【作者】 孙永军；

【导师】 童庆禧；

【作者基本信息】 北京大学 ， 地图学与地理信息系统， 2008， 博士

【摘要】 本论文研究的主要目的是综合利用遥感和地理信息系统技术查明并分析整个黄河流域湿地面积、类型、分布、变化与原因,主要研究内容包括黄河流域湿地数据处理、湿地分类方案、湿地信息提取和综合分析。论文采用的数据、方法如下:数据源包括全国1:50万地质图数据库、全国1:25万地形数据库、1:10万地形图、全国1:50万土地利用数据库、美国航天雷达地形数据(SRTM),美国陆地卫星获取的MSS(1975年)、TM(1990年)和ETM(2000年)卫星遥感数据。经过纸质图件扫描数字化、几何校正配准、镶嵌和分幅,将多元数据统一到同一地理坐标下。根据《湿地公约》分类系统和《全国湿地资源调查与监测技术规程》中的湿地分类方案,结合工作区的实际情况,将黄河流域湿地类型分为5个一级类和15个二级类。采用计算机自动信息提取和人机交互解译相结合方法提取三个时期湿地信息,利用地理空间分析技术提取变化信息,根据监测结果总结分析黄河流域湿地变化特征和原因,其中湿地变化原因分析重点考虑自然和人为两个因素。论文主要成果和创新点:1.提出了黄河流域湿地变迁的动态监测技术方案,指导编制了整个黄河流域1:25万标准分幅湿地遥感解译现状图、两期变化图和MSS、TM、ETM遥感影像图;黄河流域1:150万遥感影像镶嵌图(MSS、TM、ETM)、三期湿地现状图(1975年、1990年、2000年)和两期湿地变化图(1975-1990年、1990-2000年)为黄河流域生态环境综合评价提供了基础图件。2.提出了基于面向对象沼泽湿地信息提取方法,并解决了信息提取中的部分关键技术。根据沼泽湿地是水体、植被和土壤自然综合体的特点,采用基于面向对象分类方法提取沼泽湿地,针对河流湿地在影像上有明显河道信息的特点,采用基于边缘的方法提取河流湿地,上述方法综合利用了湿地的光谱特征和结构特征,取得较好效果,能在一定程度上减少手工操作,提高工作效率,具有推广应用价值。3.初步查明了黄河流域(1975-1990年期间和1990-2000年期间)湿地类型、面积、分布、变化和原因,得到第一手数据,填补了黄河流域近25年来湿地动态研究的空白,得到以下研究成果:(1)黄河流域湿地类型多样分布不均。湿地主要集中在黄河源区、诺尔盖、银川盆地和黄河入海口四处。黄河源区湿地类型主要是高山和冻原湿地,诺尔盖地区湿地类型主要是泥炭沼泽,银川盆地湿地类型较多,并且人工湿地数量明显增加,近海岸湿地则主要分布在黄河入海口。(2)黄河流域的湿地资源1975年为28103.51km~2,1990年为26703.66 km~2,2000年为24235.48 km~2。(3)黄河流域湿地构成按百分比由大到小顺序排列依次为沼泽湿地、河流湿地、湖泊湿地、人工湿地和近海岸湿地。(4)1975-2000年黄河流域湿地面积总体上呈减少趋势,近25年来共减少3868.03 km~2,后期(1990-2000年)的减少速度和幅度明显大于前期(1975-1990年)。(5)湿地减少的原因可分自然和人为两种,特别是人为因素应引起足够重视。更多还原

【Abstract】 The purpose of this study is making use of remote sensing and GIS technique to investigate quantity, type, distribution, change, and cause of wetland in Yellow River Drainage Area. The main content is focused on multiple data processing, wetland classification, wetland information extraction and comprehensive analysis.Data resources consists of 1:500 000 China geological map data base, 1:250000 China terrain map data base, 1:100000 terrain map, 1:500000 China land use data base, DEM of Shuttle Radar Topography Mission(SRTM), MSS(1975),TM(1990)and ETM(2000)acquired by Landsat satellite. These data are consolidated in the same coordinate system through a series of processing including scanning and digitizing of paper map, geometric correction, mosaic and segmentation. Basing on wetland classification method of“The Ramsar Convention on Wetland”and“Regulations of investigating and monitoring of wetland in China”, and also according to the practical condition, the wetland of Yellow River Drainage Area are classified into 5 classes in the first level and 15 types in the second level. Wetland temporal information of three stages was extracted by combining computer automatic information extraction and naked-eye and computer aided interpretation. Wetland change information was extracted by using of the technique of geographic space analysis. Basing on statistic result, dynamic characteristic of wetland is summarized。Two factors including nature and human being are considered in the analysis of the cause of wetland change.The innovation and conclusion of this thesis go as follows:1. Forming the scheme of remote sensing mornitoring of Wetland Dynamic Change in Yellow River Drainage Area, and directoring the making of a series of wetland maps including:Wetland map of standard scope in Yellow River Drainage Area (1:250000, 1975, 1990, 2000),Wetland dynamic map of standard scope in Yellow River Drainage Area (1:250000, 1975-1990, 1990-2000),Remote sensing image map of standard scope in Yellow River Drainage Area (1:250000, MSS, TM, ETM),Wetland map in Yellow River Drainage Area (1:1500000, 1975, 1990, 2000), Wetland dynamic map in Yellow River Drainage Area (1:1500000, 1975-1990, 1990-2000),Remote sensing image map in Yellow River Drainage Area (1:1500000, MSS, TM, ETM).These maps are the basis for the comprehensive analysis and evaluation of ecological environment in Yellow River Drainage Area.2. Proposing the method of swamp extraction basing on object-based classification and solving some key issues of information automatic extraction. Basing on the fact that swamp is made up of water, vegetation and soil, object-based method was used to extract wetland. According to the bank feature, edge-extract technique was used to extracted river wetland. Because both of spectral and structural feature of wetland were used, the above information-extract methods got better result, which can decrease workload, promote work efficiency, and be worthy of applying to other area.3. In some extent, clearing such issue as wetland type, amount, distribution, change and cause in Yellow River Drainage Area, and got the first basis data which replenish the blank of the study of Wetland Dynamic Change in the recent 25 years in Yellow River Drainage Area, and the conclution is as follows:(1) Multiple types and disproportionally distributedWetlands are mainly distributed in four regions including source area of Yellow River, Nuoergai area, Yinchuan basin and Yellow River delta area. In the source area of Yellow River, the main type of wetland is tundra, in Nuoergai area the main type of wetland is turf. The Yinchuan basin consists of varies kinds of wetland, especially there are more artificial wetland, and the offing-sea wetland is mainly distributed in Yellow River delta area.(2) Total amount of wetland in Yellow River Drainage Area is 28103.51km~2 in 1975 year, 26703.66 km~2 in 1990 year and 24235.48 km~2 in 2000 year. (3) Wetland composition proportion decreasingly ranks as swamp, river, lake, artificial wetland and sea-offing wetland.(4) From 1975 to 2000, the area of wetland in Yellow River Drainage Area took on decreasing trend. In about 25 years, the total amount of wetland reduced 3868.03 km~2,and the reduced extent from 1990 to 2000 is more than those from 1975 to 1990.(5) The cause of wetland decreasing includes nature and human being, especially the human being factor play an important rule and is worthy of more attention.更多还原