Land Surface Temperature Variation and Major Factors in Beijing, China

Land surface temperature (LST) is a significant parameter in urban environmental analysis. Current research mainly focuses on the impact of land-use and land-cover (LULC) on LST. Seldom has research examined LST variations based on the integration of biophysical and demographic variables, especially for a rapidly developing city such as Beijing, China. This study combines the techniques of remote sensing and geographic information system (GIS) to detect the spatial variation of LST and determine its quantitative relationship with several biophysical and demographic variables based on statistical modeling for the central area of Beijing. LST and LULC data were retrieved from a Landsat Thematic Mapper (TM) image. Building heights were delimited from the shadows identified on a panchromatic SPOT image. The integration of LULC and census data was further applied to retrieve gridbased population density. Results indicate that the LST pattern was non-symmetrical and non-concentric with high temperature zones clustered towards the south of the central axis and within the fourth ring road. The percentage of forest, farmland, and water per grid cell were found to be most significant factors, which can explain 71.3 percent of LST variance. Principal component regression analysis shows that LST was positively correlated with the percentage of low density builtup, high density built-up, extremely-high buildings, low buildings per grid cell, and population density, but was negatively correlated with the percentage of forest, farmland, and water bodies per grid cell. The findings of this study can be applied as the theoretical basis for improving urban planning for mitigating the effects of urban heat islands. Introduction Land surface temperature (LST) is an important parameter in study of urban thermal environment and behavior. LST modulates the air temperature of the lower layer of Land Surface Temperature Variation and Major Factors in Beijing, China Rongbo Xiao, Qihao Weng, Zhiyun Ouyang, Weifeng Li, Erich W. Schienke, and Zhaoming Zhang urban atmosphere, and is a primary factor in determining surface radiation and energy exchange, the internal climate of buildings, and human comfort in the cities (Voogt and Oke, 1998). The physical properties of various types of urban surfaces, their color, the sky view factor, street geometry, traffic loads, and anthropogenic activities are important factors that determine LSTs in the urban environments (Chudnovsky et al., 2004). The LST of urban surfaces correspond closely to the distribution of land-use and land-cover (LULC) characteristics (Lo et al., 1997; Weng, 2001 and 2003; Weng et al., 2004). Each component surface in urban landscapes (e.g., lawn, parking lot, road, building, cemetery, and garden) exhibits unique radiative, thermal, moisture, and aerodynamic properties and relates to their surrounding site environment (Oke, 1982). The myriad of these component surfaces and the spatial complexity when they are mosaicked create a limitless array of energy balance and microclimate systems (Oke, 1982). To study urban LSTs, some sophisticated numerical and physical models have been developed. These include energy balance models (Oke et al., 1999; Tong et al., 2005), laboratory models (Cendese and Monti, 2003), three-dimensional simulations (Saitoh et al., 1996), Gaussian models (Streutker, 2003), and other numerical simulations (Yang et al., 2003). Among these models and simulations, energy balance models are by far the main methods, but statistical analysis may play an important role in linking LST to related factors, especially at larger scales (Bottyán and Unger, 2003). Previous studies have focused primarily on biophysical and meteorological factors, such as built-up area and height (Bottyán and Unger, 2003), urban and street geometry (Eliasson, 1996), LULC (Dousset and Gourmelon, 2003), and vegetation (Weng et al., 2004). A less number of studies, however, have examined how population distribution influences urban heat island (UHI) intensity, although it is apparently an indicator of anthropogenic heat emission (Fan and Sailor, 2005) and the intensity of human activities (Elvidge et al., 1997). Little research effort has been made to study the intraurban variations of LST and their related biophysical and socioeconomic variables within a city. The combination of biophysical and socioeconomic data for urban studies has been hampered by the spatial unit problem, because the relevant spatial units for biophysical processes are different from the spatial units for population and other PHOTOGRAMMETRIC ENGINEER ING & REMOTE SENS ING Ap r i l 2008 451 Rongbo Xiao is with the National State Key Lab of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China ([email protected]). Zhiyun Ouyang and Weifeng Li are with the National State Key Lab of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, 100085 China. Qihao Weng is with the Department of Geography, Geology, and Anthropology, Indiana State University, Terre Haute, IN. Erich W. Schienke is with the Department of Science and Technology Studies, Rensselaer Polytechnic Institute, Troy, NY. Zhaoming Zhang is with the China Remote Sensing Satellite Ground Station, Chinese Academy of Sciences, Beijing, China. Photogrammetric Engineering & Remote Sensing Vol. 74, No. 4, April 2008, pp. 451–461. 0099-1112/08/7404–0451/$3.00/0 © 2008 American Society for Photogrammetry and Remote Sensing