The Canadian wildfires of June 2023 exposed a large portion of the Northeastern United States to unprecedented levels of smoke. A new model that combines wildfire smoke forecasts and data from ground-based sensors may help public health officials plan targeted interventions in areas most at risk for the negative health effects of unexpected smoke events and air pollution, according to a team led by Penn State scientists.
The researchers reported their findings in the journal Science of the Total Environment.
“Statistical analyses suggest that situations like last year’s Canadian wildfires, where smoke travels long distances to affect the Eastern United States, may become the norm,” said lead author Manzhu Yu, assistant professor of geography at Penn State.
“Our research can help public health officials in urban and rural areas plan targeted interventions for communities at higher risk of harmful air pollution during wildfire smoke events.”
The researchers focused on the periods between June 6–8 and June 28–30 2023, when weather conditions and a coastal storm pushed large amounts of smoke from Canada into the Northeastern United States. They used data from ground-based sensors and a form of artificial intelligence called deep learning to improve a weather forecasting model from the National Center for Atmospheric Research.
The model—the Weather Research and Forecasting model with Chemistry, or WRF-Chem—provides hourly data on surface concentrations of fine particulate matter (PM 2.5). Found in wildfire smoke and other forms of air pollution, these tiny particles can reach the lungs and cause health issues.
The scientists also studied anonymized mobility data from devices like smartphones to see how people changed their travel activities during the smoke events. Additionally, they conducted an environmental justice assessment using data from the U.S. Environmental Protection Agency to see if certain environmental and demographic factors correlated to increased vulnerability to negative health outcomes from wildfire smoke.
These factors included variables like percentage of the population with less than a high school education, minority status, heart attack and asthma hospitalization rates, and existing pollution burdens from sources like heavy traffic and power plants. They studied these factors at the county level, from Pennsylvania and New Jersey up through Maine, to see if certain communities shared a larger part of the pollution burden than others.
The team found that the refined forecasting model better estimated the magnitude and timing of PM 2.5 spikes, measured in micrograms per cubic meter of air (µg/m3), across the study area than the current forecasting model.
When looking at how predicted data matches observed data, with 0 µg/m3 of PM 2.5 signifying that the model exactly matches ground observations, the current forecasting model scored a -6.872 µg/m3, marking a large underestimation of particulate levels. The refined model scored a 0.160 µg/m3, marking a slight overestimation of particulate levels that aligned much closer to what the ground sensors measured.
In addition, the researchers found that urban and rural communities already burdened by existing environmental pollution face higher air pollution levels during unexpected smoke events than other areas.
“The good news, according to our findings, is that when people hear about wildfire smoke, they tend to reduce their mobility,” Yu said.
“But we found that during these smoke events New York City, Philadelphia and the surrounding counties still showed high mobility activities. We probably need to think about targeted interventions in urban areas because with so many people living in the area, exposure rates to unhealthy air are very high.”
Rural communities burdened by pollution from power plants and mines may have particular needs as well, she said. For example, she explained, Bennington County, Vermont, has few demographic factors that would make it more vulnerable to environmental pollution.
However, it is home to multiple mines, heavy traffic, hazardous waste storage sites and more, which all contribute to higher environmental pollution scores. Those factors amplified air pollution levels during the smoky days.
“Public health interventions are usually based on population concentrations, which are naturally higher in urban areas,” Yu said. “Knowing these existing vulnerabilities in rural areas can help officials better serve these areas and protect the public’s health.”
In the meantime, individuals can take steps now to protect their health during the upcoming wildfire season.
“I would suggest that individuals have an air filter and indoor air pollution monitor in their homes,” Yu said. “They can also enhance the insulation around their windows and doors if smoke levels are really high. I would recommend working from home if possible or getting a high-quality mask if you have to travel outdoors.
“And I think in Pennsylvania, we need to talk about standards for organizations for how we respond to smoke days, whether that’s working from home, having a day off or dismissing early. We’re not used to smoke events, and we need some sort of policy or standard for protecting the public’s health.”
In addition to Yu, contributors to this research from Penn State include Zhenlong Li, associate professor of geography, and doctoral students Shiyan Zhang and Huan Ning; and Kai Zhang, Empire Innovation Associate Professor at the University of Albany’s School of Public Health.
More information:
Manzhu Yu et al, Assessing the 2023 Canadian wildfire smoke impact in Northeastern US: Air quality, exposure and environmental justice, Science of The Total Environment (2024). DOI: 10.1016/j.scitotenv.2024.171853
Provided by
Pennsylvania State University
Citation:
Improved wildfire smoke model identifies areas for public health intervention (2024, May 9)
