城市地域地表温度-植被覆盖定量关系分析——以深圳市为例

Study on land surface temperature vegetation cover relationship in urban region:a case in Shenzhen City

地表温度-植被覆盖间的关系一直是城市热岛研究的热点之一,两者均为描述生态系统特征的重要参数。本文利用深圳市2004年的ETM+影像,基于遥感技术提取相关的下垫面类型、地表温度和植被覆盖等信息,探讨不同下垫面类型对地表温度-植被覆盖关系的影响,并结合分形维度计算方法,比较不同分辨率下地表温度、植被覆盖及其相关关系的变化。研究结果表明,植被覆盖程度与地表温度之间存在明显的负相关,并且在不同的植被覆盖程度下,地表温度-植被覆盖关系呈现分段线性关系。下垫面类型及其组合主要通过植被覆盖的分布对地表温度产生影响。而在不同空间分辨率下(30m至960m),地表温度和植被覆盖的空间变异程度均表现为先升后降,在120m的分辨率下,两者的相关程度达到最高。结果证实区域植被覆盖状况可直接影响辐射、热动力以及土壤水分等多种地表特征,从而导致地表温度分异等。

Studies on urban heat islands （UHIs） have traditionally focused on relationship between vegetation fraction（Fv） and land surface temperature（Ts）, which are important parameters to describe characters of ecosystem. The objectives of this study are： （1） to derive Ts and analyze their spatial variations using Landsat ETM＋ thermal measurements; （2） to apply linear spectral mixture analysis to estimate Fv of urban vegetation abundance; and （3） to investigate the relationship between Ts and Fv derived vegetation abundance using conventional statistics and fractal analysis. Based on examination of a Landsat Enhanced Thematic Mapper Plus （ETM＋） image of Shenzhen city, acquired on February 23, 2004, the landscape classification of Shenzhen is conducted on supervised classification and interactive modification, and the applicability of Fv is investigated as an alternative indicator of vegetation abundance, and Ts is derived from ETM＋ thermal infrared （TIR） data. Toexamine the spatial patterns of Ts and Fv, 28 transects （profiles） are drawn across through the center of the city on the image. Since these transects pass over various landscapes with different environmental settings, an inquiry into the fractal characteristics of the profiles using the divider method will help to understand the factors shaping the city＇s thermal and vegetative landscape and lead to the development of UHIs. The use of fractals for analyzing TIR images will improve our understanding of the thermal behavior of different land cover types. Results demonstrate that Ts possessed a slightly stronger negative correlation with the Fv for all land cover types across the spatial resolution （30 to 960 m）. Fractal analysis of image texture shows that the complexity of these images increased initially with pixel aggregation and peaked around 120 m, but decreased with further aggregation. Correlations reached the strongest at a resolution of 120 m, which is believed to be the operational scale of Ts and Fv. The spatial variability of texture in Ts is positively correlated with those in Fv. The interplay between thermal and vegetation dynamics in the context of different land cover types leads to the variations in spectral radiance and texture in Ts. It is suggested that the areal measure of vegetation abundance by unmixed vegetation fraction has a more direct correspondence with the radiative, thermal, and moisture properties of the Earth＇s surface that determines Ts.