【作者】 阿如旱；

【导师】 杨持；

【作者基本信息】 内蒙古大学 ， 生态学， 2007， 博士

【摘要】 中国北方在过去的近50a里，沙漠化以其广泛的分布和迅速发展构成了主要的生态环境问题，已直接影响到当地社会经济可持续发展及威胁着京津经济圈生态安全。内蒙占多伦县位于京津风沙源边缘区，面向京(北京)津(天津)，是内蒙古到北京直线距离最近的旗县，具有典型的生态脆弱性。近100多年来，由于人口剧增，大面积的自然草场被开垦为耕地：尤其是20世纪50年代以来，由于“大跃进”和“文化大革命”时期“以粮为纲”的片面路线，多伦县的滥垦、过牧现象更为严重，可称为受沙漠化威胁最严重的地区之一。尽管目前沙漠化总体比上世纪90年代有了明显改善，但局部地区仍在扩展，形势仍然严峻。实践证明，现代遥感技术在土地沙漠化研究中具有许多其他技术不能取代的优势，如宏观、准确、概括、直观、动态和现势性等。因此，采取快捷的遥感沙漠化监测方法，准确地掌握土地沙漠化的分布状况和总体趋势，对沙漠化防治战略具有重要意义。论文以地球系统科学理论为指导，运用遥感(RS)、地理信息系统(GIS)技术，以景观生态学理论与原理为基础理论，选用了1960、1975年的航片和1987、1995、2000、2005年的TM影像数据等不同时期、不同空间分辨率的遥感数据进行处理，以人机对话方式提取6期土地沙漠化信息。其中，最小图斑为10×10个像元，解译过程中图斑轮廓线误差不超过三个像元(实地100m×100m)。利用GIS的空间分析与叠加功能，生成专题地图，并选用适当的数学模型，探讨了研究区近50a土地沙漠化时空变化规律。在此基础上，结合实地考察和气象、社会统计数据，借助SPSS统计分析软件，对沙漠化土地景观格局变化、生态足迹、沙漠化的驱动因素及累加效应、沙漠化发展态势CA模拟预测等宏观层面进行了系统地综合研究。(1)近50a多伦县土地沙漠化时空变化规律。结果表明，1960～1995年间沙漠化共增加了639.88km~2，而1995～2005年间减少了427.18km2，且主要分布于中部县城南部和北部地区。说明自实施京津风沙源治理、退耕还林、禁牧舍饲和生态移民四大工程以来，多伦县沙漠化逆转的较快，流沙多被固定，出现沙漠化土地面积逐步缩小的局面。(2)采用相关景观指数，进行了多伦县沙漠化土地景观格局变化分析，并探讨了不同类型沙漠化土地景观的分维和稳定性。从各沙漠化土地景观面积来看，各个时期沙漠化土地景观类型面积分布各不相同。1960年潜在沙漠化面积最大，其次为轻度沙漠化、严重沙漠化、重度沙漠化，中度沙漠化类型面积最小。在1975年和1987年两期沙漠化土地中各沙漠化土地面积从大到小依次为潜在、轻度、中度、严重、重度沙漠化土地；1995年整个沙漠化土地景观中轻度沙漠化土地景观面积最大，为主导沙漠化景观类型。2000年中度沙漠化面积最大，其次为轻度沙漠化、潜在沙漠化、重度沙漠化、严重沙漠化类型面积最小；2005年整个沙漠化土地景观中潜在沙漠化土地景观面积最大，为主导沙漠景观类型。从整个多伦县沙漠化情况来看，6时期各沙漠化土地类型平均分维数在1.3988～1.5635之间，仍接近于随机运动值，故有潜在的流动趋势。多伦县各类型沙漠化土地中，轻度沙漠化土地图斑镶嵌结构随着时间序列的增加稳定性趋于减弱，而重度沙漠化土地图斑镶嵌结构从1987年以来稳定性逐渐变好，且稳定性相对较高。(3)多伦县沙漠化驱动因素及累加效应分析。选取年均气温、降水量、蒸发量和平均风速等自然因素指标和年末总人口、国内生产总值、牲畜头数、耕地面积、粮食产量等人为因素指标，在系统分析沙漠化自然和人文驱动因素的基础上，采用滑动平均法和灰色关联分析法进行了沙漠化驱动因素影响的累加效应分析。结果表明，年均气温的累加效应在3a尺度上最大，年均降水量的累加效应在4a尺度上最明显，蒸发量和平均风速则在10a尺度上表现得最为明显。人口数量在1a平均尺度上达到了最高值，耕地对沙漠化的影响在1-10a尺度上逐渐减少，粮食总量对沙漠化的影响在5a平均尺度上影响达到最大，牲畜头数对沙漠化的影响在2a平均尺度上表现得最为明显。从主成分因子的信息贡献看，人为因素的贡献率为57.29％，自然因素的贡献率为13.16％，自然与人为因素共同作用的贡献率为16.76％。因此，人为因素的作用是多伦县土地沙漠化的最根本原因。(4)多伦县的生态足迹分析。根据资料的可获取性，采用1986～2005年的连续时间序列数据，对多伦县的生态足迹进行研究计算。1986～2005年的多伦县人均生态足迹总体呈递增趋势，由1986年的0.699hm~2到2005年的3.505hm~2，净增加2.806hm~2。从多伦县人均生态足迹与人均生态承载力比较分析发现，1986～1989年间多伦县呈生态盈余，而1990年开始，出现生态赤字现象。说明，随着沙漠化逆转，沙漠化转变为耕地、草地及林地，生态足迹呈逐年增长态势。但是，沙漠化治理过程中缺乏科学指导，导致恢复与重建生态系统服务支持功能与迅速增长的经济、工业化的跟不上，出现1990年来生态赤子加重。(5)多伦县沙漠化发展态势CA模拟。结果表明，预计未来10年多伦县沙漠化所占比重相比较2005年土地沙漠化类型所占比重有微弱增加趋势，从2005年的409.14km~2增长到2015年的452.71km~2。其中，轻度沙漠化在总面积中的比重还将持续增长，但增长的幅度不人，2005年的轻度沙漠化共159.60km~2到2010年增长到163.91km~2。中度沙漠化呈现微弱增长的趋势，预计到2010年增长到89.25km~2，到2015年有增长到了90.45km~2。重度沙漠化和严重沙漠化将呈略增加趋势，但是增加幅度很小，仅增加了24.33％，反映了治理沙漠化的难度越来越大。可以说，在治理沙漠化政策基本稳定的情况下，上述特征基本反映了多伦县未来10年土地沙漠化类型的变化特征。更多还原

【Abstract】 Desertification has been a main ecological environment question of North part of China in the past 50 years by its wide distribution and rapid expansion, influenced the local social economic sustainable development directly and threatened the ecological security of the economic cycle of Beijing and Tianjin. Dolonnur County of Inner Mongolia Autonomous Region is located in sand source fringe area around Beijing and Tianjin, has a typical ecological vulnerability. Since nearly 100 years, as a result of the population sharp increase, the big area of natural pasture opened up for the farming, and especially from 50’s of 20th century, because the one-sided route of "regards grain as the key link" during the periods of "the Great Leap Forward" and "the Great Cultural Revolution", has made Dolonnur County one of the most serious desertification threatened area in the region. Although the present overall situation of desertification had the distinct improvement compared to that on 90’s of the last century, but the local area was still expanding. The modem remote sensing technology has the superiority in the land desertification research, like macroscopic, accurate, broad, direct-viewing, tendency and current situation and so on. Therefore, adopts the quick remote sensing monitor method, accurately grasps the distribution and changing tendency of desertification, has the vital significance to the desertification preventing and controlling strategy.Taking the theory of Earth systems science as the direction, the landscape ecology theory and principle as the basic theory, using remote sensing(RS), geographic information system(GIS) technology, aerial photo of 1960, 1975 and TM data of 1987, 1995, 2000, 2005 with different time and spatial resolution were processed, and the land desertification information in 6 periods of time were extracted by the way of human-computer conversation. Among them, the smallest patch is in 10×10 pixels, and interpretation processing error is within three pixels (100×100m). The spatio-temporal characteristics of desertification evolution of the research area in nearly 50 years were discussed by using the method of GIS spatial analysis, overlay function and suitable mathematical model.(1) The spatio-temporal characteristics of desertification evolution of Dolonnur county in nearly 50 years. The results showed that, desertification area in 1960-1995 increased 1106. 2 km~2, and in 1995-2005 reduced 487.71 km~2. It mainly distribute on central region of country and north part region. It indicated that, since the implementation of "the Beijing-Tianjin dust storms sources control project", "the project of withdrawing graze and returning to grassland", "banning herd and confinement feeding project", "eco-migration project", the area of desertification in Dolonnur county reverses quickly, most of the mobile dunes were stabilized and the total area of desertification gradually reduced.(2) The landscape pattern of land change of desertification in Dolonnur county was analyzed through the landscape metrics computation, and the fractal dimension and stability of different desertification land landscape were discussed. In 1960, the area of initial state desertification area was the largest, then for the mild desertification, the serious desertification, the severe desertification, the area of moderate desertification was the smallest. The desertification land in 1975 and 1987, the each land desertification area from large to small are: initial state, mild desertification, serious desertification, the severe desertification land. In 1995, the area of mild desertification land landscape was the largest, was the dominant type of desert landscape. In 2000, moderate desertification area was the largest, then for the mild desertification, the initial state, the severe desertification, area of the severe desertification was the smallest. In 2005, the area of the initial state desertification land landscape was the largest, also was the dominant type of desert landscape. On the whole, the average value of fractal dimension of desertification area in Dolonnur county are between 1.3988-1.5635 in 6 periods of time, which still close to the value of random motion, therefore it had the latent mobile tendency. The stability of the structure of patch mosaic of mild desertification area in Dolonnur county tended to weaken along with the time series increasing, but the stability of the structure of patch mosaic of the severe desertification area in Dolonnur county has gradually improved since 1987, also the stability is relative high.(3) Analysis of driving factors of desertification and cumulative impact of Dolonnur county. We select some natural factors, including average annual temperature, precipitation, evaporation, average wind speed and etc; and some human factors, including total year-end population, GDP, livestock number, the area of plow land, the food yield and etc. The cumulative impact of driving force of desertification was analyzed with sliding average and grey relational analysis. The results showed that the accumulative impact of annual mean temperature came to the maximum in the third year, the accumulative impact of annual mean precipitation was the most significant in the fourth year, the impact of evaporation and wind speed was gradually increased with the time scale and was in most significant in the tenth year; population reached the maximum in the first year; the impact of the area of farmland of desertification gradually decreased in the duration of one to ten years; the impact of desertification of aggregate yields of grain reached to the maximum in the fifth years; the impact of desertification of total number of livestock was the most significant in the second years. From the contribution of the information of principal components, the ratio of the contribution of human factors is 57.29％, the contribution of natural factors is 16.76％. Therefore, and the root cause of land desertification ishuman factors.(4) The analysis of desertification of Dolonnur county. The ecological foot-print of was Dolonnur county calculated by using time serial data from 1986 to 2005. The results indicated that, per capita ecological foot-print of Dolonnur county from 1986 to 2005 had an increasing tendency, 0.699hm~2 in 1986 and 3.505hm~2 in 2005, increased 2.806hm~2. In 1986-1989, it’s being discovered from comparing and analyzing per person of ecological foot-print and per person of ecosystem bearing capacity, the Dolonnur County assumed the ecology earnings, but from 1990, it appeared the ecological deficit. It showed that, along with the desertification reverses, the desert transformed into farmland, grassland and forest land, the ecological foot-print assumed increasing year by year.(5) CA simulation of desertification in Dolonnur county. The results showed that, the proportion of the area of desertification in Dolonnur county in recent 10 years compare to 2005 has small increasing tendency and the area will increase from 409.14km~2 in 2005 to 452.71km~2 in 2015. Among them, the proportion of mild desertification will continually grow, but the scope growing will be small. The area of initial state and mild desertification will grow from 159.60km~2 in 2005 to 163.91km~2 in 2010. The area of moderate desertification presents a weak growing tendency, will reach to 89.25km~2 in 2010, and 90.45km~2 in 2015. The severe and serious desertification will assume the tendency which will increase slightly, but the increasing scope will be very small, is only 24.33％, which reflects the desertification control will be more and more difficult. It can be said that, in the stable situation of desertification control policy, the above characteristics have basically reflected changing type of desertification of Dolonnur county in future 10 years.Because the desertification is a complex process of the interaction between nature and the human factors, some of the questions will need to be in thorough research and be perfect. (1)the question of time correspondence with remote sensing image;(2) microscopic and macroscopic unifies, using GIS space expansion model, distribution limited conventional view computation, spatial;(3) quantitative analysis of intrinsic mechanism of the formation and development of desertification;(4)analysis of association effect and feedback mechanism of desertification.更多还原