基于Landsat 8影像的福州市地温与土地利用及植被关系研究

Land Surface Temperature in Fuzhou City Based on Landsat 8 Image:Its Relations to Land Use and Vegetation

基于Landsat 8影像数据,根据覃志豪等人提出的单窗算法,分别用2个热红外波段反演2014年4月17日福州市区地表温度,并将反演结果与研究区部分气象站点当天的实测温度进行对比验证,发现利用波段10反演得到的地温与实际观测值更加接近,说明利用单窗算法可以较好地反演福州市地表温度,对于Landsat 8影像来说,波段10更适合用于反演地表温度。本研究采用面向对象分类方法对研究区范围内的土地利用类型进行解译,并计算归一化植被指数(NDVI),进而探讨二者与对应地表温度之间的相关性,结果显示土地利用类型与地表温度具有较强的相关性,同时NDVI与地表温度之间存在着较为显著的负相关性。

Based on Mono-window Algorithm proposed by Qin Zhihao et al, land surface temperature in Fuzhou city was retrieved using two thermal infrared bands from landsat 8 data. Then the results were com- pared with the measured temperature of some meteorological stations in study area, which indicates that in study area, the retrieved land surface temperature with Mono-window Algorithm method can be used to re- trieve land surface temperature and the use of band 10 is much closer to actual observed value. We used the object-oriented classification method to interpret land use types and calculate the normalized difference veg- etation index （NDVI）, with the correlation between the two and the corresponding surface temperature. It was found that land use types had a certain influence on the surface temperature. It also showed that there was a significant negative correlation between NDVI and the land surface temperature.