Font Size: a A A

Study On Land Use Change And Its Simulation Based On The MLP-CA-Markov Model In Jinjiang City

Posted on:2015-11-25Degree:MasterType:Thesis
Country:ChinaCandidate:J LinFull Text:PDF
GTID:2309330467961557Subject:Cartography and Geographic Information System
Abstract/Summary:PDF Full Text Request
Land use change is the result of the combination of human factors and natural factors. In recent years, due to human exploit and utilize land and make huge changes in land cover, in turn, the land cover change has a great influence on human activities, the contradiction between human activities and land use is becoming more and more serious. So it makes effective management strategy to provide scientific basis for land use by finding the rule of land use change from different research period, mastering the mechanism of land use change, and simulating projections for the future land use change situation. This paper chooses Jinjiang City as the research example, base on the land use status in2000,2005and2010, Through the change of land use quantity, the change of land use speed, land use change, land use type conversion and the landscape pattern of land use change, even the administration of land use change, traffic, topography and land use planning influence factors are analyzed and discussed, and based on MLP-CA-Markov model to forecast the future land use change situation simulation, main results are as follows:1)It presents variation characteristics that urban and rural construction land and other lands including mining lease,transportation land and other construction land continue to increase, but woodland, farmland and waters reduce continuously, mainly in urban and rural construction land increasing and farmland decreasing are most obvious.2) From the analysis of landscape pattern, the average nearest neighbor distance is more and more small, landscape patch shape index and fractal dimension increase, Shannon diversity index and landscape evenness index have increased, Contagion index and aggregation index are dropped year by year. Description in the process of urbanization, human activity has an important influence on the landscape change.3)By analyzing the factors affecting administrative center, the closer distance between urban and rural construction land administration center, the faster in increasing speed, with the increase of distance, increasing speed decreases. It has a rule that the farther the administrative center, the faster the pace of change in woodland and other lands; both sides of the traffic has inhibitory effect on the distribution of agricultural land, but the distribution of urban and rural construction land and other land; good planning factors control the distribution of the urban and rural construction land, and keep the future development of cultivated land. From the analysis of topographic factors, the appropriate distribution of slope of urban and rural construction land, agricultural land, other land and water level concentrate between0~6°, suitable for the distribution of the forest land is6。and above.4)This study based on CA-Markov model, combine with the MLP, Markov model, implement the forecast simulation of land use status of2010in the studied area. The results show that Kappa coefficient are Kno=0.7879, Klocation=0.8202, KlocationStrata=0.8202, Kstandard=0.7511respectively, not less than75%. It has the high consistency that simulation of the land use condition by CA-Markov model in2010and remote sensing interpretation land use status in2010, high reliability, simulation effect better.5)On the land use type conversion, along with the urbanization, the expansion of the urban and rural construction land in Jinjiang is mainly at the cost of farmland and forest from2010to2020. From the amount of land use change, farmland is the greatest reduction, forest land secondly, and urban and rural construction land increasing is the largest, other land secondly, the water at least from2010to2020.
Keywords/Search Tags:land use change, MLP-CA-Markov, forecast simulation, Jinjiang City
PDF Full Text Request
Related items