3.1 Background

High temperature is a natural hazard linking with excess mortality and morbidity (Kovats and Hajat 2008). High temperature can induce heat stroke, communicable disease, cardiovascular disease, and respiratory disease (Basu and Ostro 2008; Hajat et al. 2005). Individuals with limited mobility due to physical disability, or those who are medically fragile will be subject to greater risks (Díaz et al. 2006). The influence of high temperature on mortality has been found in China during the past. Additionally, the frequency and intensity of high temperature are projected to increase along with the global warming in the future in China. Therefore, the risks that high temperature imposes to human health have recently been a hot topic (Kovats and Hajat 2008). Using five models data, we assessed risks of mortality in high temperature under four future emission scenarios (RCP2.6, RCP4.5, RCP6.0, and RCP8.5) (Fig. 3.1).

Fig. 3.1
figure 1

Flow chart of assessing the risk on mortality in high temperature in the future scenarios

Full size image

3.2 Method

3.2.1 High Temperature Days and Heat Waves

The high temperature day is defined as the day with temperature above 35 °C. The heat wave is defined as the weather that continues for at least three days, during which the maximum daily temperature is above the temperature of 95th percentile in the period of 1981–2010. Another criterion for heat waves is that the daily temperature does not fall below 32 °C. In this chapter, the mean number of days with temperature exceeding 35 and 40 °C, heat waves, and the maximum temperature of a year was calculated and mapped.

3.2.2 Population Vulnerability

This study uses the mortality increment (M i ) in each degree centigrade above the threshold temperature to represent the population vulnerability on high temperature. The threshold temperatures and M i were obtained from relationships between mortality and daily highest temperature in summer of China. The model between M i and threshold temperature was made through following steps (Fig. 3.1).

Firstly, we searched literatures on the mortality in high temperature by the keywords such as “weather,” “temperature,” “mortality,” and “heat” in the major Chinese and English databases. Secondly, we selected the literatures, which studied the relationship between population mortality and the daily maximum temperature in summer, presenting a mortality threshold temperature or a curve on mortality and high temperature to judge the mortality threshold temperature and its increment value for each 1 °C above the threshold temperature. The relation as reported in the literatures was summarized in Table 3.1.

Table 3.1 The relation between mortality and maximum daily temperature in summer
Full size table

Based on Table 3.1, the relationship curves between M i and threshold temperatures is built, and the model equation is given in Eq. 3.1

$$ \begin{aligned} M_{i} & = 5179210 - 794104.82539 \times T_{v} + 47413.08018 \times T_{v}^{2} - 1431.3192 \times T_{v}^{3} \\ & \quad + 21.571523 \times T_{v}^{4} - 0.1298413 \times T_{v}^{5} \\ \end{aligned} $$
(3.1)

where M i is the mortality increment for each 1 °C above the threshold temperature, T v refers to the threshold temperature on population mortality, and the range value T v is between 30 and 38 °C. The R 2 of the fitting curve is 0.53, and the p is less than 0.01.

Based on Table 3.1, the relationship curves between the threshold temperature and mean temperature was also fitted for cities in summer during the period of 1961–2010. The threshold temperature was calculated in China according to the relation model of Eq. 3.2.

$$ T_{v} = 0.5664\bar{t} + 19.52 $$
(3.2)

where T v refers to the threshold temperature, \( \bar{t} \) is the mean temperature of the city in summer. The R 2 of the fitting curve is 0.60, and the p is less than 0.01.

3.2.3 Population Mortality Hazard

This study employed the value of exceeding threshold temperature to represent the hazard intensity for population in high temperature. The hazard was calculated according to Eq. 3.3.

$$ H_{d} = \sum\limits_{i = 1}^{7} {(T_{i} - T_{v} )} \times P_{i} $$
(3.3)

where H d refers to the heat hazard index, T i represents the temperature above the threshold temperature on mortality (°C), T v refers to the threshold temperature of population mortality (°C), and P i refers to probability caused by the daily maximum temperature above the threshold. The hazard indices were divided into ten ranks from one (low) to ten (high) by the natural break method.

3.2.4 Population Exposure

In this study, the population of every 1 × 1 km grid cell in 2000s, 2030s, 2060s, and 2090s was used to represent the exposure of population suffered from high temperature. This data was obtained from the Greenhouse Gas Initiative (GGI) of the International Institute for Applied Systems Analysis (IIASA: http://www.iiasa.ac.at). The data include projections under three scenarios, namely, A2, B1, and B2. We selected the data of A2 and B1 to represent the scenarios of RCP8.5 and RCP2.6.

3.2.5 Population Mortality Risk

Based on the risk theory, the mortality risk for each county was estimated using Eq. 3.4.

$$ POP_{d} = 365 \times H_{d} \times M_{i} \times \frac{{E \times D_{r} }}{92} $$
(3.4)

where POP d refers to the population mortality risk of a year in different periods (2011–2040, 2041–2070, and 2071–2099). M i is the mortality increment for each 1.0 °C above the threshold temperature (%). D r represents the population mortality rate during the summer of each county. E is the population exposure to high temperature. The mortality risk is divided into ten ranks one (low) to ten (high) by the natural break method.

3.3 Results

3.3.1 Spatiotemporal Distribution of High Temperature and Heat Waves

The number of days exceeding 35 °C, 40 °C, and heat waves are all increasing during the different periods under the RCP2.6, RCP4.5, RCP6.0, and RCP8.5 scenarios. When compared with the pre-2040s increase, a more speedy increase is projected after the 2040s.

The values of all variables on high temperature are almost all higher under the RCP8.5 than the RCP2.6 during the different periods. The areas of the north, center, and northwest of China are severely suffered from high temperature.

3.3.2 Spatiotemporal Distribution of Population Mortality Hazard

The mortality hazard on high temperature is the highest in the RCP8.5, and the lowest in the RCP2.6 in China. The increment trend of mortality hazard will be developed from north to south in China during the year 2011–2100. Under all the four scenarios, the highest hazard occurs in the northwest followed by the north and the mid and lower reaches of the Yangtze River.

3.3.3 Spatiotemporal Distribution of Population Mortality Risk

The population mortality risk is the highest in the middle of North China under the four emission scenarios. The area of the mid-lower reaches of the Yangtze River is the area with the second highest risk. The risks of other areas from high to low are southwest, southeast, northeast, and northwest of China.

3.3.4 Uncertainty Analysis in High Temperature Over 2071–2099 for RCP8.5

The various uncertainty of high temperature projection is analyzed by intermodal spread of temperature changes. The projected numbers of days with temperature above 35 °C have high uncertainty in southeast China. The uncertainty of projected number of days with temperature above 40 °C is largest in the middle reaches of the Yangtze River, while the uncertainty is lowest in northeast China. The uncertainty of the maximum temperature in a year is higher in northwest China than those in other regions. For the heat wave projection, the high uncertainty mainly distributes in the southeast and north of China. The higher uncertainty of population mortality hazard is found in the northwest and the middle reaches of the Yangtze River, while the northeast of China has lower uncertainty. The uncertainty of population mortality risk is high in the north China and the middle reaches of the Yangtze River.

3.4 Maps