Heat stress changes and impacts under global warming target temperatures

Abstract

This thesis evaluates the impacts of Paris Agreement target temperatures on the characteristics of heat stress based on the wet-bulb globe temperature (WBGT) which combines temperature and humidity. Potential benefits of summer heat stress reduction by global warming mitigation are analyzed and possible differences in heat stress characteristics between transient and equilibrium worlds are quantitatively assessed over 24 selected low to mid-latitude land regions. First, future changes in summer mean WBGT and summer frequency and maximum duration of daily extreme WBGT events are examined using the large-ensemble historical (2006-2015) and future simulations (1.5, 2, and 3 degree warmer conditions) from six atmospheric global climate models participating in the “Half a degree Additional warming, Prognosis and Projected Impacts” (HAPPI) project. Results show that the future increases in summer mean WBGT and global warming mitigation benefits (difference between 2.0 and 1.5 degree simulations) are similar across the regions. This is found to be due to the high humidity over the low latitude wet regions, which compensates relatively low temperature increases. On the other hand, the summer extreme heat stress changes are stronger in the low latitude regions, indicating larger potential benefits of global warming mitigation. Importantly, weaker daily variability of WBGT over the tropics is found to play a critical role in the larger increases in extreme heat stress frequency and duration. The population exposure to extreme heat stress is further analyzed based on extreme heat stress duration. Results indicate that the half degree additional warming will increase the population that would be exposed to extreme heat stress (where historical 50-year return level becomes 2-year return level) to approximately 2,498 million (48.3%) over the analyzed domain, dominated by heavily populated regions such as East Asia and South Asia. The 3.0°C warmer worlds (based on two models that provide data) would witness stronger and widespread increases in heat stress frequency and duration, resulting in most of the population experiencing extreme heat stress. Second, possible differences in future changes to heat stress characteristics between transient and equilibrium world assumptions are examined based on CMIP5 multi-model simulations. For this analysis, RCP2.6 and RCP4.5 scenario runs are selected for equilibrium world condition, which project near 1.5°C and 2.0°C warming in the late 21st century. Then corresponding 30-year periods of RCP8.5 scenario runs with the same global warming are used for transient world condition. Comparison results show that differences in land-sea contrast and regional forcing like aerosols lead to differences in heat stress changes between the two assumptions. The summer mean and extreme WBGT over the tropical and mid-latitude lands generally increase faster in the transient world than in the equilibrium world due to the intensified land-sea contrast, which is largely supported by the results based on CO2 forcing only simulations. When analyzing contributions of temperature and humidity changes to heat stress responses, a few regions of distinct behaviors are identified , including East Asia, South Asia, and North & Central Australia. East Asia exhibits a weaker increase in the transient world than in the equilibrium world, particularly for the 1.5°C warmer world. This local slowdown in heat stress increase in the transient world is found to be associated with the strong aerosol emissions that would remain by the mid-21st century. South Asia is also affected by the aerosol cooling effect, but the relative humidity rather increases related to cooling effect and dynamic factors, offsetting the aerosol-induced WBGT decrease. For North & Central Australia, a difference in precipitation response importantly affects temperature and humidity changes. The weaker precipitation decrease in the transient world than in the equilibrium world induces a relative cooling and also a relative wetting, which cancel each other, resulting in a weak difference in heat stress characteristics. A comparison of population exposure shows similar results such that generally more people will be influenced by extreme heat stress in the transient world than in the equilibrium world but that the opposite is expected to occur in East Asia due to the difference in aerosol cooling effect. Overall, results from this thesis indicate that understanding the changes in relative humidity and the aerosol cooling effects is critical to better predicting the regional-scale changes in heat stress under different global warming target temperatures.