基于MODIS日地表温度产品的兰州市热岛效应特征

由于MODIS 8 d地表温度产品和Landsat热红外波段很难准确反映地表温度的日变化，这导致由此得到的热岛效应结果不够准确，因此利用MODIS日地表温度产品1天4次的数据来分析长时间段上的热岛效应变化和已有将日地表温度转换成月和年地表温度的算法得到月和年地表温度值，以兰州市为实验区分析了2002-2015年兰州市的热岛变化特征，结果表明：兰州市热岛效应总体上处于增长趋势，2002-2009年波动变化，2010-2015年平稳增加。空间上由中部逐年向北部永登县和皋兰县方向扩展；兰州市热岛的月变化特征分为两个部分，1～5月呈“人”字形变化；6～12月整体下降，空间上表现为东北和西南向条带中心紧缩；兰州市热岛季节变化表现为：夏季>春季>秋季>冬季，空间上在兰州的中心位置以及延伸至榆中县的东南角高温区面积增减。研究表明MODIS日地表温度产品能更加准确地评价区域尺度热岛效应的演变特征。

Characteristics of urban heat island based on MODIS daily land surface temperature product in Lanzhou

Urban area has a great impact on the regional environment,the urban heat island (UHI)is one of them. In this paper,considering the problem of spatio-temporal resolution in current research of the UHI,the MODIS daily land surface temperature (LST)products (MOD11A1 and MYD11A1)are used to study the spatio－temporal variation of the UHI during 2002 to 2015 in Lanzhou,China. Firstly,by the existing methods,the daily average LST is obtained from the data that MODIS gets four times a day. The main details are filling the null value caused by the cloud cover according to the Kriging interpolation and the time-series. Secondly,normalized LST and pixel trend regression are used to analyze the spatio-temporal characteristics of UHI. The results show that as follows: (1)The annual change of the UHI in Lanzhou has distinctive characteristics according to different stages which can be divided into 2002-2009 and 2010-2015. The former fluctuated and the latter increased steadily. But in general,the UHI still has grown over the past ten years. (2)Heat island of Lanzhou is mainly distributed in the central river valley and low-lying areas. And it extends northward to the Yongdeng County and Gaolan County from the center year by year. The spatial features of the UHI are positively correlated with the topography. (3)The characters of the Monthly variation of the UHI in Lanzhou can be divided into two parts. The first one,it's in a shape of“T”during January to May,and the peak is in March. The second one,it declines overall from June to December. It shows that high temperature area narrows down from northeast and southwest to the center strip. (4)the seasonal variation of the UHI shows:summer >spring >autumn >winter. And it shows that high temperature area increases and decreases in the central location of Lanzhou and the southeast corner Extending to Yuzhong County. Comparing with 8-day LST product and Landsat image,the LST calculates by daily LST product and annual average LST algorithm can reflect the diurnal change of the regional-scale land surface temperature more accurately. And it also can evaluate the evolutionary characteristics of the regional-scale UHI more accurately. However,due to the limitation of spatial resolution,there are still many deficiencies in describing the details of the UHI. Therefore,the improving of the spatial resolution is expected in the further study on the UHI.