Research On The Northwest Pacific Tropical Cyclone: GIS-based Approach

China is one of the Countries that suffer Tropical Cyclone most frequently and severely, hence the research of Tropical Cyclone is always significant in the long run. Many researchers in meteorology department and scientific field have taken great efforts to the analysis and forecast of Tropical Cyclone.The more rapid the development of Science and Technology, the more advanced the ways human gain information. However, the traditional research methods cannot adapt to the processing and analysis of mass information. Therefore, plenty of useful knowledge cannot be fully discovered in the mass information, which lags the development of analysis and decision-making.With the support of GIS, the thesis has a discovering research on two problems, Tropical Cyclone source spatial distribution pattern and Tropical Cyclone path similarity, which are most valuable for the forecast of Tropical Cyclone. The thesis develops a system of Northwest Pacific Ocean Tropical Cyclone Path Index based on the GIS analysis toolkit and ComGIS, whose key is source analysis and path similarity.There lies four chapters in the thesis. Chapter 1 depicts the research meaning, refers the relating research in the meteorology and lists the work in the thesis.Chapter 2, Tropical Cyclone source spatial distribution pattern research. Firstly, the thesis discusses the concept, theory and method of spatial point pattern, then analyses the main character of Tropical Cyclone source spatial distribution pattern with the use of Kernel Density Estimation (KDE) and distance-based G function and F function, at last, makes an exploring analysis on the two rules with KDE: Tropical Cyclone source spatial distribution changes with the seasons and with the Tropical Cyclone Intensity Level.Chapter 3, Tropical Cyclone path similarity research. On the basis of relating research, the thesis explains the general concept of geographic objects' spatial similarity, the discusses two methods to confirm the similarity of two line geographicobjects: Peak Model and Nearest-Neighborhood distance extraction. Meanwhile, the thesis points out the disadvantages of Nearest-Neighborhood distance extraction in extracting Tropical Cyclone similar path, after analyzing the disadvantages, the thesis makes full use of GIS spatial analysis solution and proposes an improved Nearest-Neighborhood distance extraction.Chapter 4, Tropical Cyclone index and analysis System with GIS analysis toolkit in Northwest Pacific Ocean. GIS analysis toolkit incorporates the system, which is based on the key aspects of Tropical Cyclone source analysis and path similarity indexing. This chapter discusses the general design, development environment and function introduction.Chapter 5, Conclusion and Discussion. The thesis proposes several problems which need further solution on the basis of the conclusion.