基于Landsat8的城市热岛效应与地表因子关系研究——以长春市为例

Estimation of the relationship between urban heat island and land surface factors using Landsat8 data: A case study in Changchun,China

城市化进程导致的城市热岛问题成为21世纪人类需要面对的重要问题之一。文中以长春市为例,使用2014年Landsat8 OLI和TIR数据分析了城市热岛效应与地表因子之间的关系。基于Landat8 TIR数据特点,选用分裂窗算法反演地表温度用于表征城市热环境并分析其空间格局特征。基于"植被-不透水面-土壤"模型,探讨了线性光谱混合模型得到的城市不透水面和植被盖度作为表征城市人为建筑和绿地信息的能力。结果表明:长春市大部分区域有明显的城市热岛效应,占比约45%,其中最显著的区域位于缺少植被覆盖的大片连续分布的不透水面区域;不同城市土地覆被类型的地表温度差异明显,平均地表温度从高到低的土地覆被类型依次为不透水面、裸土、耕地、绿地和水体,各类型平均温度和最高温度的最大温差分别达到8.2和16.1℃;与归一化建筑指数和植被指数相比,光谱解混得到的不透水面比例和植被盖度与地表温度之间的关系更为显著,是表征城市建筑和绿地信息更优的指示因子,其相关关系表明城市建筑能够增加城市热岛效应而城市绿地具有降温效应。

In this research,the relationship between urban heat island and land surface factors was estimated using Landsat8 data in 2014 as a case study in Changchun,northeast China. The split window algorithm was applied to retrieve the land surface temperature（ LST） to characterize the spatial pattern of urban heat island. In addition,based on the ＂ Vegetation-impervious surface-soli＂（ V-I-P） model,this study investigated the applicability of impervious surface and vegetation fraction derived from spectral mixture model to serve as indicators of artificial buildings and urban green space. The results show that： there was no strong urban heat island in most areas of Changchun city. The LST of different land cover types had obvious differences. The impervious surface had the highest LST and was followed by bare soli,farmland,green space and water. Compared to normalized difference build-up index（ NDBI） and normalized difference vegetation index（ NDVI）,the LST possessed a slightly stronger correlation with the impervious surface and vegetation fraction. The positive correlation between LST and impervious surface,NDBI means that the built-up areas could strengthen the effect of urban heat island,while the green space could mitigate the negative effects in this study.