未来能源研究所-气候变化对冷热相关死亡率的影响（英文）-2021.5-43页_398kb

报告摘要

Summary of "Effects of Climate Change on Heat- and Cold-Related Mortality: A Literature Review to Inform Updated Estimates of the Social Cost of Carbon"

Core Content

This literature review examines the effects of climate change on heat- and cold-related mortality, with the aim of informing updated estimates of the social cost of carbon (SCC). The SCC is a metric that quantifies the economic damages caused by each additional ton of carbon dioxide emissions in dollar terms, and it plays a key role in shaping climate policy decisions. The review highlights the importance of understanding the temperature-mortality relationship, the role of adaptation, and the challenges in comparing studies across different regions and methodologies.

Main Points

Temperature-Mortality Dose-Response Function

The relationship between temperature and mortality is not linear and varies by region.
Dose-response functions can take different shapes: U-shaped, J-shaped, or inverted J-shaped.
U-shaped function: Mortality increases at both extreme cold and extreme heat, with the lowest risk at moderate temperatures.
J-shaped function: Mortality increases with rising temperatures, with the lowest risk at lower temperatures.
Inverted J-shaped function: Mortality increases with falling temperatures, with the lowest risk at higher temperatures.
Studies suggest that in many regions, especially high-income areas, cold-related mortality has historically been higher than heat-related mortality, but this may change with increased warming.

Regional and Population Variations

The shape of the temperature-mortality dose-response function varies across regions and populations.
Certain groups, such as children, the elderly, and rural populations, are more vulnerable to temperature changes.
The majority of existing studies are based on data from wealthier regions and cities, which may limit the generalizability of findings to lower-income and rural areas.

Future Adaptation

Adaptation strategies, such as air conditioning and migration to warmer areas, can significantly reduce mortality risks.
However, the costs and effectiveness of adaptation are not well captured in most studies.
A few studies, such as Carleton et al. (2020), attempt to quantify adaptation costs using a revealed preference approach.

Challenges in Comparing Research Results

Studies differ in how they quantify mortality risks: some report relative risk ratios, others estimate excess deaths or life years lost.
Dose-response functions are estimated using different temperature thresholds and reference points.
Many studies use the same RCP scenarios but do not report warming levels in terms of degrees Celsius, making cross-study comparisons difficult.
Monetization of mortality damages is not consistently applied across studies, with some using the value of a statistical life (VSL) or life years lost.

Key Components of an Ideal Dose-Response Function

Globally representative: Should capture temperature-mortality relationships across all regions using empirical methods.
Account for future adaptation: Include estimates of how populations will adapt through technological, behavioral, and physiological changes.
Include full age and socioeconomic distributions: Reflect how different demographic and economic groups are affected by temperature changes.
Incorporate both heat and cold effects: Cover all temperature ranges, not just extreme events.

Summary of Key Studies

Study	Region(s)	Method	Population	Temperature	Adaptation	Period(s) Examined	Key Results
Ahmadalipour & Moradkhani (2018)	Middle East, Med., most of Africa	Process-based	Elderly (>65)	Heat	Historical (assumed)	1951–2005, 2006–2100	Relative risk ratio for RCP 4.5 and 8.5 was 3–7 and 8–20 times higher than the historical baseline, respectively.
Alahmad (2019)	Kuwait	Empirical	All	Heat and Cold	Historical (assumed)	2010–2016	Mortality-minimizing temperature was 34.7°C (66th percentile). Relative risk of mortality was 1.65 at the 99th percentile (heat) and 1.67 at the 1st percentile (cold).
Anderson & Bell (2009)	United States	Empirical	All	Heat and Cold and Heat Waves	No	1987–2000	Increasing temperature from the 90th to 99th percentile increased mortality risk by 3.0% and 4.2%, respectively.
Anderson & Bell (2010)	United States	Empirical	All	Heat Waves	No	1987–2005	Mortality risk increased by 2.49% per 1°F increase in heat wave intensity, with the highest impacts in the Northeast.
Anderson et al. (2018)	United States	Empirical, process-based	All	Heat Waves	Yes (quantitative)	1981–2005, 2061–2080	High-mortality heat waves increased from ~0.25 per year to 1.8–2.4 and 2.2–3.8 per year under RCP 4.5 and 8.5, respectively.
Arbuthnott et al. (2016)	Global	Literature review	All	Heat and Cold	Yes (qualitative)	Various	n/a
Armstrong et al. (2019)	Global	Empirical	All	Heat	Historical (assumed)	1972–2015	A 23% increase in relative humidity decreased mortality by 1.1%. Results were sensitive to lag times.
Åström et al. (2013)	Sweden	Empirical	All	Heat and Cold	Historical (assumed)	1900–1929, 1980–2009	Extreme cold/heat increased mortality risk by 5.6% and 4.6%, respectively. In Stockholm, 75 lives were lost due to cold extremes and 288 due to heat extremes.
Åström et al. (2011)	Global	Literature review	Elderly	Heat Waves	No	Various	n/a
Barreca (2012)	United States	Empirical, process-based	All	Heat and Cold	Yes (quantitative)	1973–2000, 2010–2100	Air conditioning reduced heat-related mortality by 75% in the United States.

Conclusion

The literature suggests that climate change will likely increase global mortality in the long term, especially in warmer and low-income regions, while cooler, higher-income regions may experience some reduction in cold-related mortality. However, the studies show significant uncertainty, particularly in regions with limited data or where adaptation is not well quantified. The ideal dose-response function for updating the SCC should account for global coverage, future adaptation, and demographic and socioeconomic factors, but such a function remains elusive due to methodological and data limitations.

展开完整摘要