【作者】 张国平；

【导师】 刘纪远； 张增祥；

【作者基本信息】 中国科学院研究生院（遥感应用研究所） ， 地图学与地理信息系统， 2002， 博士

【摘要】 本文采用遥感和GIS技术，以遥感调查的土壤风力侵蚀数据、土地利用及其动态变化数据、中国土壤数据、中国气象数据、数字高程数据、植被指数和各种相关统计数据为基础，引用气候学、统计学、地统计学、GIS技术方法，对风蚀过程和强度、沙尘暴发生的下垫面因素、风蚀灾害、土地沙化、风蚀作用驱动力、风蚀发展趋势分析和治理对策等进行分析。在开展多因素详细分析的基础上，对各因素的关系和发展趋势进行分析，从而系统性地开展基于遥感和GIS的风力侵蚀研究。论文主要包括以下内容： 一、对中国土壤风力侵蚀的驱动力进行分析：通过遥感方法建立土壤风力侵蚀的分级标准，对全国土壤风力侵蚀的空间分布进行格局分析，从宏观意义上建立了不同侵蚀程度的空间格局。在全国范围内，对影响土壤风力侵蚀的主要因子，如：风场强度、土壤含水量、植被覆盖、土壤质地、地表起伏等进行空间分析和基于土壤风力侵蚀的知识挖掘。利用主成份分析方法，提取这些因子中的信息，生成土壤风力侵蚀动力指数，其与土壤风力侵蚀现状有很好的可对比性，可以反应上壤风力侵蚀的主要影响因子、发展方式，动力差别和土壤风力侵蚀强度，对沙漠化治理有直接指导作用。对全国风力侵蚀的驱动过程进行分析，并由此分析全国意义上的风蚀治理方案。 二、对影响土壤风力侵蚀的因子进行空间描述：分别对风场强度、土壤可侵蚀性因子、土壤结层状层状因子和积雪因子进行详细分析。 利用中国土壤数据库、中国气象数据、中国资源环境数据库等人量数据，并利用已有模型中的方法对这些因子进行空间数字化。分析它们的空间分布规律，并从控制每个因子值的角度分别探讨有关风蚀的人工防治措施。 三、村沙尘暴发生的过程进行分析：沙尘暴可分为：新疆南部区，新疆东部区，青藏高原区，陕、甘、宁、蒙区和东北区这5个区域，区域总面积达440万km~2影响我国沙尘天气存在西北路径、西方路径和北方路径3个方向，其中以西北路径为主，西方路径次之，北方路径最少。且随着土壤侵蚀程度的加深，沙尘暴日数也在增加。供沙区主要以裸地为主，占59.5％。但有近75万km~2的草地发生风力侵蚀并为沙尘暴提供物质来源，占总物质来源地的30.4％，占全国草地面积的38％。同时有5.2％的耕地发生侵蚀且为沙尘暴提供物质来源。在中国北方地区的干旱背景下，年均沙尘暴的发生频率与年均风速的相关性较强。分析结果表明了近10年来北方大部分地区的年均风速于快速下降过程，其变化幅度大于气温和降水等的变化，在这种趋势下可能会导致沙尘暴发生频率减少。近50年来多个站点的观测记录表明了沙尘暴的发生频率处于快速下降的过程，这恰好与年均风速的变化趋势较为一致。大量草地变为耕地或直接转化为沙化草地，风蚀区的耕地面积扩展了，这均会导致沙尘暴供沙区面积增大，可能会使沙尘暴的强度和频率增大。如果沙尘暴观测数据确实可靠，那么如果人为活动得到有效控制，在目前的干旱半干旱区格局下，大量治理工程对控制或抑制沙尘暴所产生的作用就需要给予新的评价。四、对近10年来中国绿洲区的耕地变化和土地沙化进行分析：利用中国资源环境数据库中遥感调查的土地利用数据，对90年代至2000年中国绿洲土地利用变化进行分析，对绿洲的耕地扩展和土地沙化关系进行探讨，分析中国绿洲区沙地与耕地的时空演化特征。五、对西藏地区的土壤风蚀沙化过程进行分析：结果表明了西藏沙地处于逆向变化过程，沙地向林地转化。这种转化是多年来沙化土地治理的结果。而80年代以来的沙漠化气候因于值偏小，表明了自然因素所导致的土地沙化趋势在减弱。而几乎所有沙地都没有明显扩张现象，这可能与CR因子值直接相关。在目前的风速和降水和温度等条件下有利于沙地的固定。几乎所有的减少的沙地均分布于西藏中部“一江两河地区”，该地区的沙地占沙地总面积的29．5％，西藏西北部和南部的沙地没有明显的缩小。并且，在“一江两河地区”，沙地的减少量只点该地区沙地总面积的1．7％，沙地总体上减少的规模不明显。六、对风力侵蚀的发展趋势进行分析：上地利用变化和气候因素变化是驱动上壤风力侵蚀过程的两个关键因子。分析对风蚀产生重大影响的土地利用变化及其规模，探讨在人类目前的土地利用程度下土壤风力侵蚀从强度上和空问分布上的发展方向。利用 1951至 2000年的气象观测数据，对相关气象要素进行时间序列分析，探讨每个因子的变化规律和趋势。利用气温、降水和风速这三个因子计算沙漠化气候因子的值，并对其进行时间序列分析，探讨中国风蚀沙漠化的发展趋势。更多还原

【Abstract】 Based on remote sensing (RS) and GIS technique, with the method of climatology, statistics, Geo-statistics, and GIS, this dissertation analyze the process and intensity of wind erosion, wind erosion hazard, sandy desertification, driving forces of wind erosion, trend of wind erosion in the future and possible countermeasures of wind erosion in China using soil wind erosion data through RS investigation, land use and land cover data, soil data, meteorological data, digital elevation data(DEM), vegetation index, and various statistical data, also, the main driving factors of sand dust storm and their spatial-temporal changes is analyzed using multi-factor analysis.The content of this dissertation are as following:l.the analysis of driving forces of soil wind erosion in China: The establishment of spatial patterns of different erosion intensity with the analysis of spatial distribution of remote sensing derived soil wind erosion in China. The analysis of spatial patterns of factors that affect soil wind erosion, such as wind field intensity, soil water content, vegetation index, soil texture, earth surface slope, is carried out through data mining theories of GIS. Soil Wind Erosion Dynamic Index(SWEDI) is set up using the method of principle component analysis. SWEI has a close relation with current soil wind erosion data in China, thus can reflect the main influential factors, developing trend, dynamic mode and wind erosion intensity, and has a direct guidance in the control of sandy desertification.2.The description of spatial distributions of factors that influence soil wind erosion. With different models, the factors of wind field intensity, soil erodibilty, soil crust and snow cover is digitalized through soil database, weather data, and resource and environmental data of China. The spatial distributions of each factors are analyzed, and the sandy desertification control measures concerned with each factor are studied. 3. The analysis of the development of sand dust storm in China: Five regions with sand dust storm can be classified with a total area of 440 X 104km2. Three possible trajectories exist for sand dust storm that influences China. Sand dust storm frequency increases as the aggravation of soil wind erosion. In China, regions as the sand sources of dust storm are mainly bare-land, which account for 59.5% of total area of sand source area. However, 70 X 104 km2 of grassland became sources of sand dust because of soil wind erosion, which account for 30.4% of all sand dust sources, and 38% of all grassland in China. 5.2% of farmland are influenced by wind erosion and became sand sources of dust storm . The frequency of sand dust storm has a close relation with meanannual wind speed in the arid, and semi-arid area in China. Analysis show a sharp decrease in annual wind speed in most part of North China in recent ten years which is more obvious than that of temperature and precipitation, and this may lead to the decrease in sand dust storm frequency. Meteorological records in the past 50 years indicate that there has been a sharp decrease in sand dust storm frequency, which is in accordance with the decrease in annual mean wind speed. The transition of grassland to sandy land, or the expansion of farmland can all leads to the expansion of sand dust sources, which in turn will increase the intensity and frequency of sand dust storm. If observation data are credible, and if human activities can be controlled effectively, then under current climatic circumstance, the efficiency of many erosion control projects should be re-evaluated.4.The analysis of spatial-temporal changes in farmland and sandy desertification in oasis in China in recent lOyears: Land use dynamic changes in oasis in China in 1990-2000 are analyzed. The relationship between farmland expansion and sandy desertification is discussed.5. The analysis of wind erosion in Tibet: Results show that sandy land in Tibet is shrinking reflecting the transition of sandy land to wood land. This is the result of years of sandy land manag更多还原