(Press-News.org) UNIVERSITY PARK, Pa. — Globally, wildfires are becoming more frequent and destructive, generating a significant amount of smoke that can be transported thousands of miles, driving the need for more accurate air pollution forecasts. A team of Penn State researchers developed a deep learning model that provides improved predictions of air quality in wildfire-prone areas and can differentiate between wildfires and non-wildfires.
“As climate change continues to cause ecological changes and challenges, it is likely that wildfire activities will continue to rise,” said Manzhu Yu, assistant professor of geography at Penn State and lead investigator on the project. “Because of this, it is an urgent research priority to accurately predict the concentration of air pollutants induced by wildfire smoke, especially in wildfire-prone areas.”
Wildfire smoke contains a combination of particulate matter and many gaseous pollutants. Fine particulate matter, referred to as PM2.5, has been associated with significant risks to human health and is regulated by the U.S. EPA.
“The fine particulate matter in wildfire smoke can adversely impact human health when the levels are high,” said Yu. “Air quality predictions for fire-prone areas can significantly help emergency managers and public health officials mitigate potentially adverse environmental and public health impacts from air pollution events.”
According to Yu, the team’s new model would be able to warn people sooner about hazardous air quality. The team reported their findings in the journal Science of the Total Environment.
Accurate prediction of air quality, especially for pollutants derived from wildfires, is challenging since it is highly related to the characteristics of the wildfire, such as atmospheric conditions, topography, fuel and moisture, according to the researchers.
“The advantage of this model is that it is able to produce better predictions that can capture the abrupt changes of PM2.5 when wildfire occur, while not underestimating the amount of PM2.5 that is present, which other models tend to underestimate,” said Yu. “Likewise, the model does not overestimate PM2.5 when there is no fire.”
The model the team developed is an iteration of an existing deep learning model called "Transformer," which is a sequence-to-sequence model originally proposed for language translation and has been successfully used for time series forecasting. The new model, called ST-Transformer, uses a novel framework that can determine trends associated with wildfires.
Using data from U.S. EPA air quality stations in the greater Los Angeles area, the model was trained to conduct time series forecasting on PM2.5 concentrations. Because air quality stations are sparsely located across large areas and collect data throughout the day, ST-Transformer must consider time and space variables as well as variable-wise dependencies, which are variables that impact each other.
“To train the ST-Transformer model, we included spatial, temporal and variable-wise dependencies of wildfire, smoke and air pollutants,” Yu said. “We also switched Transformer’s full attention mechanism to sparse attention, which can be trained to prioritize and capture the most relevant information. This allows the model to focus on only wildfire-related PM2.5.”
The traditional way to perform this kind of model work for time series forecasting is to train models separately for no-smoke or baseline scenarios as well as smoke scenarios. Then baseline models can be used to predict air pollution for days without wildfire smoke, and the smoke model to predict for days with wildfire smoke, according to the team. But Yu and her team merged these inputs into one model.
“That's where sparse attention comes into play because with sparse attention, you know which inputs will better provide an accurate forecast,” she said. “Sparse attention also provides better estimations of PM2.5, reducing overestimations during baseline times and underestimations when there is a fire.”
Yu said ST-Transformer could also be used to improve predictions in other fields, such as water quality, precipitation and solar radiation.
“ST-Transformer could be helpful to predict water quality issues where you have stations in the water, but you cannot control where you can place them,” Yu said. “For example, the nitrogen or phosphorus in runoff that can cause algal blooms are like wildfires because they are fluid and dynamically changing all the time. Your sensors may or may not capture that.”
Other Penn State researchers on the project are Christopher Blaszczak-Boxe, former associate research professor of geosciences and now at Howard University, and Arif Masrur, former graduate student in geography who earned his doctorate in 2021 and is now a senior solutions engineer at Esri.
Research was supported by the College of Earth and Mineral Sciences’ Wilson Faculty Fellowship.
New model provides improved air-quality predictions in fire-prone areas
ELSE PRESS RELEASES FROM THIS DATE:
New study finds early warning signs prior to 2002 Antarctic ice shelf collapse
UNIVERSITY PARK, Pa. — In 2002, an area of ice about the size of Rhode Island dramatically broke away from Antarctica as the Larsen B ice shelf collapsed. A new study of the conditions that led to the collapse may reveal warning signs to watch for future Antarctic ice shelf retreat, according to a Penn State-led team of scientists. “The collapse of the Larsen B ice shelf is generally thought of as an independent event,” said Shujie Wang, assistant professor of geography at Penn State. “Our ...
Oncotarget | MTAP loss in metastatic breast cancer patients: Genomic landscape
“In breast cancer, MTAP downregulation activates ornithine decarboxylase (ODC) which in turn leads to formation of putrescine which promotes tumor migration, invasion and angiogenesis .” BUFFALO, NY- March 14, 2023 – A new research paper was published in Oncotarget's Volume 14 on March 11, 2023, entitled, “Genomic landscape of metastatic breast cancer (MBC) patients with methylthioadenosine phosphorylase (MTAP) loss.” Homozygous deletion of methylthioadenosine phosphorylase (MTAP) upregulates de novo synthesis ...
Tim Michalski selected as Jefferson Lab’s engineering manager
NEWPORT NEWS, VA – The U.S. Department of Energy’s (DOE) Thomas Jefferson National Accelerator Facility has selected Tim Michalski to lead its Engineering Division as the Engineering Division Manager. In this role, Michalski oversees all aspects of the management and operation of the Engineering Division. The division includes more than 200 staff members and supports the Continuous Electron Beam Accelerator Facility and the lab’s growing project portfolio. “I am proud to name Tim as our newest Engineering Division Manager,” said Jefferson Lab Director Stuart Henderson. “Tim has ...
Princeton Chem, IAS uncover spatial patterns in distribution of galaxies
In late 2021, Salvatore Torquato, on sabbatical from Princeton’s Department of Chemistry, reached across the aisle as it were and invited a young astrophysicist at the Institute for Advanced Study to apply the tools of statistical mechanics to his own work on the distribution of galaxies. The astrophysicist, Oliver Philcox, now a postdoc at the Simons Foundation, was intrigued. A year-long collaboration ensued. The questions at the heart of their unusual partnership were straightforward: can the statistical descriptors Torquato has worked with throughout his career find application in unlikely places like cosmology, and can they accurately characterize the complexity in the distribution ...
Researchers look to AI for decision-making in extreme situations
Imagine you are a doctor managing the emergency room of a large hospital. You suddenly get a call reporting a mass shooting at a nearby concert. In 20 minutes, you will be responsible for triaging more than 200 patients with a range of injuries. You do not have enough staff or resources and the hospital policies are not designed for a situation this dire. “When people respond to emergencies, many decisions they face are quite predictable. They’re trained on them, and there’s policy,” said UMass Lowell’s Neil Shortland, associate professor in the School of Criminology and Justice Studies. “But every now and then, they get stuck with a really ...
Mix-and-match kit could enable astronauts to build a menagerie of lunar exploration bots
When astronauts begin to build a permanent base on the moon, as NASA plans to do in the coming years, they’ll need help. Robots could potentially do the heavy lifting by laying cables, deploying solar panels, erecting communications towers, and building habitats. But if each robot is designed for a specific action or task, a moon base could become overrun by a zoo of machines, each with its own unique parts and protocols. To avoid a bottleneck of bots, a team of MIT engineers is designing a kit of universal ...
Common dry cleaning chemical linked to Parkinson’s
A common and widely used chemical may be fueling the rise of the world’s fastest growing brain condition – Parkinson’s disease. For the past 100 years, trichloroethylene (TCE) has been used to decaffeinate coffee, degrease metal, and dry clean clothes. It contaminates the Marine Corps base Camp Lejeune, 15 toxic Superfund sites in Silicon Valley, and up to one-third of groundwater in the U.S. TCE causes cancer, is linked to miscarriages and congenital heart disease, and is associated ...
Molecular component of caffeine may play a role in gut health
Brigham researchers studying how and why certain cell types proliferate in the gut found that xanthine, which is found in coffee, tea and chocolate, may play a role in Th17 differentiation Insights may help investigators better understand gut health and the development of conditions such as inflammatory bowel disease The gut is home to a cast of microbes that influence health and disease. Some types of microorganisms are thought to contribute to the development of inflammatory conditions, such as inflammatory bowel disease (IBD), but the exact cascade of events that leads from microbes to immune cells to disease remains ...
Ochsner Health announces new Aortic Center; subscribes to cutting-edge imaging with Cydar Technology
NEW ORLEANS, La. – Global death rates from aortic diseases have steadily increased over the past two decades. Studies show up to 8% of individuals will develop an aortic aneurysm during their lifetime, higher than rates of colon and lung cancer combined, and over 80% of ruptured aortic aneurysms cause sudden death. To save lives and improve outcomes for patients with aortic disease of all kinds, Ochsner Health is excited to announce the establishment of The Ochsner Aortic Center. Outfitted with cutting-edge imaging technology that allows medical staff to make faster, easier, and safer decisions, this dedicated, comprehensive aortic center is now the ...
Aston Pharmacy School researchers develop new technique mixing oil and water to improve drug delivery
A team of researchers from Aston University has developed a new technique that could be a game changer for the medical and drinks industries. The technology enables insoluble drug/oil to be dissolved in water. The technique is novel because it doesn’t just mix the two together to make an emulsion, it makes oil soluble in water and has the potential to revolutionise a variety of medical treatments and improve drug delivery. The technique was developed with Max Bio+ a spin out company founded by Professor Sunil Shah, a consultant ...