Contrasting impacts of heat stress on violent and nonviolent robbery in Beijing, China

Abstract

Previous studies investigating the relation between heat stress and crime incidents often focus on violent crimes. In this study, the impacts of heat stress on two types of robbery (violent and nonviolent) in China are compared using crime statistics collected in Beijing and heat stress indices that consider the combined effects of temperature and humidity. The results indicate that the abrupt change in the trend of robbery rates is affected by the 2008 Beijing Olympic Games. The nonviolent robbery rates have a more pronounced seasonality and are better correlated with heat stress at daily scales, especially during the period from 2009 to 2014 when no trend exists. The results also demonstrate that both violent and nonviolent robbery rates significantly increase with heat stress in spring. The nonviolent robbery rates also significantly increase with heat stress in summer. The influence of heat stress on violent robbery rate is more complicated and nonlinear.

Appendix

The heat stress indices used in this paper are the discomfort index (DI), the simplified wet-bulb globe temperature (sWBGT), the Humidex (HUMIDEX) and the temperature–humidity index for comfort (THIC). Their definitions are given in Table 2. The risk levels associated with different heat stress values are shown in Table 3 (Buzan et al. 2015).

Table 2 Heat stress indices

Table 3 Risk levels of different heat stress indices

In order to estimate the heat stress indices sWBGT and HUMIDEX, the vapor pressure e _RH (Pa) is firstly calculated by:

$$e_{\text{RH}} = ({\text{RH}}/100)e_{{s\text{Pa} }}$$

(3)

The saturated vapor pressure e _sPa is calculated using Magnus form approximation:

$$e_{{s\text{Pa} }} = 6.1078\exp \left[ {\frac{17.13(T - 273.16)}{T - 38}} \right] \times 100$$

(4)

where T is the temperature in Kelvins.

In this paper, to calculate heat stress indices THIC and DI, it is necessary to first calculate the wet-bulb temperature T _w (°C). Here, we use an approximation from (Oleson et al. (2015)) to calculate the wet-bulb temperature:

$$\begin{aligned} T_{\text{WS}} = T_{\text{C}} \arctan \left( {0.151977\sqrt {{\text{RH}} + 8.313659} } \right) + \arctan \left( {T_{\text{C}} + {\text{RH}}} \right) - \arctan \left( {{\text{RH}} - 1.676331} \right) + 0.00391838{\text{RH}}^{3/2} \arctan \left( { 0. 0 2 3 1 0 1 {\text{RH}}} \right) - 4.68035 \hfill \\ \end{aligned}$$

(5)