Study On Land Use Dynamic Evolution Simulation Of The North Branch Of Yangtze River Estuary

River-sea basin has diverse type, high intensive land-use style, active land-use change and significant spatial variation because of some natural and human factors such as the special geographical position along the coast of River-sea basins, silting and erosion of beach, reclamation and utilization of beach and so on.With the implementation of "river-sea linkage" development strategy, the high-speed development of socio-economic, the development of the river-sea coast industrial park, as well as the construction of river-sea coast ports, land-use change is extremely active in the North Branch of Yangtze River estuary basin.This paper takes land-use change in the North Branch of Yangtze River estuary as the research object and administrative region of Qidong city as study scope, study time span of 24 years, time interval of 6 years.Raised coastal images fusion model of lifting wavelet transform, selected ETM+ and ALOS images as reseach images and compared with the traditional fusion methods by combining subjective and objective evaluation. It is considered that the effect of lifting wavelet image fusion method is superior to the traditional method. Lifting wavelet method improved the image spatial resolution and maintained image spectral information betterly; Image classification Kappa coefficient showed that image of lifting wavelet fusion has the highest classification accuracy; lifting wavelet fusion can make full use of spatial resolution from panchromatic image and multi-spectral information from spectral bands, and effectively improved the image classification accuracy.Combined with GIS and RS technology, human-computer interact interpreted remote sensing images of 1989, 1995, 2001 and 2008, researched land-use pattern changes of 1989-1995, 1995-2001 and 2001-2008. Using the parameters Construction method, respectively analysis the North Branch of Yangtze River estuary land use changes from six aspects of land use structure, land-use transfer matrix, land use degree, land change extent, land use space change and diversification; Using landscape index method, chose landscape indexes such as concentration, separation, diversity, evenness and landscape fractal dimension, the study shows that landscape overall fragmentation degree is enhanced, landscape patch shape is tended to complicate, landscape diversity and evenness is increased in the North Branch of Yangtze River estuary; Using statistical methods, analysis land use sample semivariogram fitting and structure of the North Branch of Yangtze River estuary, research shows that the spatial correlation is greater , as the land-use distance between points is smaller in the range of 61 kilometers. When sample interval is greater than 61 kilometers, the space-relation of regionalization variables will not exist. The structure factor of land-use sample semivariogram fitting is around 0.7, which indicated that variation among the land use samples are more caused by random factors, land-use spatial dependence between samples is weak, but the space dependence is trend to enhance.Two different methods Statistics and spatial analysis is used to study the main driving factors causing land-use change in the North Branch of Yangtze River estuary. Combined with correlation analysis and principal component analysis, human driving force statistical analysis of land use change shows that human driving forces can be divided into three main aspects: (1) speed up the process of urbanization, economic development, residents improve living standards; (2) technological progress, population; (3) policy factors, traffic factors. Space-driven analysis showed that the main space affect factors of construction land in the North Branch of Yangtze River estuary are roads, river waters, the city center, the "River-sea linkage" development strategy and so on.Introduced the classic cellular automata SLEUTH model using for simulating urban growth and land use evolution and improved model practicality in the north branch of the Yangtze River estuary through improving the screening methods of Cellular automata control coefficient, adjusting control coefficient model by combining the study of the driving force and setting model calibration parameters in accordance with the actual situation in the north branch of the Yangtze River estuary. Designed Qidong City and North Branch of Yangtze River estuary two different scenarios, simulated and predicted land-use evolution of the two scenarios, analysis and compared of the simulation accuracy of the two scenarios. It is shows that SLEUTH model can accurately predict land-use change, especially for the prediction of the city's expansion, but SLEUTH model did not put socio-economic factors such as the humanities into the model and model calibration process is very time-consuming.Combined with genetic algorithm, artificial neural network and cellular automata, improved and expanded the ANN-CA model, used object-oriented modeling idea, built GANN-CA model based on Matlab, which can make full use of artificial neural network to obtain the variable space parameters and simplify the land use transfer rule. Model considered more comprehensive space impact factors and not only take into account of spatial distance, neighborhood conditions, property, but also give full consideration of land policy, economic development strategy, and take the north branch of the Yangtze River estuary as an example, research shows that the GANN-CA model has better simulation results.There are 60 figures, 49 tables, 98 formulas and 207 references.