As an important method for the evaluation of quantity of area rainfall, Thiessen polygon method is widely applied because of its high calculation accuracy and fast computation. The calculation of Thiessen polygon method is simple because only area data of sample point is needed. When dataset is stable, accuracy of Thiessen polygon method is higher than the arithmetic average method. But there are still problems, for example boundary treatment and efficiency of Thiessen polygon method.Because border points are complex, current boundary treatment method is unsystematic or feasible theoretically without operability in existing research results. In this paper vector method is used to make unified treatment for boundary points, and avoid emergence of singular points. When dataset is massive and complex, the efficiency of Thiessen polygon method becomes an important indicator. Large amount of data search operation will take a great deal of time in the process of Thiessen polygon method. A higher efficient and practical topological structure is proposed, which is used to construct the topological relationships among the points, edges and triangles, to achieve more efficient search speed. Based on the research of the above two aspects, Thiessen polygons are built up on the base of Delaunay triangulation and the corresponding topological structure among the points, edges and triangles are set up, and network model about three entities is built based on HASH table principle to provide a basis for the following steps. Then the search of Thiessen polygon is to be carried out to find the border points by vector dot product approach to solve the border issue.Above optimized Thiessen polygon method was applied to the test project of Shijingshan District of Beijing with more than 20000 data points. The result shows topological structure method is efficient in the fast search of topological relations of the points, edges and triangles, and the treatment of boundary and singular points within massive and complex data can be improved evidently.
Keywords-Thiessen polygon method; boundary treatment; topologyI.