基于CA-Markov模型的吉林省西部土地利用景观格局变化趋势预测  

The Tendency Forecast on Land Use Landscape Pattern Change in Western Jilin Province Based on CA-Markov Model

作　　者：汤洁[1] 汪雪格[2] 李昭阳[1] 毛子龙[1] 韩维峥[1] 徐小明[1] 

机构地区：[1]吉林大学环境与资源学院,长春130026 [2]中国市政工程东北设计研究院,长春130021

出　　处：《吉林大学学报（地球科学版）》2010年第2期405-411,共7页Journal of Jilin University：Earth Science Edition

基　　金：国家自然科学基金项目(40572170;40871088)

摘　　要：Markov模型与CA模型均为时间离散、状态离散的动力学模型,CA模型具有模拟复杂系统空间变化的能力,Markov模型具有长期预测的优势。CA-Markov耦合模型既可进行长期预测,又可有效地模拟空间变化。以吉林西部为例,以RS-GIS技术为支撑进行土地利用遥感解译,得出1989年和2004年两期土地利用景观格局分布图;在此基础上利用CA-Markov耦合模型,进行了土地利用景观格局变化趋势预测。结果表明:与1989—2004年的变化趋势一样,到2019年,水田、旱田、居民地和盐碱地面积分别比2004年增加,高、中、低覆盖草地、水域和湿地面积分别比2004年减少。增加的盐碱地主要是在原盐碱地的基础上扩张,侵占草地、水域和湿地,其分布主要集中在中部的地势低洼地带,大安市、通榆县、长岭县和乾安县的盐碱地将连成片,生态环境被严重破坏。修正的Lee-Sallee指数为0.64,表明CA-Markov模型模拟精度较高,该模型预测土地利用景观格局变化是可行的。Markov and CA model are the dispersed dynamics model with time and state,i e.CA model has the ability of simulated complex systematic space change,and markov model has the advantage of long-term forecast.CA-Markov couple model may not only carry out long-term forecast,but also imitate efficiently the space change.With Jilin west as example,RS and GIS technology had been explored to interpret TM data,and reached landuse landscape pattern distribution picture in the year of 1989 and 2004,then on the foundation,CA-Markov model was applied to carry out the ecological landscape pattern change tendency forecast.The results showed that the changing tendency from 2004 to 2019 was same with the changing tendency from 1989 to 2004.The area of paddy field,dry land,resident land and saline-alkali soil increased respectively,and the area of the higher,lower coverage grassland,waterland and wetland decreased respectively.The increased area of saline-alkali soil expand mainly on the foundation of the saline-alkali soil,and occupy grassland,water area and wetland.The saline-alkali soil concentrate mainly in the low-lying region,the saline-alkali soil in Daan,Tongyu,changling and Qianan will join in piece,ecological environment will be broken seriously.The revised Lee-Sallee index was 0.64,showed that CA-Markov model simulated precision was higher,and it confirmed that model was applicable for prediction of the landuse landscape pattern change.

关 键 词：CA MARKOV 土地利用 景观格局 吉林省西部 

分 类 号：X171.1[环境科学与工程—环境科学]