QQ客服 官方微博 反馈留言

题名	西双版纳天然林与人工林变化遥感监测及局域气候影响研究
作者	刘恩勤
学位类别	博士
答辩日期	2013
授予单位	中国科学院研究生院
授予地点	北京
导师	周介铭
关键词	自然林 人工林 决策树 变化检测 气候变化
其他题名	Remote Sensing Monitoring of Natural and Artificial Vegetation Changes in Xishuangbanna and Local Climate Effects Research
学位专业	自然地理学
中文摘要	人类活动引起的土地覆被变化被认为是全球变化的重要原因之一。土地覆被在很大程度上取决于地表植被的覆被状况。对植被覆被变化的格局和过程的研究是土地利用/覆被变化研究的重要方面。近年来，地球植被变化与气候变化的关系研究也是全球变化的研究焦点之一，引起了人们的广泛关注。西双版纳良好的热带气候条件为植物和作物的生长提供 良好的基础，同时也是人工橡胶林，茶园在该地种植的重要基础条件。近几十年来，西双版纳人工橡胶林大量种植的同时，包括热带雨林在内的原始森林面积大量减少。西双版纳 是热带森林植被变化的典型区域，也是在人类活动影响下土地利用/覆被发生巨大变化的区域。另一方面，西双版纳的近年来极端天气频发，气候问题突出。 本文基于“3S”技术，应用多平台、多时相的遥感影像，结合DEM、野外采样光谱数据、野外样点调查数据等数据源，监测西双版纳近21年自然林和人工植被变化的格局和过程 ，揭示了数量、空间分布和人工种植林的地形等的变化特点。另外，本文利用景洪市近31年的气象数据分析了该区域近年来的气候变化情况。并利用提取的植被数量数据与气候因 子数据分析了二者的相关性。研究西双版纳近年来的植被变化特点、热带气候的局域变化特征，以及植被变化与气候变化的关系，该研究对于加强对热带自然林的保护和管理具有重要意义，同时对于人们认识和研究局域尺度上的LUCC气候效应也具有重要作用。本论文的主要内容和取得的创新成果有如下几个方面： （1）在考虑遥感影像的时相、不同类型地物之间的可分性等多方面因素的基础上，综合利用地物的光谱、纹理、地形和气候等多特征，构建基于地物多特征的决策树分类模型， 将其用于土地利用/覆被信息的提取工作。建立了2005年西双版纳土地利用/覆被数据库和植被空间数据库，为西双版纳的植被分布的长期监测提供了基础数据。通过与其他分类方 法的对比和对分类结果的精度验证表明，该分类方法能综合考虑地物的光谱特征、纹理特征、地学特征和物候特征，使之有别于传统仅考虑光谱特征的信息提取方法，能大大提高 土地利用信息提取的精度。 （2）针对传统遥感影像变化检测方法的缺点以及利用光谱特征进行检测的局限性，同时考虑到多特征的决策树方法有效性，本文将多特种的决策树方法引入遥感影像的变化检测中，提出了一种基于差值影像和地物多特征的决策树变化检测方法，为遥感影像的变化检测提供了新思路和新方法。对多特征影像各波段的重要性分析表明，纹理特征信息和面向对象影像分割信息在模型建立中起到了重要作用。本文利用该方法检测出1988-1999年，1999-2005年，2005-2010年共三组影像间的变化区域及变化前后的属性信息，建立1988-1999年，1999-2005年，2005-2010年的植被空间变化信息数据库。结合提取的2005年的植被空间分布数据进行空间分析，得到1988年，1999年，2005年和2010年的植被空间数据库。通过精度验证并与其他变化检测方法的结果对比，证明了该变化检测方法能有效检测出植被变化信息，并降低漏检率。 （3）利用获取的1988-1999年、1999-2005年、2005-2010年以及1988-2010年几个不同阶段的植被变化空间数据，并结合土地利用模型从数量特征、空间特征以及土地利用变化速度等方面揭示近年来西双版纳自然林和人工林的变化的时空过程和格局。利用人工橡胶林的空间分布数据与DEM数据进行叠加分析，探讨了橡胶林的空间分布模式和地形变化特征。结果表明，近年来西双版纳的自然林面积呈持续减少，而人工林面积呈逐渐增长趋势，人工林面积增长的部分主要是由自然林转入的；另外，人工林的种植海拔和坡度有增高的 趋势。 （4）利用景洪市的1980-2010年的气象数据，分析了31年间其气候变化特征。结果表明，近年来景洪市的气候发生了明显的变化，年平均温度和平均高温呈波动上涨趋势，云量减少，同时无降水天数越来越多，年平均风速明显增大，干旱天气发生的越来越频繁。近年来西双版纳与云南省的气候变化情况有部分相似之处，但是西双版纳的气候变化具有非常明显的区域特点。 （5）利用获取的1988年、1999年、2005年及2010年共四期植被数据库，以及1980-2010年共31年间的气象数据做相关性分析，探讨植被变化对气候的反馈作用。研究表明，橡胶林的面积和年平均温度、年均高温、无降水天数、平均风速、云量、最大蒸发量等气象要素具有很强的相关性，并且气象数据的变化趋势和与同时间段的橡胶林种植情况的变化能够很好的吻合。这说明了西双版纳人工林的大量种植和自然林的大量砍伐对当地气候产生了巨大影响。 笔者在此呼吁保护天然森林植被，遏制天然林面积快速减少的恶劣现状和趋势，限制人工林的种植，特别是高海拔陡坡人工林。政府及相关部门能制定相应的森林资源保护策略和科学规划，妥善保护天然森林等各类宝贵资源。
英文摘要	Human-induced land cover change is considered to be one of the important reasons of global changes.Land cover largely depends on the surface condition of vegetation cover.The study of patterns and processes of vegetation cover change is an important aspect of land use / cover change research. Vegetation is the main part of the terrestrial ecosystem. It adapted to the climate, topography, and soil conditions. And it was controlled by natural and man-made factors. So it was sensitive to changes in various factors. In recent years, Study of the relationships between vegetation changes and climate changes become one of the most widely concerned issues. The tropical climatic conditions of Xishuangbanna provide a good foundation for the growth of plants and crops. And it was also the foundation conditions of artificial rubber and tea plantations. In recent decades, a large number of artificial rubber forests were planted in Xishuangbanna and following a significant reduction of original forest area including tropical rain forests. On the one hand, Xishuangbanna is a typical area of tropical forest vegetation changes.The land use/cover area changed dramatically under the influence of human activities. On the other hand, the frequency of extreme weather rises in recent years. The climate issues become more serious every day. The "3S" technology, including remote sensing (RS), Geographic Information system (GIS) and Global Positioning System (GPS) was the foundation of this paper. The remote sensing images of multi-platform, multi-temporal, DEM, spectral data form field sampling, field survey data and other data sources was applied for monitoring the pattern and process of vegetation change in Xishuangbanna nearly 21 years .In addition, meteorological data nearly 31 years of Jing Hong City(State capital of Xishuangbanna) was used for analyzing the climate changes. The correlation between extracted vegetation data and climate  data was analyzed. Research of the vegetation change characteristics of Xishuangbanna in recent years, the local variation of the tropical climate, and the relationship between vegetation change and climate change was significant for understanding the vegetation climate effects on the local scale. And it was also meaningful for strengthening protection and management for tropical forests. The main content of the paper and the innovations was as follows: (1) multi-features decision tree classification model and multi-features decision tree change detection model were build for monitoring time series of natural forests and plantations vegetation distribution information. The separability between the different types of surface features, and many other factors were under the consideration. Surface features of the spectrum, texture, terrain and phenology characteristics were comprehensively utilized for building multi-features decision tree classification model. And the model was used for land use/cover information extracting, building land use / land cover database and vegetation database of the year 2005 in Xishuangbanna. It provided the basic data for long-term monitoring of the distribution of vegetation in Xishuangbanna. The applicability and effectiveness of the classification was discussed by comparison with other classification methods and the accuracy of classification results was verified. Considering the shortcomings of the traditional remote sensing image change detection method, a new image change detection method which based on the difference between images and multi-features decision tree was introduced in this paper. It provides a new idea and method for remote sensing image change detection. This method was used for detecting the changed area and recognizing attributes information in total three groups images: from1988 to1999, from 1999 to 2003, from 2005 to 2010. Then vegetation database of the year 1988, 1999, 2005 and 2010 was also established by spatial analysis between vegetation changes spatial information and the vegetation data of the year2005.The effectiveness and accuracy of the method was discussed by comparison with other classification methods. Vegetation change data and land-use model were used for reveal the spatial and horal process and pattern of vegetation changes in recent years in Xishuangbanna considering the quantity, spatial distribution, and the pace of change. The changes in the topographic characteristics of the rubber plantation was recognized by overlay analysis both spatial distribution data of artificial plantation and topography data which including elevation, slope, aspect, etc. The result shows that area of natural forest in Xishuangbanna was continued to decline, while the artificial plantation area was gradually increasing in recent years. The growing area of artificial plantation was mainly transferred from natural forests. And artificial plantation was grown to a higher slope and higher elevation in recent years. (2) Meteorological data of Jing Hong City from1980 to 2010 was used for analysis the climate change characteristics. The result shows that the annual average temperature, average high temperature and average wind speed were on the rise in recent years. Cloudiness decrease at the same time, and precipitation-free day’s increased, dry weather is occurring more and more frequently. (3) The feedback effect of vegetation change on climate was addressed by correlation analyzing between vegetation data of year 1988, 1999, 2005 and 2010 and  meteorological data from the year from 1980 to 2010.It has been shown that there is a strong correlation between the rubber plantation area and annual mean temperature, average annual high temperature, non-rainy days, average wind speed, cloud cover, maximum evaporation and other weather elements. Furthermore, the trends of rubber-planting and Climate Change were in good agreement. It indicates that Climate Change and the planting of rubber plantation have a tremendous impact on the local climate in Xishuangbanna. Author calls for protection of natural forests, especially tropical rain forests and other natural forest vegetation. Do not plant rubber any more. The rapid decrease of natural forest area and the expansion of the plantation must be curbed. Furthermore, government and relevant departments should develop appropriate strategy for the protection of forest resources.
语种	中文
公开日期	2014-07-26
内容类型	学位论文
源URL	[http://ir.imde.ac.cn/handle/131551/7058]
专题	成都山地灾害与环境研究所_数字山地与遥感应用中心
推荐引用方式 GB/T 7714	刘恩勤. 西双版纳天然林与人工林变化遥感监测及局域气候影响研究[D]. 北京. 中国科学院研究生院. 2013.

个性服务

查看访问统计

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。