Font Size: a A A

Research On Land Use Change In Poyang Lake Region Based On Logistic-CA-Markov Model

Posted on:2022-06-17Degree:MasterType:Thesis
Country:ChinaCandidate:Y T WangFull Text:PDF
GTID:2480306557961929Subject:Surveying and Mapping project
Abstract/Summary:PDF Full Text Request
The influence of wetlands on the entire biosphere is self-evident.The Poyang Lake wetland is extremely important in our country's wetland system.Recently,affected by natural changes and human activities,the ecology of the Poyang Lake wetland area has been damaged to a certain extent.The development of remote sensing technology provides a practical approach for wetland research.Using relevant research models and methods to study and predict land use/cover changes will help deepen the understanding of land use structure changes and provide a scientific basis for relevant departments to manage land.In this paper,the Poyang Lake area is selected as the research area to study the change process and regularity of land use in this area on time and space scales,and to predict the future situation.First,interpret the remote sensing images around Poyang Lake in 1995,2000,2005,2010,and 2015.After comparing different classification methods,the classification method with the highest accuracy was selected to obtain the classification data of land use in five periods.The transfer matrix and the center of gravity migration model are used to analyze the law and characteristics of land use change.The distance factor from the protected area was added as a special factor of the study area,and 8 driving factors of nature,social economy and accessibility were selected,the driving forces of land use change were analyzed by Logistic model,and the conversion rules for CA correction from the atlas of suitability were obtained.Based on the two scenarios of natural evolution and ecological environment improvement,the Logistic CA-Markov coupling model was used to forecast the land use situation around Poyang Lake in 2025 under the two scenarios,and the results were analyzed.The research indicates that:(1)The Kappa coefficients of maximum likelihood method,decision tree method and support vector machine method are 0.8855,0.8769 and 0.8979,respectively.The comparison shows that support vector machine method has higher accuracy.(2)From 1995 to 2015,the stability of all land use types around Poyang Lake was relatively low,in which the water area decreased by 181.07km~2,and the area of construction land increased by 1301.77km~2.The degree of mutual conversion between cultivated land and unused bare land was the highest among all land use types.From1995 to 2015,there were significant changes in land types at the spatial level,with significant changes in local migration.(3)The driving force of land use change was explored according to the coefficient obtained from the Logistic regression model,and it was found that each driving factor was mutually restricted,promoted and closely related.The special driving factor of distance from the protected area in this study showed a strong influence effect,and natural factors were the basic factors affecting land use change.Socioeconomic factors have great influence on construction land,and accessibility factors mainly guide the evolution of land use.(4)The Kappa coefficient of the Logistic Ca-Markov model is 0.0428 higher than that of the CA-Markov model,which is more reliable.The forecast results of land use in 2025 show that:According to the laws of nature evolution,construction land growth fast,water body,by contrast,the trees area is reduced,the ecological environment is destroyed,according to the improvement of ecological environment evolution,trend,increase of construction land and water area of trees increased,to alleviate contradiction between human and land,said the implementation of sustainable development planning can help improve ring poyang lake region ecological environment.
Keywords/Search Tags:Poyang Lake Wetland, Remote sensing image, Land use change, Logistic-CA-Markov model, Land use prediction
PDF Full Text Request
Related items