基于MAS-LCM耦合模型的孟连县土地利用时空变化模拟预测研究

发布时间：2018-01-04 20:39

  本文关键词：基于MAS-LCM耦合模型的孟连县土地利用时空变化模拟预测研究　出处：《云南大学》2016年硕士论文　论文类型：学位论文

  更多相关文章： 多智能体系统模型(MAS) Logistic-CA-Markov模型(LCM) 孟连县 土地利用时空变化 模拟预测

【摘要】：随着人类活动范围的扩展,特别是不合理人类活动导致的土地利用变化,不利于土地系统的可持续发展。为了更好地引导人类行为,改变不合理的土地利用活动,对土地利用模拟预测研究就显的很重要。孟连县近十多年来人工经济园林种植面积快速增长,取代了相当面积的耕地和天然林地,这些经济活动的干扰使孟连县土地利用结构和空间布局发生了剧烈的改变,引发更多的经济、社会和生态过程的变化。因此,有必要模拟预测孟连县的土地利用时空变化,以便积极应对可能出现的问题。本研究以大面积人工经济园林种植区孟连县为研究对象,采用遥感数据、社会经济统计数据、调查访谈问卷数据,分析孟连县土地利用变化特征。在此基础上,采用Logistic-CA-Markov模型和MAS-LCM耦合模型,以孟连县2015年土地利用现状为基期年数据,模拟预测孟连县2025年土地利用时空变化,并对预测结果进行分析和比较。结论如下：(1)土地利用变化特征。①孟连县近10年土地利用以林地、耕地和经济园地为主,10年间林地、耕地和水域面积减少,经济园地、建设用地和其他用地面积增加,其中林地面积减少最多,经济园地面积增加最多。②10年间孟连县土地利用类型流向主要发生在林地、耕地和经济园地之间。(2)采用二元logistic回归方法确定孟连县土地利用时空变化的驱动因子。耕地、建设用地、林地、经济园地、其他用地和水域的ROC值都大于0.7,表明所选的坡度、坡向、海拔、距离河流距离、乡镇农业劳动力数量、距离主要道路距离、距离农村居民点距离、距离乡镇距离和距离县城距离9个驱动因子分析各地类空间分布适宜性概率是合适的。(3)基于Logistic-CA-Markov模型的2025年孟连县土地利用时空变化模拟预测。① Logistic-CA-Markov模型模拟精度的有效性检验,得到的Kappa系数是0.8598,模型模拟的准确性很高,模型可以使用。②孟连县2025年土地利用时空变化模拟预测结果与2015年土地利用现状相比,耕地、林地和水域面积减少,经济园地、建设用地和其他用地面积增加,尤其是经济园地快速扩张；孟连县西部的勐马镇、公信乡和富岩乡大量的耕地和林地被占用,预测结果与孟连县土地自然社会条件影响下的土地利用变化趋势一致。(4)基于MAS-LCM耦合模型的2025年孟连县土地利用时空变化模拟预测。①在MAS模型构建中,耕地、建设用地和林地主要受到基本农田保护区、建设用地空间管制分区和林地利用保护规划三个土地利用政策影响,经济园地近5年主要由耕地流入,受农户种植行为影响最为明显。设定农业部门、国土部门、林业部门和农户四类土地利用智能体,农业部门、国土部门和林业部门以相关规划为行为规则,分别对应选择耕地、建设用地和林地的种植行为,农户以农业劳动力数量和种植经济效益高低为行为规则,判断经济园地由耕地转入的种植行为。②孟连县2025年土地利用时空模拟预测结果与2015年土地利用现状相比,经济园地、耕地、建设用地和其他用地面积在增加,林地和水域面积在减少,基本农田和二级保护林地得到保护,建设用地在规划设定范围内有序发展；土地利用变化在空间上主要体现在勐马镇、公信乡、富岩乡和娜允镇四个地方,预测结果符合孟连县土地自然社会条件影响下和地利用主体行为选择下的土地利用时空变化趋势,与实际状况更相符。(5)两种模型模拟预测结果对比。MAS-LCM耦合模型模拟预测结果与Logistic-CA-Markov模型模拟预测结果相比,耕地和林地面积增加,建设用地和经济园地面积减少,其他用地和水域面积几乎相同。在MAS-LCM耦合模型模拟预测下,经济园地种植面积增长的趋势放缓,建设用地也在有序稳步发展,特别是基本农田和二级林地保护区避免了被经济园地和建设用地占用,说明MAS-LCM耦合模型的模拟预测比Logistic-CA-Markov模型的模拟预测更有利于孟连县土地可持续发展。同时也说明在Logistic-CA-Markov模型基础上,通过基本农田保护区、建设用地空间管制分区和林地利用保护区三种政策的设定,特别是农户以农业劳动力数量和种植经济效益高低选择经济园地种植行为,土地利用模拟预测结果更符合孟连县现实状况。MAS-LCM耦合模型在土地利用的自然社会经济背景环境下,阐明土地利用主体对环境反应机制,科学分析土地利用主体行为,把影响土地利用主体行为的人文因子添加到模型中,符合土地利用变化的自然性、社会性和人文性特征,对提高模型模拟预测结果的科学性有重要作用。
[Abstract]:With the expansion of human activities, especially the unreasonable human activities lead to the change of land use, is not conducive to the sustainable development of land system. In order to better guide human behavior, change the unreasonable land use activity is very important for land use simulation and prediction of explicit. In Menglian County, nearly ten years of artificial economic landscape planting the area of rapid growth, instead of a considerable area of arable land and natural forest, interfere with the economic activity to Menglian County land use structure and spatial layout of a dramatic change, lead to more economic, social and ecological change process. Therefore, it is necessary to predict the spatio-temporal change of land use in Menglian County, in order to actively respond to possible problems. In this study, a large area of artificial economic landscape planting area in Menglian County as the research object, using remote sensing data and social economic statistical data, survey interview On the questionnaire data, the analysis of land use change characteristics of Meng Lian county. On this basis, using Logistic-CA-Markov model and MAS-LCM model, with the status of land use in Meng Lian County in 2015 for the base year data, simulation and prediction of spatial and temporal changes of land use in Meng Lian County in 2025, and the forecast results are analyzed and compared. The conclusions are as follows: (1) land use the variation characteristics of Meng Lian County in recent 10 years of land use in woodland, farmland and garden land based economy, 10 years, cultivated land and water area decreased, increasing the economic field, land and other construction land area, the forest area decreased most economic garden area increased most. The land use types to 10 years of Meng Lian county mainly occurs in woodland, between cultivated land and economic field. (2) by two yuan logistic to determine the Meng Lian county land use change in the driving factor regression method. The cultivated land, construction land, Economic forest, garden, other land and water ROC value is greater than 0.7, showed that the selected slope, slope, elevation, distance from the river, the township agricultural labor force, from the main road distance, distance of rural residential distance, distance and distance from the County Township from 9 driving factor analysis suitability probability the space distribution is appropriate. (3) simulation and prediction of temporal and spatial variation of land use in 2025 in Menglian County, based on the Logistic-CA-Markov model. The validity of Logistic-CA-Markov model simulation accuracy, 0.8598 Kappa coefficients, the accuracy of model simulation is high. The model can be used in Menglian County, 2025 land use change simulation results with the 2015 land use status compared to farmland, woodland and water area, economic field, increase land and other land area construction, especially in the economic field quickly Expansion; Menglian County West of mengma Town, township public trust and rich Rock Township large amount of arable land and forest land is occupied, the prediction results and effects of land natural and social conditions in Menglian County, favorable land use change trend. (4) land in 2025 in Menglian County MAS-LCM coupling model with spatial variation in the prediction based on simulation. MAS model, farmland, construction land and forest land is mainly affected by the basic farmland protection areas and construction land space control zoning and forest conservation and utilization planning three land use policy, economic field for nearly 5 years, mainly by land inflow, by farmers planting effect is most obvious. Set the Department of agriculture, Department of homeland. The forestry departments and farmers, four types of land use agent, the Department of agriculture, Department of land and forestry departments related to planning for the corresponding selection rules, cultivated land, planting behavior and woodland construction, farmers in agricultural The amount of labor and planting industry in economic level as a code of conduct, to judge the behavior of economic field by planting cultivated land into. The 2025 Menglian County land use spatio-temporal simulation results and 2015 land use status in the economic field, land, and other land uses in the area to increase the construction of forest and water area decreased in basic farmland and two protected forest protection, set construction land within the scope of planning and orderly development; land use change in space is mainly reflected in the mengma Town, township public trust, Fu Yan Xiang and Nayun town four place, the prediction result accords with the land use change trend under the choice of subject behavior with influence in Menglian County the land of natural and social conditions and, more consistent with the actual situation. (5) two kinds of model prediction results and Logistic-CA-Markov model comparison of.MAS-LCM coupled model simulation results The prediction results compared to cultivated land and woodland increased, reduce the area and economic construction land, other land and water area is almost the same. In the simulation and prediction of MAS-LCM coupling model, economic growth slowed the garden planting area, construction land in the orderly and steady development, especially basic farmland and forestland protection level two to avoid the area occupied land economic garden and construction, which shows the simulation prediction of MAS-LCM coupling model is more than the simulation prediction Logistic-CA-Markov model to Menglian County land sustainable development. At the same time also shows that based on the Logistic-CA-Markov model, the basic farmland protection area, construction land space control zoning and forest land reserve policy set three in particular, the number of agricultural labor and planting farmers to the economic level of economic garden planting behavior, simulation and prediction of land use. The fruit is more in line with the actual situation in Menglian County land use.MAS-LCM model in the background of natural and social economic environment, to clarify the mechanism of reaction of land use subject to the environment, scientific analysis of land use behavior, the human factors affecting land use behavior is added to the model in accordance with the land use change of natural, social and cultural the characteristics of the simulation, scientific prediction results have an important role to improve the model.

【学位授予单位】：云南大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F301.24
，

本文编号：1380030