Dynamic Model Research Of Urban Economic Development And Landscape Pattern Changes

Urban ecosystems have the most special landscapes in the Earth’s major ecosystems. Relative to forest ecosystems and other natural ecosystems, ecological structure of urban ecosystem is very simple, and usually the dominant landscape is urban land, moreover, urban ecosystems is profoundly influenced by human activities. Economic development, technological development, and social development have a critical impact on urban landuse conversion namely urbanization, in turn, urbanization has profoundly changed urban ecosystem structure, function and dynamics. Today, many ecological problems, environmental problems confronting humanity are related very closely with urbanization, therefore the quality of the ecological environment status can fully reflect the rationality of urban land use. Quantifying landscape patterns is an essential part of monitoring and evaluating ecological consequences of urbanization. In many city models, cellular automata have been widely used as a dynamic model. Using this model, researchers can simulate and predict urban land changes.Paper chooses Haidian District and Chaoyang District in Beijing as the study area for their rapid urbanization. Time width is 1984 to 2009, a total of 25 years, at intervals of five years. Paper uses the RS and GIS software for remote sensing image interpretation, access to land use classification maps. And then according to the theory of landscape ecology, using landscape pattern index methods and FRAGSTATS software, landscape index results can be computed by land use classification maps. At last, paper analyzes landscape pattern changes in a historical retrospective, and makes recommendations of urban landscape planning based on landscape patterns. The results showed that: generally, as the development of urbanization landscape should be gradually fragmented, but for the study area, by contrast, various indicators show that with the increase of artificial building area and the reduce of bare land, landscape fragmentation decreases, the extent of landscape structural complexity decreases and landscape diversity decreases. The biggest difference between the urban landscapes and the natural landscapes is the strong human interference on urban landscapes, but it is precisely because of the human management, we can optimize the urban landscapes based on understanding of landscape patterns.In addition, in order to simulate the dynamic change of urbanization, the study implements dynamic simulation of urban land use change based on cellular automata theory using logistic regression. In Arc GIS software support, paper accesses a series of natural and economic and social constraints of urbanization. Paper uses the 1984 and 1994 classification maps as the training set to generate cellular automata rules based on logistic regression. Then applying the rules to the 1994 classification map to generate the 2004 land use simulation map, comparing it with the 2004 remote sensing image classification map, paper verifies the accuracy of the model is 70.02% in the Haidian District and 64.94% in the Chaoyang District.