基于CA-Markov模型的老河口市土地利用格局变化研究

发布时间：2018-08-12 17:33

【摘要】：近年来,土地利用格局的变化随经济社会的快速发展而愈发显著,人地之间的矛盾日益突出。近几年来老河口市在全市人民的努力下经济得到了飞速的发展,随之而来的是农用地面积的减少,粮食产量的降低、生活质量的下降、生态环境的破坏,这与老河口市希冀的可持续发展背道而驰的,这使得土地利用格局的研究就显得格外重要。我国近年来,不论从研究范围还是研究方法、模型上,土地利用／土地覆被研究不断推进,关于土地利用/土地覆被变化模型研究中,CA-Markov模型综合了CA模型和Markov模型两者的优势,拥有十分强大的空间预测能力与预测数量变化能力,因此,在土地利用预测中的应用范围不断扩大。本文以老河口市为研究对象,以该研究区2005年、2010年两个时期土地利用矢量数据成果,模拟预测未来土地利用变化格局。首先,运用ArcGIS10.0软件对两期数据进行各土地利用结构变化分析,继而通过景观动态度、信息熵等指数研究土地利用数量变化；并运用Fragstats4.0软件,分析老河口市在研究期间的景观格局指数的变化及其特征,之后探讨了该研究区土地利用格局演变与自然、社会和经济等影响因素之间的关系,制定出转换规则、土地利用类型的转移矩阵和适宜性图集,在此基础上,利用IDRISI Andes15.0软件中的CA-Markov模型,对2020年老河口市的土地利用格局进行预测,最后提出能够促进该区域可持续发展的措施与意见。本文的研究目的是分析老河口市土地利用的现状以及变化规律,探讨促使这种格局演化的形成及变化的影响因素,形成CA-Markov模型模拟所需要的转移规则,进而对2020年老河口市的土地利用格局进行预测模拟,提出能够促进可持续发展的建议。本文研究的主要结论如下： (1)老河口市土地利用结构格局在2005-2010年研究期间变化显著,面积增加程度最高的是城乡建设用地,其次为交通水利用地。耕地为主的农用地较少程度最高,水域及自然保留地面积也有所减少,但土地利用系统的方向趋于有序。 (2)老河口市景观格局在2005-2010年研究期间变化较为显著,通过两期景观指数的分析,城乡建设用地景观和耕地景观类型的优势斑块作用明显,两者的景观指数反映出景观整体的特性,由于在研究期间城乡建设用地和耕地景观的景观格局趋于多样化和破碎化程度较增加,所以老河口景观格局整体趋于多样化、景观的破碎化程度加深。 (3)为了老河口市2020年的土地利用格局的模拟预测提供转换规则,本文分析影响土地利用格局变化各种因素分析表明,对于老河口市研究区而言,造成老河口市景观格局显著变化的原因主要有自然因素中的坡度、行政中心和交通线路；以及社会经济因素主要包括,伴随着人均收入水平提高过程中非农人口的增长、社会经济的发展以及相关政策法规等因素。基于上述影响因子,本文定性和定量的对各影响因子进行分析,从而为模拟图做充足的准备。 (4)用2010年的实际土地利用格局图与CA-Markov模型模拟老河口市2010年的土地利用格局模拟图作对比,通过Kappa系数来进行精度验证,来检验该模型的预测效果,若Kappa系数较低说明前期数据的收集以及数据的分析存在很大的误差,若Kappa系数较大,精度较高,预测结果有足够的可信度,再来模拟预测老河口市2020年的土地利用格局。 (5)针对老河口市土地利用格局的预测分析结果,本文提出促进可持续发展的几点建议。
[Abstract]:In recent years, with the rapid development of economy and society, the change of land use pattern has become more and more remarkable, and the contradiction between people and land has become increasingly prominent.In recent years, the economy of Laohekou City has developed rapidly with the efforts of the people of the whole city, followed by the reduction of the area of agricultural land, the reduction of grain production, the decline of the quality of life, and the ecological environment. In recent years, no matter from the research scope or the research method, the research on land use / land cover model has been advancing. In the research on land use / land cover change model, CA-Marko V model combines the advantages of CA model and Markov model, and has a very strong ability of spatial prediction and forecasting quantitative change. Therefore, the application of V model in land use forecasting is expanding.
This paper takes Laohekou City as the research object and takes the land use vector data of the study area in 2005 and 2010 as the research object to simulate and predict the future land use change pattern. By using quantitative change and Fragstats 4.0 software, the change and characteristics of landscape pattern index in Laohekou city during the study period were analyzed. Then the relationship between land use pattern evolution and natural, social and economic factors in the study area was discussed. The conversion rules, transfer matrix and suitability map of land use types were worked out. On this basis, the CA-Markov model of IDRISI Andes 15.0 software is used to predict the land use pattern of Laohekou City in 2020. Finally, the measures and suggestions to promote the sustainable development of the region are put forward. The formation of evolution and the influencing factors of change form the transfer rules needed by CA-Markov model simulation, and then forecast and simulate the land use pattern of Laohekou City in 2020, and put forward some suggestions to promote sustainable development.
The main conclusions of this study are as follows:
(1) The land use structure pattern of Laohekou City changed significantly during the study period from 2005 to 2010. The highest increase of land area was urban and rural construction land, followed by traffic and water use land.
(2) Landscape pattern of Laohekou City changed significantly during the study period from 2005 to 2010. Through the analysis of landscape indices, the dominant patches of urban and rural construction land landscape and cultivated land landscape types played an obvious role. The landscape indices of the two types reflected the overall characteristics of the landscape, because of the landscape pattern of urban and rural construction land and cultivated land landscape during the study period. The situation tends to be more diversified and fragmented, so the overall landscape pattern of the old estuary tends to be diversified and the landscape fragmentation deepens.
(3) In order to provide conversion rules for the simulation and prediction of land use pattern of Laohekou City in 2020, the analysis of various factors affecting the change of land use pattern shows that for Laohekou Research Area, the main reasons for the significant change of landscape pattern of Laohekou City are the slope of natural factors, administrative center and traffic routes. And the social and economic factors mainly include the growth of non-farm population, the development of social economy and related policies and regulations in the process of per capita income increase.
(4) Using the actual land use pattern map of 2010 and CA-Markov model to simulate the land use pattern simulation map of Laohekou City in 2010, and verifying the accuracy of the model by Kappa coefficient to verify the prediction effect. If the Kappa coefficient is low, it shows that there is a great error in data collection and data analysis. If Kappa coefficient is low, there is a big error in data collection and data analysis. The coefficient is bigger, the precision is higher, the forecast result has enough credibility, then simulates and forecasts the land use pattern of Laohekou City in 2020.
(5) According to the forecasting results of land use pattern in Laohekou City, this paper puts forward some suggestions for promoting sustainable development.
【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F224;F301.2

【参考文献】