Abstract
This study examines the spatio‐temporal changes of heat risk in South China and the possible effects of urbanization using a dynamic method for classifying urban and rural sites based on time‐varying DMSP‐OLS night‐time light images. Ten different indicators are analysed, which include daily means and consecutive heat wave (HW) measures at 86 stations in Guangdong Province from 1971 to 2014. Results show that these heat risk indicators have been intensifying across the study area, and the intensifying trends are more profound in urbanized and populated areas (e.g., the Pearl River Delta) than less urbanized inland areas. In addition, the mean of daily maximum, minimum, and mean temperatures in urban (rural) areas increased by 0.35 (0.24), 0.21 (0.11), and 0.23 (0.13) °C per decade, and urbanization contributes to 31, 48, and 43% of the total trends, respectively. We also estimate that the urbanization effects on (contributions to) the increasing hot days and warm nights are 2.52 (36%) and 4.21 days per decade (53%). Moreover, urbanization induces an additional increase in the frequency and participating days of HW by 0.44 events (49%) and 3.96 days (51%) per decade, respectively.