BACKGROUND Investigating future changes in temperature-related mortality as a function of global mean temperature (GMT) rise allows for the evaluation of policy-relevant climate change targets. So far, only few studies have taken this approach, and, in particular, no such assessments exist for Germany, the most populated country of Europe. METHODS We assess temperature-related mortality in 12 major German cities based on daily time-series of all-cause mortality and daily mean temperatures in the period 1993-2015, using distributed-lag non-linear models in a two-stage design. Resulting risk functions are applied to estimate excess mortality in terms of GMT rise relative to pre-industrial levels, assuming no change in demographics or population vulnerability. RESULTS In the observational period, cold contributes stronger to temperature-related mortality than heat, with overall attributable fractions of 5.49% (95%CI: 3.82-7.19) and 0.81% (95%CI: 0.72-0.89), respectively. Future projections indicate that this pattern could be reversed under progressing global warming, with heat-related mortality starting to exceed cold-related mortality at 3 °C or higher GMT rise. Across cities, projected net increases in total temperature-related mortality were 0.45% (95%CI: -0.02-1.06) at 3 °C, 1.53% (95%CI: 0.96-2.06) at 4 °C, and 2.88% (95%CI: 1.60-4.10) at 5 °C, compared to today's warming level of 1 °C. By contrast, no significant difference was found between projected total temperature-related mortality at 2 °C versus 1 °C of GMT rise. CONCLUSIONS Our results can inform current adaptation policies aimed at buffering the health risks from increased heat exposure under climate change. They also allow for the evaluation of global mitigation efforts in terms of local health benefits in some of Germany's most populated cities.

中文翻译：

---

2°C和全球变暖程度更高时，德国城市中与温度相关的过度死亡率。

背景技术调查与温度有关的死亡率的未来变化作为全球平均温度（GMT）上升的函数，可以评估与政策相关的气候变化目标。到目前为止，只有很少的研究采用这种方法，特别是对于欧洲人口最多的国家德国，尚无此类评估。方法我们采用两阶段设计中的分布滞后非线性模型，基于1993-2015年期间全因死亡率的每日时间序列和每日平均温度，评估了德国12个主要城市的温度相关死亡率。假设人口结构或人口脆弱性没有变化，则将所得风险函数用于估计相对于工业化前水平而言的GMT上升带来的超额死亡率。结果在观察期，寒冷对温度相关的死亡率的影响要强于热量，总归因分数分别为5.49％（95％CI：3.82-7.19）和0.81％（95％CI：0.72-0.89）。未来的预测表明，在全球变暖加剧的情况下，这种模式可能会逆转，在3°C或更高的GMT上升时，与热相关的死亡率开始超过与冷相关的死亡率。在各个城市中，预计在3°C时与温度相关的总死亡率净增加分别为0.45％（95％CI：-0.02-1.06），在4°C时为1.53％（95％CI：0.96-2.06）和2.88％（ 95％CI：1.60-4.10）在5°C，而今天的升温水平是1°C。相比之下，在2°C相对于GMT升高1°C时，预计的与温度相关的总死亡率之间没有显着差异。结论我们的结果可以为当前的适应政策提供信息，该政策旨在缓解气候变化下因增加的热量暴露而造成的健康风险。它们还允许根据德国一些人口稠密的城市对当地健康的益处来评估全球缓解措施。