Evaluation of the mortality risk reduction effects of countermeasures to nocturnal UHI

Heatstroke mortality during nighttime sleep is expected to rise as global warming and urban heat islands (UHI) become more prevalent. Measures that can reduce nighttime temperatures include greening and anthropogenic heat reduction. However, it is difficult to compare health impacts using the same temperature index because the peak time period when temperature reduction is most effective, varies greatly between measures. Therefore, this study proposed a daily temperature index that can consider health effects by time of day and used this index to compare and evaluate the mortality risk reduction effects of measures on nocturnal UHI.

When the human body is exposed to a certain temperature, health effects can last for days or even weeks. This study used a distributed lag nonlinear model (DLNM) to assess the relationship between temperature indices and mortality risk taking into account the lag effect. The study was conducted in the Tokyo metropolitan area from 2000 to 2010, using data from demographic surveys on air temperature and mortality. First, a mortality risk prediction model was developed using the DLNM, with the temperature every hour of the day as the explanatory variable, and the optimal temperature with the lowest mortality risk was determined for each hour. Next, the mean deviation between the optimal temperature and the observed temperature at each time of day was calculated for each of three time periods (2–9, 10–17, and 18–1 h), and the daily weighted average deviation ΔT_day was calculated using the combination of weight coefficients that minimizes the Quasi Akaike Information Criterion (QAIC) of the forecast model. This combination (0.67, 0.13, and 0.20) suggests that the temperature deviations have a five-fold greater impact at night (2–9 h) than during the daytime (10–17 h). The model with ΔT_day had a lower QAIC and higher prediction accuracy than models with daily maximum or daily mean temperatures.

Three countermeasure scenarios for nocturnal UHI were compared and evaluated: ground greening (doubling the area of green space), electric vehicles (with 100% penetration), and air-source heat pump (HP) water heaters (with 100% penetration) that emit cold exhaust heat at night. The evaluation target area included the entire 23 wards of Tokyo, and the daily ΔT_day was estimated on a mesh-by-mesh basis using an urban canopy model to compare and evaluate heat-related excess mortality risk reduction rates in August. The results showed that the three scenarios had similar mortality risk reduction effects for the entire area, but their effects had different geographic distribution characteristics. The HP water heater scenario showed particularly large impact in residential areas with large nighttime populations.