Building And Simulation Of Fuzhou City Land Use Change LCM Model

For some natural and human factors such as the special geographical positionalong the coast of River-sea basins, population increasing sharply, acceleratedgrowth of Urbanization and so on, the scale of Fuzhou city expands continuously,land use contradictions become more and more serious, and regional land use covertakes a lot of changes. Land use change in Fuzhou city is typical in the southeastcoastal city.Therefore the simulation of land use cover change in Fuzhou city isvery important for providing guiding role for the reasonable spatial layout of theurban development/the effective development and utilization and management of theregional land resource/the protection and government of the ecological environment,also for providing the basis for land utilization, management and the formulationof land use policy in Fuzhou city.Took Fuzhou city as the research area, in the ENVI and Arcgis software,extracted three periods of Fuzhou city remote sensing image information in2000/2005and2010, got three periods of land use remote sensing interpretation maps andanalysed them. Used Cramer'V index to analyze the influence factors of land usechange. On the basis of them, modeled the LCM, and used it to simulate the Fuzhoucity land use change dynamically in2010and2015in Idrisi15.0. Get the followingconclusions:(1)Took Fuzhou city land use change research as the example,modeled theintegrated LCM gathered MLP/Markov chain/soft prediction and hard prediction models. Used Multi-layer perception method (MLP) to create transition potential model, theaccuracy of MLP reached70.38%, meeting the requirement of the study, and obtainednine transition potential maps of the paddy field into the water and constructionland, dry land into construction land,forest land into paddy field/dryland/construction land and the road,the water into paddy field and the unused landinto the garden. Then predicted all kinds of change variables of the transitionpotential maps through the Markov Chain model, getting the transition probabilitymatrix.Used the hard and soft prediction model to predict Fuzhou city land usepattern change in2010, the simulation result shows: their AgreementQuantity/DisagreeQuantity and AgreementChance is the same, but the AgreementGridcell of thehard prediction model is less than that of the soft prediction model, itsDisagreeGridcell is more than that of the soft prediction model. And the four kappaindexes of the soft prediction are as follows: Kno=0.9223,Klocation=0.9063,KlocationStrata=0.9063,Kstandard=0.8657; that of the hard prediction modelare relatively0.9178/0.8980/0.8980and0.8579.Therefore the accuracy of the twokinds of the models' prediction results is relatively higher, and they are almostconsistent with the actual land cover maps with small difference, but it is necessaryto improve the AgreementQuantity. And the accuracy of the soft prediction is higherthan the hard prediction, generating the biodiversity vulnerability distributionmap, which is very important to provide the basis and foundation for the subsequentanalysis of the biodiversity and the evaluation of the specific species habitatdistribution. And the ROC result is84.70%, it shows the location distribution effectin the prediction map is comparatively well. So the research shows that LCM modelhas good simulation result.(2) This study selected four spatial variables: distance from the citycenter/distance from the road/distance from the human disturbance and distance fromthe water; two natural attribute variables: slope and elevation; one policy variable:land use policy; one intermediate variable: the possibility of conversiondistribution. Distance from the water/the possibility of conversion distribution/slope and elevation were set as the static variables, and distance fromthe city center/distance from the road and distance from the human disturbance wereset as dynamic variables. In the study of Fuzhou city land use change, not onlyconsidered the social economic environment factors, but also made full considerationon the policy and other factors.The Cramer' V of distance from the city center was0.1711, and the Cramer'V of distance from the road/distance from the water/distance from the humandisturbance and the possibility of conversion distribution was relatively0.1220/0.0865/0.0515and0.3785. So the possibility of conversion distribution hasthe strongest explanation, and followed by distance from the city center/distancefrom the road/distance from the water and distance from the human disturbance. Slopeand elevation is very important to change land use, and land use policy influencedon the simulation of the land use change as encouragement or restrictive variable.(3) In the research area, forest land and paddy field is the main land use types,and then is the water/construction land/dry land/unused land/garden and the road.With the acceleration of Fuzhou urbanization and the implementation of "eastwardexpansion into south, along the river-sea to development"/land consolidation anddevelopment and the related policy, on the whole, land use pattern presents theincreasing trend on the construction land/garden and the road; paddy land/dryland/unused land and forest land continue to decrease; the water basically take nochange. Through calculation and analysis of Fuzhou city land use dynamic change,learned it: from the year of2000to2005, land use dynamic degree of constructionland is largest, garden is the second largest, then followed by dry land/thewater/the road/paddy fields/unused land and forest land. From2005to2010, thechange degree of construction land and garden is still the top two, but the roadchanges very dramatically, unused land change is relatively large, next is dryland/paddy field/the water and forest land.(4)Used LCM to predict Fuzhou city land use change in2015.Result shows: in theland use pattern, the construction land/the road and other land types of Fuzhou city expand in the east and south along Minjiang river, but the north and west ofthe study area restrict the development of all kinds of the land use types becauseof Gushan mountain/Qishan mountain and so on, which is accord with the developmentplanning of Fuzhou.In the land use structure,Fuzhou city in2015keeps the area of constructionland/garden and the road increasing, and dry land increases in a certain extent.Theirpercentage of the total land area are respectively by11.38%/2.36%/1.34%and2.87%in2010to by15.19%/2.93%/1.89%and3.52%in2015. The area of the constructionland increases from43947.1149hm~2in2010to58672.9181hm~2in2015.The area of thegarden increases2201.0777hm~2,the road area adds7284.34hm~2;but paddy field/forestland and unused land continue to decrease, their percentage of the total land areaare respectively by60.33%/11.29%/2.02%in2010to by55.05%/10.97%/1.89%in2015,and decreases respectively20390.5386hm~2/1249.2208hm~2/1300.31hm~2;the waterchange a little.This states the area of construction land and the road substantial increasebecause of the development of the social-economy/growth of the population and theacceleration of the urbanization in Fuzhou city, but the area of the paddyfield/forest land is decreasing due to the agricultural structure adjustment andoccupied by construction. This is also reflected that it should be intensive andeconomical to use land, protect paddy fields/dry land and other agricultural land,and should also strengthen forest land protection,keep the characteristics of"Green Fuzhou" and realize sustainable unity between land use and the ecologicalenvironment.