In March and April, firefighters grappled with disastrous blazes in the Wears Valley area west of Great Smoky Mountains National Park. Hundreds of structures were destroyed in the hillside firestorm but there were no reported deaths and few injuries, unlike the epic Thanksgiving wildfires of 2016 that originated in the Smokies and killed at least 14 people in developed areas of nearby Sevier County.
It could get worse, according to Jiafu Mao, an Oak Ridge National Laboratory scientist who works with computational models to gauge potential climate conditions and create more credible wildfire simulations. And it could get very expensive.
His team’s conclusions were recently published in Nature, a highly regarded academic and scientific research journal. While it described enhanced wildfire risks around the world, such as in Australia and Africa, it also included a warning for the Southern Appalachians.
Mao works with a group at ORNL that uses computational Earth system models to simulate various conditions of the planet’s climate and environment.
The model paired regional gross domestic product numbers with areas of enhanced wildfire risk to see what regions would be most likely to see and suffer the effects of increased wildfire activity.
“At least over these mountain areas, our initial results did show an increasing (wildfire) chance in this region,” he said.
The Southern Appalachian region “is projected to be more vulnerable to wildfire changes, but this needs further investigation,” he said.
From the report:
“Through global climate change, human influence on fire ignition, land-use/land-cover change, and complex response of the land biosphere to human-induced climate change and CO2 fertilization, anthropogenic activity has remarkably altered wildfire behavior and its environmental risks at various temporal and spatial scales. These scale-dependent human-fire feedbacks also complicate the future projection of wildfire regimes across the globe (e.g., size, frequency, and intensity, and their socioeconomic impacts),” Mao and fellow researchers concluded.
“Their approach used machine learning and historical data to improve projections from Earth system models in regard to future fire carbon emissions and the associated socioeconomic risks,” according to Sara Shoemaker, media relations specialist at ORNL.
“Technically, that’s the first time we developed a new kind of machine-learning analytical framework of future wildfire emissions … and impacts,” Mao said of the model.
While carbon emissions from wildfires were determined by the model to be potentially less than that of previous forecasting, the economic costs could be dire, especially in the Southern Appalachians.
ORNL senior research scientist Peter Thornton said that though the work is at global scale, researchers found “very interesting local results,” indicating enhanced wildfire likelihood for the Knoxville metro region and its mountainous areas. The study also predicted an increased chance for wildfires in the Western U.S., with accompanying effects on local economies.
The model took into account the natural prevalence of wildfires. Climate change offers a couple of curveballs: Even in the midst of potentially wetter climates, increased temperatures mean moisture may collect in the lower atmosphere in ever-greater amounts, thus reducing humidities and enhancing fire conditions at the ground level. Higher carbon dioxide concentrations might actually spur vegetative growth that could one day power forest fires.
Mao and Thornton emphasized the need for thinning forest fuel stocks in anticipation of increased wildfires, and also establishing real-time monitoring of fire risk and consequences.
“Potential mitigation or adaptation strategies are needed for these particular areas,” including the Southern Appalachians, Mao said.
Those strategies include more clearing and prescribed burns of thick forest, as Great Smoky Mountains National Park has done in the years following the deadly 2016 Thanksgiving wildfires.
The ORNL Earth system model analysis provides “vital information for those who make economic development decisions to consider,” Shoemaker said.
“We can give examples of what mitigation and/or adaptions strategies could be, which could help minimize potential socioeconomic loss caused by wildfires. Specific recommendations would need to come from municipalities and decision-makers,” she said.
“We need more data, and more investigation is needed,” Mao said.
Many climate models don’t yet agree on what areas or regions will get wetter or drier in coming decades.
It will likely take another decade or two to fully suss out the difference between weather and climate, but facts are starting to gel, Thornton said.
Temperature increase is always a consistent player in long-range climate forecasts, and that further leads to an unfortunate, yet workable, conclusion:
“We think we are seeing evidence of the real fire occurrence following the pattern that’s predicted by these models,” Thornton said.
“The conditions for more fire are increasing over time in this region. And that should be the expectation.”