Project Summary Climate change is the greatest public health challenge of the 21st century. While numerous pathways of the health impact of climate change have been proposed, the “climate penalty” effect, i.e., a warming temperature worsens ambient air quality and consequently influences human health, remains poorly understood, resulting in an underestimated public health burden associated with global warming. Our previous epidemiological studies have reported that higher summer mean temperatures and higher PM2.5 concentrations are each associated with increased all-cause mortality in the Medicare population (aged ≥65) in the Southeastern US (SEUS)1, 2. Satellite and ground-based observations suggest a strong dependence of air pollution on interannual variabilities of summer mean temperature in SEUS3. These findings suggest that the indirect health effect of temperature via the climate penalty on air quality can be potentially important in the SEUS region, in addition to the direct adverse effects that we observed. However, clear epidemiological evidence of the air pollution serving as a mediator for the health effects of temperature, and accurate estimate of this effect is still missing in current literature. Herein, drawing on our preliminary results, we hypothesize that rising temperature can indirectly affect all-cause mortality via worsening both PM2.5 and ozone levels in the SEUS. We propose a study that will leverage the Medicare cohort from 2000-2016 (124 million person-years), the largest longitudinal cohort available for the SEUS and the high-resolution temperature, PM2.5, and O3 data, to investigate all-cause mortality in response to the “climate penalty” effect using a mediation statistical analysis. Specifically, in this project we will (1) update the present- day temperature and ozone predictions at 1-km2 grids across the SEUS through 2016 by incorporating ensemble averaging of machine learning models; (2) quantify the health effect of “climate penalty” on all-cause mortality using a mediation analysis, and explore whether mitigating anthropogenic air pollution emissions might serve as a pathway of climate change adaptation; (3) perform a risk assessment on the excess deaths related to the climate penalty on air pollution for the mid- (2050) and late-21st century (2100), using climate model output, chemical transport modeling, along with the top-down estimate of “climate penalty” from Aim 2. The proposed research will improve understanding of the interplays between climate, air pollution, and human health based on real-world big data, and provide epidemiological evidence of an important pathway that climate change adversely affects human health, with immediate relevance to climate and environmental policymaking.