West J Emerg Med. 2012 Aug;13(3):239-46. doi: 10.5811/westjem.2012.3.11746.
To investigate relationships between ambient temperatures and violent crimes to determine whether those relationships are consistent across different crime categories and whether they are best described as increasing linear functions, or as curvilinear functions that decrease beyond some temperature threshold. A secondary objective was to consider the implications of the observed relationships for injuries and deaths from violent crimes in the context of a warming climate. To address these questions, we examined the relationship between daily ambient temperatures and daily incidents of violent crime in Dallas, Texas from 1993-1999.
We analyzed the relationships between daily fluctuations in ambient temperature, other meteorological and temporal variables, and rates of daily violent crime using time series piece-wise regression and plots of daily data. Violent crimes, including aggravated assault, homicide, and sexual assault, were analyzed.
We found that daily mean ambient temperature is related in a curvilinear fashion to daily rates of violent crime with a positive and increasing relationship between temperature and aggravated crime that moderates beyond temperatures of 80°F and then turns negative beyond 90°F.
While some have characterized the relationship between temperature and violent crime as a continually increasing linear function, leaving open the possibility that aggravated crime will increase in a warmer climate, we conclude that the relationship in Dallas is not linear, but moderates and turns negative at high ambient temperatures. We posit that higher temperatures may encourage people to seek shelter in cooler indoor spaces, and that street crime and other crimes of opportunity are subsequently decreased. This finding suggests that the higher ambient temperatures expected with climate change may result in marginal shifts in violent crime in the short term, but are not likely to be accompanied by markedly higher rates of violent crime and associated increased incidence of injury and death. Additional studies are indicated, across cities at varying latitudes that experience a range of daily ambient temperatures.
为了研究环境温度与暴力犯罪之间的关系，以确定这些关系是否在不同的犯罪类别中具有一致性，以及它们是否最好描述为线性递增函数，或者是否描述为超过某个温度阈值后呈曲线递减的函数。次要目标是考虑在气候变暖的背景下，观察到的这些关系对暴力犯罪所致伤害和死亡的影响。为了解决这些问题，我们研究了 1993 年至 1999 年期间德克萨斯州达拉斯市每日环境温度与每日暴力犯罪事件之间的关系。
我们使用时间序列分段回归和每日数据图分析了环境温度日波动与其他气象和时间变量以及每日暴力犯罪率之间的关系。分析的暴力犯罪包括严重攻击、杀人罪和性侵犯。
我们发现，日平均环境温度与每日暴力犯罪率呈曲线关系，温度与加重犯罪之间呈正相关且递增关系，在 80°F 以上温度时逐渐减弱，在 90°F 以上温度时转为负相关。
虽然有人将温度与暴力犯罪之间的关系描述为持续增加的线性函数，从而为在更温暖的气候下加重犯罪将会增加的可能性留出了余地，但我们的结论是，达拉斯的这种关系不是线性的，而是在高温下减弱并转为负值。我们假设，较高的温度可能会促使人们在较凉爽的室内空间寻求庇护，街头犯罪和其他机会犯罪随后会减少。这一发现表明，气候变化预计会导致环境温度升高，短期内可能会导致暴力犯罪略有转移，但不太可能伴随着暴力犯罪率明显上升以及相关伤害和死亡发生率的增加。需要在不同纬度的城市进行更多的研究，以涵盖不同的日环境温度范围。
