基于Markov模型的松嫩平原西南部土地盐碱化预测研究  

RS and GIS Supported Simulation and Prediction on Soil Salinization in Southwest Songnen Plain Using Markov Model

作　　者：李建平[1] 赵江洪[2] 张柏[1] 王宗明[1] 宋开山[1] 郑树峰[1] 

机构地区：[1]中国科学院东北地理与农业生态研究所 [2]北京建筑工程学院,北京100044

出　　处：《农业系统科学与综合研究》2006年第4期264-267,共4页System Sciemces and Comprehensive Studies In Agriculture

基　　金：国家自然科学基金项目(40401003)

摘　　要：摘要：以松嫩平原西南部土地盐碱化典型区之一的吉林省大安市为研究对象,该区土地盐碱化的主要动态变化因子是草地、耕地、水域、林地与盐碱地之间的转化,采用时间尺度为12年的1989年和2001年两期遥感影像,结合野外调查、各种图件和文字资料,利用GIS软件人机交互式解译提取大安市盐碱地、草地、耕地、水域、林地信息,获取统计数据,并以此为基础利用马尔柯夫模型预测和分析大安市未来10年土地利用格局的演化。结果表明。在保持当前人为干扰条件不变的情况下,2010年该区盐碱化土地面积占上述5种土地利用类型总面积的比例将达到38．67％,2015年将达到40．75％,土地盐碱化严重。因此,这一问题不容忽视,必须采取有效措施,维护该区生态环境的健康发展。图2,表4,参12。Daan city in Jilin province is one of the most typical soil salinization regions in southwest Songnen plain. The dynamic change factor of saline-alkali soil is mutual transition among grassland, cultivated land, water area, forest and saline-alkalized land in the region. By using TM image in 1989 and ETM ＋ image in 2001 from American LANDSAT satellite and integrating field surveying, all kinds of maps and text datum, soil salinization information for Da＇an city are extracted by man-computer interactive interpretation method in AreView and Arcinfo GIS software. Based on the statistics data acquired, development trend of land use types in the next 15 years is forecasted and analyzed by Markov model. The results indicate that the problem of soil salinization in study area is quite severe. If the current development trend is kept, the area of saline-alkalized land in Daan city will be 36.99% of total area of five above-mentioned land use types in 2010, and 40.75% in 2015. Saline-alkalized land will be the dominant. So the problem must be emphasized, and effective measures should be taken to maintain healthy development of environment.

关 键 词：遥感 地理信息系统 土地盐碱化 影像解译 马尔柯夫模型 

分 类 号：P208[天文地球—地图制图学与地理信息工程] F301.24[天文地球—测绘科学与技术]