Researches On Land Use Change Analysis Of Spatio-temporal Visualization

Land use change has become a hot topic in global change research. The existing land-use change analysis model has its own scope of application and suitable to solve problems. Spatio-temporal visualization is the important method of land use change analysis. It enables people to in-depth and comprehensive understand the spatio-temporal changes of land. In order to strengthen scientific research in land use change, the thesis built the spatial effect model to analysis the process of land use change and its influence factors, proposed in spatio-temporal visualization method of land use change, and discussed the new update method of spatial database from the perspective of artificial intelligence.Land database must appropriately be organized and constantly be updated to make better use. This thesis outlines the current spatial database update modes and discusses some related practical issues. In order to shorten land database update periods and improve the intelligence level of system, an omnidirectional and incremental spatial database update mechanism was proposed by using the cooperative work characteristic of Wiki idea and the multi-agent technology in artificial intelligence research fields. Special resolve method, mechanism elements and data update processes are presented in detail. This multi-agent spatial database update mechanism based Wiki idea can arouse the interaction of users, shorten the period of spatial data update, improve the use value of land data and make the spatial database update more convenient and flexible.In the process of land use change, cultivated land is more concerned for food safety and ecological security. Because of the limitations of obtainable data, by looking year as time scale, Chengdu plain and its county as spatial scale, as viewed from the amount and spatial change of arable land, this research analyzed spatio-temporal change of cultivated land in Chengdu plain from 1998 to 2007. Based on the spatial autocorrelation of cultivated land in Chengdu plain, its intrinsic factors and outer factors are systemically analyzed by setting up spatial regression models. The result indicates that the downtrend of cultivated land in Chengdu plain become slow, spatial structure change go stable, influence factors and their influence degrees is different every year. Accordingly, several suggestions are proposed in the study to realizing the harmonious development of people and cultivated land by adjusting industry structure and land consolidation and reclamation.To fit the higher complexity of area boundary in map, we derive the equal-distance dichotomy method that determines the centre and radius of maximum inscribed circle. It creates equal-distance polygon based on the method of circular arc for convex vertex. Then, the maximum inscribed circle's centre and radius is computed by using dichotomy. Comparison with existing methods, the equal-distance dichotomy method can obtain the best results and be applied to any shape of polygons in cartography. Determining visualization scope is an important part of map design. And it has a direct impact on the beauty of map and expression of information. To solve problem of overlap with other visualization results or other graphical features, looking upon geometric center of point or maximum inscribed circle's centre of area as vision center of its visualization, in the four ways that view constraints, no view constraints, and whether and the characteristic values associated with the size of visualization results, the maximum circular and square sizes of points or areas are obtained by using breaking point theory. Experimental results show that this method is simple and effective.The existing space-time visualization methods of land use change are difficult to meet users' requirements by lack of comparability. Considering the users' specific requirements in analyzing and understanding land use change, the static space-time visualization method for land use change is proposed by improving display mode, reducing dimension and compounding graphs. In our method, colors are used to distinguish land type and size is used to represent land amount. Users may intuitively know the trend of land use change with time and compare the land use change of regions in different time. Selecting comparability, readability, whether overlap, and data clarity as evaluation factors by analyzing existing visual evaluation methods, the experiment result shows that this method is simple and has more comparability.