interpolated point cloud by bilinear interpolation method after segmenting the cells，wherein we assumed that the reflectivity factors of the same data point at two adjacent moments of body scanning are essentially identical. Next，we set the threshold value of the reflectivity factor and using the Marching Cubes method，constructed two surfaces from point clouds with reflectivity factors greater than or equal to the threshold at the two adjacent moments of volume sweeping. Then，for this pair of surfaces，we designed a nonrigid surface matching method to describe the three-dimensional airflow field of severe convective weather. In this method，we design the shape feature vector of each point of the surface，including infor-mation about the normal vector and curvature，limit the matching area，and set additional conditions to solve the uncertainty problem of one point matching many points. To approximately describe the three-dimensional airflow field of severe convection weather，we obtain the displacement vector of each data point of the three-dimensional airflow field by surface matching. Finally，we tested the surface matching method by simulated surface data obtained from various three-dimensional space transformations and found its comprehensive average matching accuracy to be 87.0%．
Keyword_Chinese:强对流天气；多普勒天气雷达；三维流场；曲面匹配；三维可视化
Keywords_English:severe convective weather；Doppler weather radar；three-dimensional airflow field；surface matching；three-dimensional visualization

PDF全文下载地址:http://xbzrb.tju.edu.cn/#/digest?ArticleID=6591