Knowledge Resource Center for Ecological Environment in Arid Area
| 城市群土地覆被变化及其对夜间光环境变化影响研究 | |
| 其他题名 | Study on Urban Agglomeration Land Cover Change and its Impact on The Nighttime Lighting Environment Change |
| 韩鹏鹏 | |
| 出版年 | 2015 |
| 学位类型 | 博士 |
| 导师 | 李仁东 |
| 学位授予单位 | 中国科学院大学 |
| 中文摘要 | 土地利用/土地覆被变化（Land-Use and Land-Cover Change, LUCC）是全球变化及其区域响应研究的核心领域。土地利用/土地覆被变化对人类改变地球表面空间格局的活动进行了客观记录，再现了地表景观的时空动态过程，与全球气候的变化、生态环境的演变、生物多样性的减少以及人类与自然环境间的相互作用的可持续性等休戚相关。土地利用变化通过改变地表覆被情况来影响陆地生态系统的机构和分布，从而改变陆地生态系统的碳循环和碳储量（柳梅英等, 2010），是造成全球碳循环失衡的重要原因之一（Jean and Anver, 1994）。我国人口-资源-环境问题日益凸显（Wu et al., 2008; 史培军等, 2006; Hepinstall Jeffrey et al., 2008），土地利用/土地覆被变化已成为可持续发展的核心问题之一。对土地利用/土地覆被变化进行深入研究，可为我国国家和地方的国土规划编制、土地可持续利用标准的建立等提供科学决策的依据，对实现我国土地资源的合理开发、利用和保护，实现经济、生态和环境的协调发展具有重要的现实意义。 20世纪以来，人类在追求快速发展的过程中深刻地影响着全球环境的变化，与此同时，一系列全球性重大环境问题的产生也威胁着人类的生存和发展，特别是在发展中国家，环境问题更为严重。目前，中国作为世界上最大、发展最快的发展中国家，大气污染、水污染、土地荒漠化等已成为最主要的环境问题，同时，随着人口膨胀和城市化过程的加快，如噪声污染、电磁污染和光污染等物理污染问题正变得日益严峻。城市夜景照明技术改善了人类的生存环境，提高了人们的生活质量，延长了人们的活动时间。但随着人类无序、过度地开发和滥用，有些夜间照明带来的光成为了多余有害的污染源，对人的身心健康、动植物的生产发育造成不同程度的影响，甚至威胁全球生态系统平衡。对城市夜间光污染的研究有助于城市生态系统保护，促进城市夜景照明科学和可持续发展研究，掌握我国城市夜间光环境变化趋势和空间格局，为我国光污染防治体系建设提供决策依据。 本研究以多时相HJ星数据、地面调查样点和DEM等辅助数据，利用面向对象的方法，基于eCognition软件平台按工程分块的原则对福建省2010年进行土地覆被分类；以福建省2010年土地覆被分类结果为基础，利用基于像元和面向对象的变化检测方法，得到海西城市群1990、2000和2010年三期土地覆被数据，并进行土地覆被变化分析；从时间和空间角度分析海西城市群、武汉城市群和南北钦防城市群1990、2000和2010年土地覆被变化特征，重点分析三大城市群城市扩张引起的土地覆被变化特征；基于DMSP/OLS和NPP/VIIRS夜间灯光遥感数据，进行不同尺度上我国光污染变化趋势遥感监测分析，并结合土地覆被数据进行空间关联分析，探究城市土地覆被与城市光污染变化趋势的相关关系。 本文的主要研究结论如下： （1）基于多时相HJ星数据、野外调查数据和DEM等辅助数据，利用面向对象的分类方法，在eCognition软件平台支持下，按照工程分块的原则进行层次分类，得到2010年三大城市群土地覆被分类结果。利用地面调查的实测样点分别对海西城市群、武汉城市圈和南北钦防城市群2010年土地覆被一级分类和二级分类结果精度进行精度检验，其中海西城市群一级分类总体精度为92.88%，Kappa系数为83.96%；二级分类总体精度为83.21%，Kappa系数为79.62%。武汉城市圈一级分类总体精度为91.11%，Kappa系数为83.74%；二级分类总体精度为75.50%，Kappa系数为68.08%。海西城市群一级分类总体精度为90.76%，Kappa系数为82.13%；二级分类总体精度为81.52%，Kappa系数为76.84%。 （2）基于相近时相的Landsat数据，利用基于像元和面向对象的变化检测方法进行土地覆被数据更新，得到三大城市群1990、2000和2010年三期土地覆被数据，在此基础上进行土地覆被变化分析。海西城市群1990-2000年间变化面积为4204.52km2，占海西城市群总面积的7.69%；海西城市群2000-2010年间变化面积为4378.58km2，占海西城市群总面积的8.01%。2000-2010年变化面积比1990-2000年多174.06km2，可以看出，海西城市群在1990-2010年期间，土地利用强度持续增大，人地关系作用强烈。海西城市群1990-2010年间土地覆被变化总体呈现林地面积稳中有升，耕地面积持续减少，人工表面明显扩张，湿地和草地面积相对保持稳定。武汉城市圈1990-2000年间变化面积为1339.49km2，占武汉城市圈总面积的2.32%；武汉城市圈2000-2010年间变化面积为1833.74km2，占武汉城市圈总面积的3.17%。2000-2010年变化面积比1990-2000年多494.25km2。南北钦防城市群1990-2000年间变化面积为2660.45km2，占南北钦防城市群总面积的6.26%；南北钦防城市群2000-2010年间变化面积为1000.66km2，占南北钦防城市群总面积的2.36%。2000-2010年变化面积比1990-2000年少1659.79km2。三大城市群1990-2010年间土地覆被变化共同呈现出人工表面明显扩张的趋势。 （3）基于多时相土地覆被数据，分别提取海西城市群、武汉城市圈和南北钦防城市群1990、2000和2010年三期城市扩张土地覆被变化数据，重点从城市扩张的角度进行探究。海西城市群人工表面从1990年的2453.21km2扩张到2010年4124.34km2；武汉城市圈人工表面1990年为2236.30km2，到2010年增长到4062.47km2；南北钦防城市群人工表面从1990年1321.96km2增长到1627.62km2（2010年）。三大城市群2000-2010时段城市扩张面积均大于1990-2000时段，但城市扩张地区差异各具特色，海西城市群城市扩张中心区具有南移特点；武汉城市群城市扩张呈现“中心高四周低”；南北钦防城市群城市扩张面积呈现从南宁分别向东南和西南方向递减的趋势。1990-2000时段内，海西城市群城市扩张速度最快，为55.04km2/a；其次为武汉城市圈和南北钦防城市群，分别为24.68km2/a和13.08km2/a。2000-2010时段内，武汉城市圈超越海西城市群成为三大城市群中城市扩张速度最快的地区，武汉城市圈城市扩张速度为157.94km2/a。可以看出，1990-2000年间东部地区城市群快速发展，城市扩张度速度快；而在2000-2010年间，由于中部崛起，中部地区城市化快速发展，城市扩张速度明显加快，因此出现2000-2010年间武汉城市圈城市扩张速快于海西城市群。海西城市群城市扩张的主要来源是耕地和园地；武汉城市圈城市扩张主要是由耕地转化得到；南北钦防城市群城市扩张则主要是由耕地、林地和湿地转变而来。 （4）基于1992-2012年DMSP/OLS夜间灯光遥感数据，利用一元二次回归模型进行灯光数据相互校正，并分3个时段1992-1996、2000-2004和2008-2012进行年均值计算，对中国大陆进行光污染变化趋势遥感监测分析。中国光污染变化趋势与我国城市空间分布具有密切的相关性；我国光污染变得越来越严重，从大城市向中小城市扩张；光污染变化趋势具有明显的东西差异，东部地区光污染升高趋势明显；大城市中心区光污染变化趋势相对稳定或降低，城市外围光污染呈升高扩张趋势。 （5）三大城市群城市扩张与夜间光环境变化具有较好的空间分布关联性。通过分析1990-2000和2000-2010两个时段土地覆被类型对夜间光环境变化的影响值，发现城市扩张土地覆被类型整体上对夜间光环境的影响逐渐升高。 （6）基于2012-2014年NPP/VIIRS夜间灯光遥感数据，利用Robust regression模型进行影像校正，在城市尺度上结合土地覆被数据进行光污染变化趋势分析。深圳市高大建筑对应的夜间光环境平均值最大；乡村居民点对应的夜间光环境平均值最小。高大建筑与乡村居民点夜间光环境平均值差距从2012年的15.3758减低到2014年的4.4167，这在一定程度上表明深圳市夜间光环境趋向于均一化、平衡化。深圳市光环境变化在2012-2014时期内总体呈现城区降低，郊区、远郊区相对升高的的趋势。 |
| 英文摘要 | Land use/land cover change is a global change and its regional response core areas. Land use/land cover changes in the objective record of the Earth’s human alter the spatial pattern of surface activity, reproduce the spatial and temporal dynamics of the surface landscape, associated with global climate change, ecological evolution, reduce biodiversity, sustainability between humans and the natural environment interactions, and at all. Land use change is one of the important causes of the imbalance of the global carbon cycle, and alters terrestrial ecosystem carbon cycle and carbon stocks by changing land cover conditions to affect institutions and the distribution of terrestrial ecosystems. As population, resources and environmental issues in China become increasingly prominent, land use/land cover change has become one of the core issues of sustainable development. Land use/land cover change in-depth study provide a scientific basis for decision-making in national and local territorial planning and sustainable land use standards, has important practical significance in rational development of land resources,use and protection, and coordinated development of economy, ecology and the environment. Since the 20th century, human profound impact changes in the global environment in the process of pursuit of rapid development. At the same time, a series of global environmental issues are also a major threat to human survival and development particularly in developing countries. Now, air pollution, water pollution, desertification, etc., have become the most important environmental issues in China which is the world largest and fastest growing developing countries. With the acceleration of population growth and urbanization process, physical problems such as noise pollution, electromagnetic pollution and light pollution are becoming increasingly serious. Urban landscape lighting change the human living environment, improve the quality of people’s lives, and extend people’s active time. But, as human disorder, excessive exploitation and abuse, some urban night lighting changed to harmful pollution and affect people’s physical and mental health, production and development of plants and animals, and global ecosystem balance. Study of urban light pollution can be helpful in urban ecosystem protection, urban lighting science and sustainable development, understand trends and spatial patterns of changes in the urban nighttime environment, and provide decision-making basis for the construction of light pollution prevention and control system in our county. In this study, based on HJ multitemporal satellite data, ground survey samples, DEM and other auxiliary data, we used object-oriented approach in eCognition software with engineering block principle in the land cover classification of Fujian province in 2010. We obtained land cover classification of Haixi urban agglomerations in 1990, 2000 and 2010 from and cover classification of Fujian province in 2010 using pixel-based and object-oriented change detection method, and analyzed land use change from 1990 to 2010. After that, we analysis land-use change characteristics from the perspective of time and space through compare Haixi urban agglomeration, Wuhan urban agglomeration and Nan-Bei-Qin-Fang urban agglomeration in 1990, 2000 and 2010, and focuses on three major urban agglomerations land use changes features caused by urban sprawl. In additional, we monitor trends of light pollution in our county from different scale based on DMSP/OLS and NPP/VIIRS nighttime light satellite data, and analyze spatial relationships between light pollution changes and land use changes in urban areas. The main conclusions are as follows: (1)Based on HJ multitemporal satellite data, ground survey samples, DEM and other auxiliary data, we used object-oriented approach and hierarchical classification method in eCognition software with engineering block principle in the land cover classification of three urban agglomerations in 2010. We used land survey samples to assess land cover classification results in three urban agglomerations. Haixi urban agglomeration’s first-type overall accuracy is 92.88%, kappa coefficient is 83.96%, second-type overall accuracy is 83.21%, kappa coefficient is 76.92%. Wuhan urban agglomeration’s first-type overall accuracy is 91.11%, kappa coefficient is 83.74%, second-type overall accuracy is 75.50%, kappa coefficient is 68.08%. Nan-Bei-Qin-Fang urban agglomeration’s first-type overall accuracy is 90.76%, kappa coefficient is 82.13%, second-type overall accuracy is 81.52%, kappa coefficient is 76.84%. (2) Based on similar temporal Landsat data, we obtained land cover classification of three urban agglomerations in 1990, 2000 and 2010 using pixel-based and object-oriented change detection method, and analyzed land cover change from 1990 to 2010. Between 1990 and 2000, Haixi urban agglomeration change area was 4204.25 km2, accounting for 7.69% of the total area of Haixi urban agglomeration, and between 2000 and 2010 change area was 4378.58km2 accounting for 8.01% of the total area of Haixi urban agglomeration. Change area between 2000 and 2010 more 174.06 km2 than the period between 1990 and 2000. In the period from 1990 to 2010, land use intensity continues to increase and man-land relationship become more intense. The overall land use change in Haixi urban agglomeration from 1990 to 2010 presents woodland area steadily, arable land continued reduction, significant expansion of artificial surfaces, and wetland and grassland areas remained relatively stable. Wuhan urban agglomeration change area was 1339.49 km2, accounting for 2.32% of the total area of Wuhan urban agglomeration from 1990 to 2000, and between 2000 and 2010 change area was 1833.74 km2 accounting for 3.17% of the total area of Wuhan urban agglomeration. Change area between 2000 and 2010 more 494.25 km2 than the period between 1990 and 2000. Nan-Bei-Qin-Fang urban agglomeration change area was 2660.45 km2, accounting for 6.26% of the total area of Nan-Bei-Qin-Fang urban agglomeration from 1990 to 2000, and between 2000 and 2010 change area was 1000.66 km2 accounting for 2.36% of the total area of Wuhan urban agglomeration. Change area between 2000 and 2010 less 1659.79 km2 than the period between 1990 and 2000. Land cover change trend in three urban agglomerations from 1990 to 2010 commonly show significant expansion of artificial surfaces. (3) We analysis land cover change characteristics from the perspective of time and space through compare Haixi urban agglomeration, Wuhan urban agglomeration and Nan-Bei-Qin-Fang urban agglomeration in 1990, 2000 and 2010, and focuses on three major urban agglomerations land use changes features caused by urban sprawl. Haixi urban agglomeration artificial surface area increased from 2453.21 km2 in 1990 to 4124.34 km2in 2010; Wuhan urban agglomeration artificial surface area expanded from 2236.30 km2 in 1990 to 4062.47 km2 in 2010; Nan-Bei-Qin-Fang urban agglomeration artificial surface area spread from 1321.96 km2 in 1990 to 1627.62 km2 in 2010. All of the three urban agglomerations expansion area in the period of 2000-2010 larger than 1990-2000, but with different characteristics. Urban sprawl center show southward trend in Haixi urban agglomeration, while urban expansion present “center high low around” pattern in Wuhan urban agglomeration. In addition, Nan-Bei-Qin-Fang urban agglomeration increased urban area show decreasing from Nanning to southeast and southwest, respectively. In the time from 1990 to 2000, Haixi urban agglomeration urban expansion speed is 55.04km2/a, which is the fastest among of the three urban agglomerations. Urban sprawl speed are 24.68 km2/a and 13.08 km2/a in Wuhan urban agglomeration and Nan-Bei-Qin-Fang urban agglomeration, separately. In the time from 2000 to 2010, Wuhan urban agglomeration urban sprawl speed change to 157.94 km2/a and exceed Haixi urban agglomeration to be the fastest in three urban agglomerations. Urban sprawl show rapid growth in eastern region in the time 1990-2000. Due to implementation of the strategy of Central China, the central region show rapid urbanization process with significantly faster urbanization speed, and urban expansion speed in Wuhan urban agglomeration located in central region faster than in Haixi urban agglomeration located in eastern region in the time from 2000 to 2010. Arable land and garden are the main sources of urban expansion in Haixi urban agglomeration, while arable land change to artificial surface in Wuhan urban agglomeration. In addition, arable land, woodland and wetland are the major resources of urban sprawl in Nan-Bei-Qin-Fang urban agglomeration. (4) Based on long term DMSP/OLS nighttime lighting satellite data from 1992 to 2012, we obtain mean year value data in three time period, which are 1992-1996, 2000-2004 and 2008-2012, after intercalibration through second order regression model, and monitor trend of light pollution in China. Light pollution trends show a strong correlation with urban spatial distribution in China. Light pollution is becoming more seriously and expand from lager cities to medium and small cities. Light pollution trends have significant spatial differences and increase trend in eastern region larger than central and western region. light pollution trends in central areas in large cities show decrease or keep stable, while increase outside of central cities. (5) Urban expansion in three urban agglomerations show a spatial association with night light change. The land cover class impact value on night light change show a growing trend from compare impact value in 1900-2000 and 2000-2010. (6) Based on NPP/VIIRS nighttime lighting satellite data from 2012 to 2014, we combine land use data to monitor light pollution trends in urban scale after intercalbration through robust regression model. In Shenzhen, high buildings has the highest mean nighttime lighting value, while rural settlements has the lowest mean nighttime lighting value. The difference of mean nighttime lighting value between high buildings and rural settlements change from 15.3758 in 2012 to 4.4167 in 2014. The change difference show that nighttime lighting environment is becoming homogenization and balance in Shenzhen between urban core area and suburbs. The total change of nighttime lighting show decrease trend in urban area while suburbs display increase relatively in Shenzhen from 2012 to 2014. |
| 中文关键词 | 土地利用/土地覆被变化 ; 面向对象 ; 变化检测 ; DMSP/OLS ; NPP/VIIRS |
| 英文关键词 | land use/land cover change object-oriented change detection DMSP/OLS NPP/VIIRS |
| 语种 | 中文 |
| 国家 | 中国 |
| 来源学科分类 | 自然地理学 |
| 来源机构 | 中国科学院测量与地球物理研究所 |
| 资源类型 | 学位论文 |
| 条目标识符 | http://119.78.100.177/qdio/handle/2XILL650/287472 |
| 推荐引用方式 GB/T 7714 | 韩鹏鹏. 城市群土地覆被变化及其对夜间光环境变化影响研究[D]. 中国科学院大学,2015. |
| 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 |
| [韩鹏鹏]的文章 |
| 百度学术 |
| 百度学术中相似的文章 |
| [韩鹏鹏]的文章 |
| 必应学术 |
| 必应学术中相似的文章 |
| [韩鹏鹏]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。
