景观格局对城市地表热岛效应的影响研究--以北京市为例
文献类型：学位论文
| 作者 | 陈爱莲 |
| 学位类别 | 博士 |
| 答辩日期 | 2014-05 |
| 授予单位 | 中国科学院研究生院 |
| 授予地点 | 北京 |
| 导师 | 陈利顶 |
| 关键词 | 城市热岛 格局指数 地表温度 源汇景观理论 位置加权景观指数 （LWLI） Urban heat island landscape metrics “source”-“sink” landscape theory location weighted landscape index （LWLI） |
| 其他题名 | Study on effects of landscape pattern on surface urban heat island -a case of Beijing |
| 学位专业 | 生态学 |
| 中文摘要 | 城市热岛效应是世界范围内普遍存在的城市气候现象，在中纬度的夏季其危害及伴生效应影响众多。如何缓解城市热岛效应所带来的危害是国内外学者和政府部门关心的特点问题之一。探究和评价景观格局对城市热岛效应的影响可为城市（镇）规划提供理论基础，也为缓解城市热岛效应提供客观可行的方法参考。 本文以北京五环内城市核心区作为研究区，以2002 年Quickbird 和2012年IKONOS 高分辨遥感影像数据作为景观分类数据。首先通过非监督和决策树相结合的分类方法，以二维角度对待研究区城市景观，将其分成4 个土地覆盖类型：即水体、植被、裸土和建设用地，并将局部区域的数据通过人工目视解译分成15 个小类，以此作为景观格局分析的基础数据；其次，以Landsat5 和7的TM/ETM+数据反演地表温度，作为地表热岛分析的基础数据；最后，采用传统景观格局指数分析研究区景观格局对城市地表热岛的影响，评价两者关联分析的干扰因子，并引入改建的源汇景观空间负荷比指数(LWLI)来指示城市热岛效应。 本研究比较系统地评价了景观格局对地表热岛的影响大小及其干扰因子，为景观格局与城市热岛分析提供了指数挑选、尺度选择、粒度选择等方面的方法参考，为格局分析方法提供了新的思路，也为城市规划提供了一些科学依据。研究得到如下结论： （1） 景观格局分析显示：2002 年至2012 年北京市五环内地区建设用地增加了近12%，主要由裸土地（耕地）转变而来，植被减少了3%，水体几无变化；地表热岛分析显示：2002 年北京市五环内夏季（7 月）热岛区域总面积和最大面积均大于2002 年其他时相，年际热岛区域对比显示，5 个年份内2002年夏季的热岛区域总面积和最大面积也大于2009 年夏季，2009 年夏季的热岛区域总面积、最大面积又大于2007 年、2011 年及2012 年夏季。 （2） 从二维景观（土地覆盖）的角度上看，热岛区域下方的景观格局主要受建设用地主导。在北京城市核心区，建设用地所占比例达到或超过84%时，这个区域在夏季会形成高于整个研究地区高达8°的热岛区。 （3） 景观水平的格局指数与热岛及地表温度的相关性受建设用地组成百 分比的影响十分明显；类型水平的景观格局组成和配置对景观单元的平均温度和中心温度均有影响，其中类型组成的影响最大。 （4） 夏季，建设用地格局对夏季景观中心温度的解释度最高达 58%，其中，其组成（PLAND）解释度一般在53%左右，而配置的解释度约5%左右；建设用地格局对夏季景观平均温度的解释度最高达 79% ， 其中其组成（PLAND）解释度一般在73%左右，而配置的解释度约在6%左右。 （5） 景观格局对地表温度的解释度随景观格局分析方法、景观分类数据、温度时相的不同而不同。在建设用地格局与景观分析单元平均温度或中心的温度的回归分析中，圆形分析单元较方形分析单元回归系数高，小尺度分析单元较大尺度分析单元回归系数高；分类数据的分类精度越高，越有利于景观格局对地表温度的解释，分类数据的粒度越小越有利于景观格局对地表温度的解释；另外，格局指数对不同时相（或称季节）的温度的解释度不同，夏季最高，春秋季较低。 （6） 改建的景观空间负荷比指数——位置加权的景观指数LWLI，集成了景观组成和配置于一体，对景观中心温度具有较好的指示作用。 |
| 英文摘要 | Urban heat island (UHI) is a worldwide urban climate phenomenon. Its harm is prevalent in the summer of mid-latitudes areas，with associated effects.How to mitigate urban heat island effects and alleviate its harm is therefore concerned by domestic and foreign scholars and government affiliations.Exploration and evaluation of the impact of landscape pattern on urban heat island not only can provide a theoretical basis for the city (town) planning, but also provide an objective and feasible method as reference to alleviate the urban heat island effect. In this paper, core urban area of Beijing within the Fifth Ring Road was taken as study area. Firstly, High resolution remote sensing (RS) images, the Quickbird images acquired in 2002 and the IKONOS images acquired in 2012 image was used as a basis for landscape classification. The urban area was treated as a two-dimensional landscape. Classification was done by a combination of unsupervised classification with a decision tree method. The study area was classified into four land cover types: water, vegetation, bare soil and built-up areas,or known as impervious surface. Also, a small part of the study area was further classified into 15 subtypes through visual interpretation for the need of analysis. These classification results were used as the basis for the landscape pattern analysis. Secondly, Landsat5 and 7 TM / ETM+ data was used to retrieve land surface temperature (LST), which was the basis for the UHI analysis; Finally, relationship between landscape pattern and UHI was analyzed first the landscape metrics. The affecting factors during relationship analysis were evaluated, and an alteration of the located weighted landscape index (LWLI) was introduced and utilized to indicate the urban heat island effect. This study systematically evaluated the impact of landscape pattern on surface heat island and the effects of analyzing sizes and confounding factors. The results provide methods references for metrics selection and analysis, scale selection,granularity selection and other landscape pattern analyzing aspects. Additionally, the study provides a new method for pattern analysis and some theoretical basis regarding urban planning.The following conclusions were drawn: (1) Analysis on the landscape pattern shows that: from 2002 to 2012, the impervious surface within the Fifth Ring Road of Beijing has increased by nearly 12%，mainly transformed from bare land ( arable land ); the vegetation decreased by 3% and water hardly changed. The surface heat island analysis shows that in 2002 the total UHI area and the largest UHI area of Beijing in the summer (July) are larger than the other phase, and annual UHI area comparison shows that the total UHI area and the largest UHI area in the summer, 2002 is greater than that in the summer, 2009, which is then greater then that in the summer of 2007, 2011 and 2012. (2) From the two-dimensional landscape (land cover) point of view, the landscape pattern beneath the heat island region is dominated mainly by the impervious surface. In the core area of Beijing, a neighborhood with impervious surface proportion reaching or exceeding 84% , tends to form a heat region with an intensity of 8 ℃ in summer, compared to the whole study area。 (3) The correlation between the landscape level metrics and LST is significantly affected by proportion (PLAND) of impervious surface; Both composition and configuration at class level has effects on mean LST and center LST of the landscape unit, and composition affects much more effects than configuration. (4) Impervious surface pattern can explain center LST of the landscape analyzing unit in summer up to 58%, of which the composition (PLAND) of the generally explains about 53%, and the configuration 5%; Impervious surface pattern can explain mean LST of the landscape analyzing unit in summer up to 79% of which its composition (PLAND) explain it is generally around 73%, while the configuration 6%. (5) Interpretation of landscape pattern on the surface temperature varies with the method of landscape pattern analysis, landscape classification data, the different date or time of LST data. During regression analysis between class level pattern and LST, circular analyzing units resulted in better regression coefficients than quadrate units; a small-scale analyzing units resulted in better regression coefficients than large-scale analyzing units; higher classification accuracy, or finer resolution classification data also resulted in better regression coefficients. Regression coefficients for different time (or season) temperatures and class level metrics vary,too. The higher coefficients exist in the summer, and lowest exist in spring and. (6) The modified landscape index LWLI - the location of landscape indices weighted, has integrated landscape composition and configuration in one, and can well predict central LST of the landscape unit. |
| 公开日期 | 2015-06-16 |
| 源URL | [http://ir.rcees.ac.cn/handle/311016/13478]  |
| 专题 | 生态环境研究中心_城市与区域生态国家重点实验室 |
| 推荐引用方式 GB/T 7714 | 陈爱莲. 景观格局对城市地表热岛效应的影响研究--以北京市为例[D]. 北京. 中国科学院研究生院. 2014. |
入库方式： OAI收割
来源：生态环境研究中心
浏览0
下载0
收藏0
其他版本
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。