Remote sensing of the urban heat island effect in a highly populated urban agglomeration area in East China

Decheng Zhou; Stefania Bonafoni; Liangxia Zhang; Ranghui Wang

doi:10.1016/j.scitotenv.2018.02.074

Remote sensing of the urban heat island effect in a highly populated urban agglomeration area in East China

Sci Total Environ. 2018 Jul 1:628-629:415-429. doi: 10.1016/j.scitotenv.2018.02.074. Epub 2018 Feb 13.

Authors

Decheng Zhou¹, Stefania Bonafoni², Liangxia Zhang³, Ranghui Wang⁴

Affiliations

¹ Jiangsu Key Laboratory of Agricultural Meteorology, Nanjing University of Information Science and Technology, Nanjing 210044, China. Electronic address: zhoudc@nuist.edu.cn.
² Department of Engineering, University of Perugia, via G. Duranti 93, 06125 Perugia, Italy. Electronic address: stefania.bonafoni@unipg.it.
³ Jiangsu Key Laboratory of Agricultural Meteorology, Nanjing University of Information Science and Technology, Nanjing 210044, China. Electronic address: brightzlx@126.com.
⁴ Jiangsu Key Laboratory of Agricultural Meteorology, Nanjing University of Information Science and Technology, Nanjing 210044, China. Electronic address: rhwang@nuist.edu.cn.

PMID: 29448025
DOI: 10.1016/j.scitotenv.2018.02.074

Abstract

Increasingly urban agglomeration, representing a group of cities with a compact spatial organization and close economic links, can rise surface temperature in a continuous area due to decreasing distance between cities. Significant progress has been made in elucidating surface urban heat island intensity (SUHII) of a single city or a few big cities, but the SUHII's patterns remain poorly understood in urban agglomeration regions. Using Aqua/Terra MODIS data over 2010-2015, we examined the SUHII variations and their drivers in Yangtze River Delta Urban Agglomeration (YRDUA) of east China. Instead of using the widely-used suburban/rural areas as references, this study predicted the unaffected reference temperature wall-to-wall from natural forests by a simple planar surface model. Results indicated that urbanization warmed the land surface regardless of urban area size in YRDUA, with the SUHII clearly larger in the day (2.6±0.9°C) than night (0.7±0.4°C). The SUHII varied markedly by cities, yet the largest did not happen in the presumed core cities. Also, the SUHII differed greatly in a seasonal cycle, with summer-winter difference of 4.2±0.9°C and 2.0±0.5°C in the day and night, respectively. Particularly, cooling effects of urban areas were observed in winter for the majority of cities at night. These spatiotemporal patterns depend strongly on the background climate (precipitation and air temperature), vegetation activity, surface albedo, and population density, with contrast mechanisms during the day and night. Further, we showed that ignoring urban agglomeration effect (using suburban/rural areas as the unaffected references) would lead to large biases of SUHII estimates in terms of magnitude and spatial distribution. Our results emphasize the necessity of considering cities altogether when assessing the urbanization effects on climate in an urban agglomeration area.

Keywords: City clusters; Driving forces; MODIS land surface temperature; Spatiotemporal trends; Urbanization.