Study On Succession Of Spatial Landscape Pattern Of Land Use/cover In Mountains Around Sichuan Basin

In the development campaign of the western regions and the sustainable development of society and economy, Sichuan province, which holds rich nature and culture resources in the hinterland of western China, play a very important strategic role. Mountains around Sichuan Basin have rugged terrain, complex topography and geological structure, multiplicity lithology, and fragile ecological and geological environment. With the rapid development of resource exploitation and economic construction, the equilibrium of environmental carrying capacity is broken. Sequentially, Sichuan province is confronted with increasingly serious environmental problems. To solve this problem, the basic idea is to fully understand the objective nature laws in the environmental changing, and achieve sustainable development in resource exploitation and socio-economic development. Land use/cover is closely linked with the environment, which is the most prominent landscape features in the surface system of earth. Its changing reveals the important evolution law in ecology and environment, and become an important content in changing research of global environment.The dissertation chose mountains around Sichuan Basin as study area. Surrounded with spatial pattern succession of land use/cover landscape, it study spatial patterns characteristics and evolution laws of land use/cover landscape in complex mountainous by adopting principles and methods in ecology, environment and system science, utilizing the technology of remote sensing and geographic information system, and combining with geographical statistical analysis and dynamic simulation. Main knowledge and result as following:(1) The spatial distribution pattern and changing characteristics of main land use landscape, such as the arable land, grassland, woodland, water, urban and rural construction land and unused land, are studied by analyzing land use/cover landscape in the patch level and landscape level. Studies have shown that natural conditions of mountains restrict the type of land use. Therefore, the landscape of land use has significantly different distribution patterns and trends within different mountain areas. Their distribution patterns are characterized by the natural environment (topography, temperature, precipitation, etc.) and human activities together. In the low altitude, human activities have a dominant impact on the land-use. With the elevation, the effect of topography increased gradually on land use, and from the combined effect of topography and human activities to the dominant effect of terrain, which change distribution pattern of landscape of land use by bringing about influence on fragmentation, uniformity, connectivity and the complexity.(2) The spatial characteristics of the NDVI vegetation index time series are studied. The results showed that NDVI index was significantly affected by altitude, in different altitude, have independent wave curve. Except for alpine region, NDVI index of mountains around Sichuan basin have decreased tendence. Annual average NDVI vegetation index have obvious characteristics of gradation distribution, but the order of gradation distribution aren't consistent with the altitude. Quarter(Month)NDVI vegetation index has obvious seasonal cycle trend. In addition, the type of land use bring a direct impact on vegetation index., and there are important links between land use type and NDVI vegetation index.(3) In order to achieve a short-term forecast of NDVI vegetation index, chaos theory and chaotic time series prediction method are adopted to reconstruct the phase space of time series of NDVI vegetation index. Studies have shown that the NDVI time series is non-linear deterministic system with chaotic characteristics. Linear fitting analysis, for new time series, shows the NDVI vegetation index is still decreasing in the study area. In different mountain regions, the NDVI vegetation index has different changing degree in changing direction. In the short term, ecological environment become worse in non-mountain and alpine areas, however in low-mountain, middle mountain and high mountain region, ecological environment is improved.(4) In land use pattern research, the dominant factor and drive mechanism are the cores of the study. Association rules data mining model is constructed to extract multidimensional association rules between elements of landscape and environment,and to explain the environmental driver factors of landscape pattern of land use. Conclusion about the association rules shows that elevation, slope, climate, soil, and vegetation are important driving factors which affect landscape pattern of land use, but the correlation between slope and land use is relatively weaker. At the same time, it prove that there are complex multiple correlation among topography, vegetation, soil and climatic. These correlations reflect intrinsic link between spatial distribution laws of landscape pattern of land use and environmental factors in the complex natural environment of mountains.(5) Based on the relationship research of land use and environment, relevant environmental factors are selected to establish dynamic evolution model of spatial pattern of land use landscape, which predict changing of the landscape. In the model, geographical cellular automation is introduced as the core of mode. Neural network is used to construct transition rules of cellular automata state, and genetic algorithm is adopted to optimize neural network. Simulation results show that arable land is decreasing and large numbers of arable land was converted to other types of land use, therefore, protection of farmland is faced with increasingly severe situation. Forest area increases year by year, forest resources get better protection, and so ecological environment restored some. However, the development and utilization of land resources has led to rapidly increasing scale of urban and rural construction land. Land use changing is the results of the natural and human factors together. After taking full account many relevant factors, the model can forecast the trend of land use well, and has good reliability and credibility. Furthermore, through the dynamic simulation of cellular automata, the evolution of land use can directly be displayed.Through the study, spatial and temporal variations of land use/cover are analyzed, and simulation system and model for forcasting evolution of land use pattern is constructed. Finally, those objective laws of mountain eco-environmental changing can be understanded by analyzing appearances from changing of land use pattern. As a result, it will provide references for land management and planning, and it is helpful for promoting sustainable development of resources and environment in mountains.