Search

Under the background of global warming, interaction between heat waves (HWs) and urban heat island (UHI) has led to trends of increase in the intensity, frequency, and duration of extreme heat events in urban areas, seriously threatening the health of urban populations. Taking Guangzhou (a tropical megacity in China) as an example, this study used automatic weather station data and ERA5 reanalysis data to explore the interaction between HWs and UHI, and to elucidate the effects of wind speed and local climate zones (LCZs) on such interaction. Results revealed obvious HWs–UHI interaction in Guangzhou, whereby HWs induces an amplification effect on UHI intensity (UHII) that was most significant at night. In the main urban area, UHII and HWs both weakened with increasing wind speed, indicating that low wind speeds contribute to increased occurrence of HWs and enhancement of UHII. Differently, in some areas peripheral to the main urban area, the UHII at medium wind speeds was stronger than that at low wind speeds, which reflect the impact of heat advection from the urban center. For different LCZs in the main urban area, the strongest UHII, highest risk of HW occurrence, and most significant HWs–UHI interaction were found in the compact mid-rise buildings and compact low-rise buildings (LCZ2 and LCZ3, respectively), followed by the compact high-rise buildings (LCZ1), which was mainly affected by the shading effect of high-rise buildings. The weakest UHII and lowest risk of HW occurrence were found in open high-rise buildings and open mid-rise buildings (LCZ4 and LCZ5, respectively), which generally have good ventilation conditions. Our findings will help to understand urban warming and its association with UHI and HW events in tropical urban regions, which has implications for rational improvement of the urban thermal environment in other tropical urban regions globally.

To investigate the values of 10-m drag coefficient (CD) in different coastal areas under the influence of tropical cyclones, the present study used the observational data from four towers in different coastal areas of the South China Sea (SCS) during six tropical cyclone (TC) passages, and employed the eddy covariance method and the flux profile method. The analysis of footprint showed that the fluxes at Zhizai Island (ZZI), Sanjiao Island (SJI) and Donghai Island (DHI) were influenced basically by the ocean, and the flux at Shangyang Town (SYT) was influenced mainly by the land. The results showed that the dependence relationships of CD on 10-m wind speed (U10) in four different coastal areas under the influence of TCs were different. CD at ZZI and SJI initially increased and then decreased as U10 increased, similar to the pattern over the ocean. CD at ZZI and SJI represented the values over shallow water with seawater depths of ~7 m and ~2 m, respectively. Moreover, the critical wind speed at which CD peaked gradually decreased as the seawater depth became shallower in the coastal areas. CD at DHI and SYT decreased monotonously as U10 increased, similar to the pattern over the land. CD at DHI represented the value over the transition zone from shallow water to coastal land, and CD at SYT represented the value over the coastal land. Meanwhile, the eddy covariance method and the flux profile method were compared at ZZI and SYT during TC passages. It was found that their CD values obtained by the two methods were close. Finally, the parameterizations of observed u* and CD as a function of U10 over four different coastal areas were given under the influence of high winds. These parameterizations of observed CD may be used in high-resolution numerical models for landfalling TC forecast.