Knowledge Resource Center for Ecological Environment in Arid Area
| 科尔沁沙地土地退化遥感监测研究 | |
| 其他题名 | STUDY ON LAND DEGRADATION MONITORING IN HORQIN SANDY LAND USING REMOTE SENSING |
| 杜子涛 | |
| 出版年 | 2009 |
| 学位类型 | 博士 |
| 导师 | 王长耀 |
| 学位授予单位 | 中国科学院遥感与数字地球研究所 |
| 中文摘要 | 土地资源是人类进行一切社会经济活动及其赖以生存和发展的物质基础，是一种不可再生的最基本的资源，它的合理利用是人类生存和社会经济发展的基础。但是近几十年来，由于人口急剧增长带来的食物需求以及人类不合理的开发利用土地，土地资源不断退化，严重威胁到人类社会的可持续发展。我国是世界上土地退化比较严重的国家之一，土地退化问题已严重影响到人们的生存环境和社会经济的可持续发展。鉴于土地退化给农业生产、经济发展带来的巨大危害，本次论文以“国际合作项目《荒漠化评估和监测系统(A Surveillance System for Assessing and Monitoring of Desertification)》”为依托，选取我国北方农牧交错区的科尔沁沙地为研究区，以TM和MODIS遥感数据以及详尽的地面调查资料为数据源，综合运用RS、GIS、GPS技术，进行土地退化监测与评价方法的研究。\n论文的主要研究成果和创新点包括：\n1、在前人研究成果的基础上，本研究以土壤调节植被指数(MSAVI)和阴影指数(SI)为基础，借鉴ITTO(International Tropical Timber Organization)提出的FCD (Forest Canopy Density Model) 模型，构建了一个新的复合植被指数，该复合植被指数能够有效地削弱土壤、阴影等背景信息的影响。\n2、评价了遥感反演方法的适用性，实现了土地退化指标因子的遥感反演。\n土地退化指标因子的遥感反演是建立土地退化评价体系的基础，本文选取了土地利用/覆被方式、复合植被指数、风蚀土地占像元面积百分比和土壤湿度作为土地退化评价体系的四个指标因子，以奈曼旗为试验区，利用TM数据获取了四个指标的反演数据。其中，土地利用/覆被方式采用支持向量机分类方法，风蚀土地占像元面积百分比应用线形光谱混合模型进行遥感反演。\n3、建立了基于遥感的土地退化监测指标体系和综合评价方法。\n本文使用层次分析法构造了四个指标因子的判断矩阵，并通过判断矩阵特征向量的计算，得到了土地退化指标因子的权重，构建了土地退化综合评价公式。结合综合评价指标公式，计算出奈曼旗的土地退化结果，将土地退化分为未退化、轻度退化、中度退化、重度退化和严重退化五个等级。1987年、1991年、1996年、1999年、2002年和2007年的土地退化评价结果表明，奈曼旗的土地退化在1987-1999年期间处于“局部好转、整体恶化”的局面，2000年以后土地退化扩展趋势得到遏制，土地退化治理取得了良好的效果。这和当地的气候、经济等有一定联系，更和国家在奈曼旗实行的土地退化治理措施密切相关。2000年国家启动京津风沙源治理工程，奈曼旗在国家的支持下开始实施“退耕还林还草”工程，全旗开始实施有计划的分时、分地的禁牧尝试，使生态建设的速度由“局部好转、整体恶化”转向“治理与破坏速度相持”。2002年初奈曼旗实现了全旗禁牧，使大自然得以休养生息、自我修复。加上国家对生态建设投入的加大，农牧民的税赋负担的减轻，之后又有退耕还林补助、种粮补贴等，农牧民的生活有了保障，植被恢复开始加快。\n4、Sayizky-Golay平滑滤波法能得到高质量的EVI时序数据。\n最大值法(MVC)合成数据也会受到云的影响，仍存在异常值或缺失数据等噪声。本文采用Sayizky-Golay滤波器对16天合成MODIS-EVI时序数据进行平滑滤波处理，该方法采用EVI数据的上包络线来拟合EVI时序数据的变化趋势，通过迭代过程得到高质量的EVI时序数据。经过试验表明，该滤波方法能够有效的去除受云和大气条件影响的异常值，并能最大限度的保存原时间序列中的有效信息。\n5、利用Sayizky-Golay滤波平滑处理后的2000-2008年9年16天合成的空间分辨率为250m的MODIS植被指数产品，采用均值法、差值法、变化矢量分析方法和一元线性回归模拟法，分析了科尔沁沙地的植被年际变化特征，并模拟了9年EVI年最大化值的变化趋势线和增长幅度。最后，利用EVI与植被盖度的高相关性，以EVI为土地退化评价的定量依据来评价科尔沁沙地土地退化状况及动态变化。研究结果表明，近9年来科尔沁沙地植被覆盖总体上表现为稳中增加的趋势，局部区域有植被退化的趋势，但退化幅度小于改善幅度；对比研究2000与2008年不同等级退化土地面积所占比例发现，科尔沁沙地土地退化总体上呈现减少趋势，部分轻度和中度退化土地经过治理变为有利用价值的非退化土地，而严重退化土地由于治理难度大，可利用价值少，治理效果不是很明显。 |
| 英文摘要 | Land resource is the basic material for social activities, living and development of human beings. It is nonrenewable, so its reasonable utilization is the foundation for survival and economic development of human beings. While in recent decades, land has been degrading and threatening the sustainable development of human society because of significant population growth and unreasonable utilization of land resources. China is one of countries with serious land degradation. Land degradation has become so serious that environment and sustainable development of social economy have been impacted. As land degradation causes great disaster to agriculture and economic development. With Horqin Sandy land as a focus, this paper studies land degradation monitoring and evaluation with RS，GIS and GPS through TM and MODIS on the basis of A Surveillance System for Assessing and Monitoring of Desertification.\nThe major results and innovation of the study include: \n1. On the basis of MSAVI and SI and previous study achievement, with FCD model in ITTO as a reference, this study establishes a new VI which can effectively reduce the effect of soil and shadow.\n2. It evaluates the applicability of remote sensing retrieval methods, and land degradation indexes are obtained through remote sensing retrieval. \nLand degradation index retrieved through remote sensing images is the foundation of land degradation evaluation system. Four land degradation indexes, i.e. land use and land cover change, new vegetation index, sandy area ration in a pixel and soil moisture, are selected. With Naiman county as study area, retrieval data of four indexes is obtained with TM image. Land use and land cover change is retrieved with support vector machine method and sandy area ratio in a pixel is retrieved with line spectral mixture model. \n3. Remote sensing land degradation indexes and comprehensive evaluation method is established in the study. \nThis paper constructed the judge matrix of four indexes using analytical hierarchy process, calculated the eigenvector of judge matrix, and obtained weights of five indicators. The model of land degradation estimation system had been built with four weights. Land degradation of Naiman county is classified into no degradation, slight degradation, medium degradation, bad degradation and serious degradation 5 levels with evaluation formula. Land degradation evaluation of 1987, 1991, 1996, 1999, 2002 and 2007 demonstrates in that the land degradation treatment of Naiman county is locally improved, but wholly worsened during 1987-1999, while the land degradation started to have improvement since 2000. It is due to the local climate and economy, in particular to land degradation treatment policy implemented in Naiman. The state started Jing-Jin sand source treatment project from 2000, and Naiman county started Conversion of Cropland to Forest and Grassland Project. The county implemented grazing-banning according to time period and land position in a planned manner. As a result, the ecology is shifted to a balance of destroy and treatment from “local improvement and overall worsening”. Grazing banning was implemented in the whole county, and the nature gained recovery at the beginning of 2002.The life of peasants and herdsmen is improved and revegetation is accelerated because of more investment to ecology establishment, tax cuts, subsidy for conversion of cropland to forest and subsidy to crops.\n4. High quality EVI time serial data can be obtained with Sayizky-Golay smooth filtering method.\nMaximum Value Composite (MVC) is also influenced by clouds, so it also has noise like abnormal value or missing data. This method was developed to make data approach the upper EVI envelop and to portray patterns of EVI change through an iteration process. The method was tested with a 16-day MVC MODIS EVI time serial data. It is demonstrated in the study that the filtering method can effectively eliminate abnormal value and missing data impacted by clouds and atmospheric conditions, and reserve effective information in original time serial to the greatest extent. \n5．After Sayizky-Golay smooth filtering treatment, 16-day MODIS vegetation index data of 250 m spatial resolution from 2000 to 2008 is processed with averaging, difference, change vector analysis and simple linear regression to analyze internal change trend of Horqin sand land and simulate change trend and increase range of maximum EVI in 9 years; Land degradation and dynamics of Horqin sand land is evaluated with EVI qualitative method on the basis of high correlation of EVI with vegetation coverage. It is showed in the study that the vegetation in Horqin sand land showed overall stable improvement, local degradation, and improvement more than degradation in the 9 years. It is found in comparison of degraded land percentage in 2000 and 2008 that Horqin sand land degradation has overall decrease, some land of slight and medium degradation is converted into useful non-degraded land, while land of serious degradation does not have significant improvement because of great difficulty in treatment.\n\nKeywords: remote sensing, Horqin sandy land, land degradation, land use and land cover change, analytical hierarchy process (AHP), Moderate Resolution Imaging Spectroradiometer (MODIS), Enhanced Vegetation Index (EVI) |
| 中文关键词 | 遥感 ; 科尔沁沙地 ; 土地退化 ; 土地利用/覆被方式 ; 层次分析法 ; 中分辨率成像光谱辐射计 ; 增强型植被指数 |
| 英文关键词 | remote sensing Horqin sandy land land degradation land use and land cover change analytical hierarch |
| 语种 | 中文 |
| 国家 | 中国 |
| 来源学科分类 | 地图学与地理信息系统 |
| 来源机构 | 中国科学院遥感应用研究所 |
| 资源类型 | 学位论文 |
| 条目标识符 | http://119.78.100.177/qdio/handle/2XILL650/286809 |
| 推荐引用方式 GB/T 7714 | 杜子涛. 科尔沁沙地土地退化遥感监测研究[D]. 中国科学院遥感与数字地球研究所,2009. |
| 条目包含的文件 | 条目无相关文件。 | |||||
| 个性服务 |
| 推荐该条目 |
| 保存到收藏夹 |
| 导出为Endnote文件 |
| 谷歌学术 |
| 谷歌学术中相似的文章 |
| [杜子涛]的文章 |
| 百度学术 |
| 百度学术中相似的文章 |
| [杜子涛]的文章 |
| 必应学术 |
| 必应学术中相似的文章 |
| [杜子涛]的文章 |
| 相关权益政策 |
| 暂无数据 |
| 收藏/分享 |
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。
