Publications

Sun, YW; Gao, C; Li, JL; Li, WF; Ma, RF (2018). Examining urban thermal environment dynamics and relations to biophysical composition and configuration and socio-economic factors: A case study of the Shanghai metropolitan region. SUSTAINABLE CITIES AND SOCIETY, 40, 284-295.

Abstract
Multi-scale assessment of urban thermal environment dynamics and its influencing factors have been considered an essential precondition for mitigation and regulation of urban heat island (UHI) effects. However, the annual cycle behavior of satellite-derived urban land surface temperature (LST) is still unclear due to the limitation of irregular and infrequent satellite LST data. This study investigated annual dynamics of LST and the UHI effects of Shanghai by using an annual temperature cycle (ATC) model to reconstruct the Moderate-resolution imaging spectroradiometer (MODIS) 8-day LST data with 1 km(2) resolution in 2015. Ordinary least squares (OLS) and spatial regression models were further applied to investigate the relationships between day/night LST and urban biophysical composition and configuration and socio-economic characteristics in the metropolitan area of Shanghai, China. The results indicate that ATC model performed well with overall root mean square errors (RMSE) of 2.6 K. The performance of ATC is better in night-time than that in day-time. The thermal gradient analysis showed the urban center was hotter than outskirts by up to 2.0 K in day and 0.8 K in night. Spatial patterns of UHI intensity between day-time and night-time had significant differences in central urban areas. We also observed higher amplitude of the annual temperature cycle in central urban areas than that in rural. Furthermore, our statistical model showed biophysical indicators and land composition had a stronger influence on LST than the other explanatory variables. Nighttime light composite data from the Visible Infrared Imaging Radiometer Suite (VIIRS) on board Suomi National Polar-orbiting partnership (NPP) as a spatial proxy for socioeconomic factors had significant positive relationships with mean LST. Compared with OLS and spatial lag model (SLM), spatial error model (SEM) is more appropriate to predict the urban LST. We conclude that distinct control strategies between day-time and night-time may improve efficiency of attenuating the UHI for a metropolitan region, and the spatial and temporal elements should be considered in landscape and urban planning.

DOI:
10.1016/j.scs.2017.12.004

ISSN:
2210-6707