基于Landsat-8热红外数据的太湖地区地表温度反演与热岛效应分析

Retrieval of land surface temperature based on Landsat-8 thermal infrared data and heat island effect analysis over the Taihu Lake region

应用Landsat-8 TIRS遥感数据结合气象资料,使用三种不同的反演方法获得苏南太湖地区的地表温度(LST)数据.分析发现,辐射传输方程法反演的结果最接近于实测值;在缺乏实时大气廓线的情况下,普适性单通道算法更适用于Landsat-8热红外波段的地表温度反演;分裂窗算法在三种算法中精度最差.在反演结果的基础上,分析了不同地物的LST特征,建筑用地温度最高,水体最低.归一化植被指数(NDVI)和LST之间存在明显的负相关关系,证明了城市绿地是缓解城市热岛效应、改善城市热环境的重要途径.最后对2000年和2014年的两时相LST进行标准化和分区处理,分析发现,苏南太湖地区的热岛现象近些年来在不断加剧,城市热岛由原来单个存在逐渐成为多个热岛并存的局面,建筑用地的增加是导致整个城市热岛加剧的主要原因.

Using three different retrieval algorithms based on Landsat-8 thermal infrared data and meteorological data, land surface temperature （LST） of the Taihu Lake region is obtained. After comparing the values of three different algorithms with measured values, we find that the result of radiative transfer equation （RTE） is a good approximation to the measured values. When the real-time atmospheric profile is absent, general single-channel algorithm （SC） is more suitable for land surface temperature retrieval based on Landsat 8 thermal infrared data.The split window algorithm （SW） has the lowest accuracy. After analyzing the LST characteristics of different land cover based on the result of RTE, we find that built-up areas has the highest LST while water body has the lowest LST. There is a negative correlation relationship between the normalized difference vegetation index （NDVI） and the LST values, which indicates that urban green space is a key way to improve urban thermal environment and alleviate urban heat island effect. Finally, after comparing the LST results in year 2000 and 2014, we find that the heat island phenomenon of the Taihu Lake region is increasing in the past 14 years and single existent urban heat islands has gradually become coexistent heat islands now. Building land expansion is the main reason that leads to the increasing of urban heat island.