A multi-perspective study of atmospheric urban heat island effect in China based on national meteorological observations: Facts and uncertainties
Decheng Zhou,Shanlei Sun,Yu Li,Liangxia Zhang,Lin Huang
DOI: https://doi.org/10.1016/j.scitotenv.2022.158638
2022-09-22
Abstract:The atmospheric urban heat island (AUHI) effect, traditionally measured by in-situ sensors mounted on fixed meteorological stations, has been extensively studied by different and imperfect methods. However, facts and uncertainties of the AUHI estimates revealed by the different methods are not well understood at a large scale. Here we examined the spatial-temporal variations of the AUHI effects from multiple perspectives in China's 86 large cities as revealed by national-level meteorological observations at 2-m height from 1981 to 2017. We find relatively consistent patterns of larger urban heating effects in daily minimum temperature, winter, and Northeast China than their counterparts in terms of multiyear mean intensity (AUHII), long-term trend (△AUHII), and contribution to local warming (according to the CTR UMR "urban minus rural" and CTR OMR "observation minus reanalysis" methods). Concurrently, a cooling impact or a reduction in the heating effect has been observed in some cities randomly, especially in daily maximum temperature. On average across cities, the AUHII, △AUHI, CTR UMR , and CTR OMR for the daily mean temperature amount to 0.33 °C, 0.013 °C 10a −1 , 53 %, and 23 % at an annual mean time scale, respectively. Nevertheless, the poor representativeness of weather stations, discrepancies among the quantification methods, nonlinearity of the long-term tendencies, and coupling effects with rural crop land use activities lead to large uncertainties of the AUHI estimates. Our results emphasize the limitations of national-level meteorological stations in characterizing AUHI in China and suggest that the urban heat island remains a "well described but rather poorly understood" phenomenon warranting further investigation by a combined uses of multiple techniques like high-density sensor networks, remote sensing techniques, and high-resolution numerical models.
environmental sciences