Description And Application Of Three-dimensional Convective Airflow Field Based On Weather Radar Data

Severe convective weather has always been a difficulty and important point in weather forecasting and analysis.It is sudden and destructive,which poses a great threat to people's property and personal safety.Severe convective weather is caused by the intense vertical movement of air,dominated by upward and downward airflow.Based on doppler radar data,this paper aims to describe the three-dimensional airflow field of severe convective weather.Under the assumption that the reflectivity factor of the same precipitation particle is basically unchanged at two adjacent times of volume sweeping,the displacement vectors of the particles in the three-dimensional airflow field are obtained by surface matching,thus approximating the three-dimensional airflow field in severe convective weather.The main work of this paper can be summarized as follows:(1)preprocessing the doppler radar data: after the cell is extracted,the reflectivity factor data of nine elevation angles are bilinearly interpolated and seventy layers threedimensional reflectivity factor lattice data are obtained.(2)Surface reconstruction for point clouds whose reflectivity factors are larger than the threshold value: denoising for point clouds whose reflectivity factors are larger than or equal to the threshold value,then two surface reconstruction methods are tried for point clouds: Delaunay triangulation and Marching Cubes method.Through the indicator designed in this paper to evaluate the effect of surface reconstruction,combined with some test cases,the advantages and disadvantages of the two methods are compared.(3)A new surface matching method is proposed to describe the three-dimensional airflow field in severe convective weather: A new surface matching method based on fusion of normal vector feature and curvature feature is proposed,which is accurate to point-to-point.In addition,this paper also designs a "relative coordinate vector" to restrict the matching area.Finally,Euclidean distance is used as similarity measure to realize the matching between surfaces and form "matching point pairs".For each set of "matching point pairs",the three-dimensional displacement vectors of the precipitation particles in the adjacent volume scanning interval are obtained by the coordinate difference between the two points.For these displacement vectors,this paper visualizes these displacement vectors through VTK,which can facilitate meteorological forecasters to observe directly the three-dimensional airflow field of severe convective weather from multiple angles and give quantitative description.In addition,this paper tests the effect of surface matching by simulating data.(4)Two applications based on three-dimensional airflow field description: based on this paper's method of surface matching to describe three-dimensional airflow field,two applications are carried out: the ability of this method to reveal the "three-stage" development law of convective cells is verified and a preliminary attempt of three-dimensional radar extrapolation is made.