Motivation. The World Health Organization declared COVID-19 a global pandemic on March 11, 2020 just as the southwestern region begins to see increased fire activity. The project PIs had been collaborating on other wildfire projects but also had expertise in infectious disease modeling. We rapidly developed a model of COVID-19 in a single incident to gauge the potential for widespread transmission on moderate to large incidents. We quickly recognized the need for risk assessment tools as well as a model capable of evaluating systemic risks across the entire season. Our project has two primary objectives: 1) design a risk assessment tool applicable to the unique risks posed by the COVID-19 pandemic, and 2) develop an agent-based model of infectious disease transmission across all potential incidents over the course of the season.

Methods. The team developed very different outputs to accomplish the two objectives (i.e., risk assessment tool and seasonal model). The risk assessment tool was designed in an interactive dashboarding environment known as R Shiny. The intention was to create a simple and intuitive decision support tool that could be used by internet-connected devices or printed out and used offline. We took inspiration from existing risk assessment tools but pioneered a new visual presentation of the information. We analyzed usage trends as a measure of value to the wildland fire community. The project's second objective necessitated the development of a novel agent- based model of SARS-CoV-2 transmission across the season. Agent-based models simulate disease transmission between individuals and track the health status of those individuals over the course of the simulation. We based the simulation on observed resource assignment patterns in recent fire seasons. We simulated health interventions' potential health and workforce capacity effects, including vaccination uptake rates and module-as-one.

Key Findings. The risk assessment tool was reviewed by experts and used in the field during the 2020 and 2021 seasons. The PIs received positive feedback over the course of the two seasons, which was corroborated by detailed usage metrics. Further assessment of the risk assessment tool is challenging because we lack systematic data on critical outcomes. The agent-based model yielded several findings. First, the model suggests that off-fire transmission (firefighters off duty) poses a larger risk to workforce capacity and firefighter health than on-fire transmission. Second, workforce capacity depends on individual behaviors both on and off of incidents - high compliance and vaccine uptake can reduce workforce capacity impacts.

Policy Implications. The COVID-19 pandemic posed operational and health risks to the wildland firefighting system. The model and risk assessment tool provided relevant and timely information when little information was available on the potential risks to firefighter health and workforce capacity.