Long Enough? Uncertainty-driven Metrics for Modeling Air Quality Health Impacts under Climate Change

Abstract:

There are many factors to consider when simulating health impacts of air pollution under a changing climate. Uncertainties related to projecting climate and health systems make it difficult to develop robust insights about the effects of climate policy in protecting public health. Here, we consider several sources of uncertainty and develop relevant metrics to inform simulation lengths. Specifically, we use a large ensemble of simulations of the future economy, climate, air quality, and human health impacts to project mortality and morbidity associated with the so-called “climate penalty” on ozone and fine particulate matter at mid and end of century. This multi-decadal, multiple initial condition ensemble reflects variation due to natural variability and different levels of climate policy ambition. Across U.S. regions, we find that health incidence rates and associated economic damages vary significantly. So, too, do demographics, health characteristics, and the natural variability in the climate. Consequently, in some regions, relatively few simulation years are needed to adequately address natural variability. Five years or less can suffice to significantly reduce this source of variability across multiple mitigation scenarios and time periods. In other regions, it may take a decade or more to reliably distinguish the effects of a policy. These insights can provide guidance for modeling, communicating, and reducing uncertainty in the future health impacts of air pollution under a changing climate.