Abstract:
Communities across the western United States experience hazardous smoke exposures from multiple fire sources. As wildfires become more frequent and severe, smoke exposures stemming from these fires are also worsening. Prescribed burning is a promising forest management strategy that can mitigate future wildfire risk, but also contributes to biomass burning emissions and human exposure impacts. Agricultural burning is another commonly used management tool, implemented on agricultural landscapes across the West, but also contributes to ambient air pollution. Despite these multiple sources of smoke exposure, few studies have examined their differential exposure impacts, particularly among communities most at-risk. This dissertation seeks to address this gap, by characterizing smoke exposures from each of these fire types in the recent past and by examining how these exposures may change due to the implementation of new forest management strategies in the future. First, we generate a fire type-specific 1 km biomass burning emissions inventory, using the Fire INventory from NCAR (FINN) and a series of federal and state-level fire and fuel treatment inventories to distinguish between wildfire, prescribed, and agricultural burn emissions across Washington, Oregon, and California. We then use that emissions inventory to model surface-level PM2.5 concentrations and population-level exposures, using the GEOS-Chem atmospheric chemical transport model at a 0.25° x 0.3125° resolution from 2014-2020. We identify distinct spatiotemporal exposure patterns for each fire type, which differentially impact population sub-groups within states. For example, we observed disproportionately higher exposures to wildfire smoke among Native communities and higher exposures to agricultural burn smoke among lower socioeconomic groups in California. Next, we specifically focus on the relationship between wildfires and prescribed burns to assess the exposure and health impacts of six forest management scenarios proposed for a 2.4 million acre landscape in the Central Sierra, California. Using wildfire and prescribed burn emission estimates generated using a landscape forecasting model, we modeled fire type-specific smoke exposure impacts using the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model at a 27 km resolution. Among the general population, we estimated that moderate amounts of prescribed burning can reduce wildfire-specific smoke exposures and asthma-related health impacts. We observed a similar pattern when examining exposure impacts among outdoor agricultural workers in California, in which total smoke exposure is lowest under scenarios that call for moderate amounts of prescribed burning; however, we also observe a decreasing exposure benefit under scenarios that call for greater amounts of prescribed burning due to the smoke contributions from the fuel treatment themselves. Together, this two-part analysis describes the distinct exposure patterns from different types of fire on the landscape and what role management can play in reducing exposure burdens. The results of this dissertation emphasize the need for more tailored exposure reduction strategies that consider the source of smoke. Additionally, it highlights the importance of increased collaboration between public health and natural resource management agencies in a way that can optimize the achievement of management objectives, while simultaneously minimizing harmful exposure burdens among at-risk communities.