When Mauna Loa erupted in 2022, the largest lava flow headed on a path headed directly toward Daniel K. Inouye State Highway 200, also known as Saddle Road, a critical route that carries many residents from their homes on one side to their jobs on the other.
No one could accurately predict whether the lava would continue to flow and eventually block the highway, or stop short, sparing the road.
However, when the volcano next erupts scientists will be better able to monitor the eruption in real-time and make more accurate predictions about where the lava will flow and when the volcano might erupt. These advances are thanks to the availability of satellite data from public and private sources as well as machine learning algorithms developed at Pitt with help from a colleague in Italy, as highlighted in a recent publication in the Journal of Volcanology and Geothermal Research .
During the 13-day Mauna Loa eruption, Ian Flynn, research assistant professor in the Department of Geology and Environmental Science in Pitt’s Kenneth P. Dietrich School of Arts and Sciences, wasworking in the lab of Professor Michael Ramsey.
At the time, more data from privately launched satellites was becoming available to researchers. Ramsey wondered if those new sources could be combined with traditional government satellites to make better predictions. “He asked if I could map the lava flow in real time and actually see the flow-front advancing toward the only road that cuts across the island,” Flynn said.
He could. He was able to watch as the lava made its way toward the Saddle Road. “The concern was that lava was making a beeline toward the road,” Flynn said. “It stopped about 1.5 miles from the road.”
The best way to keep people safe in the event of an eruption, however, is to know as soon as possible before lava begins running down hillsides.
Researchers already knew that increased heat and seismic activity are indicators of an upcoming eruption, but how hot? How much activity? How early? These questions are difficult to answer in general.
Working with a colleague, Dr. Claudia Corradino, from the Italian National Institute of Geophysics and Volcanology (INGV) the team was able to use a machine learning algorithm to identify a thermal increase one month before the start of the eruption. While this signal that an eruption was coming was identified after the eruption ended, any new insights into how a volcano behaves prior to erupting adds to scientists’ ability to predict when they’ll occur for the next eruption.
“Every volcano has its own personality,” Flynn said. “Yes, it’s cheesy, but it’s the truth. They’re all different.” His research has been focused on Mauna Loa for years, trying to decode how those changes relate to its eruptions.
Combining public and private data did just that. But Flynn thought there might be more useful information to extract. Particularly, the thickness of the lava flow. He reached out to Dr. Shashank Bhushan, a colleague working at NASA’s Goddard Space Flight Center.
Bhushan had done similar work with glaciers. “I reached out and asked, ‘can we use this methodology that you apply to glaciers and adapt it lava flows?’ He said, ‘I don’t know. Let’s try.’” It did work, and it gave Flynn and collaborators another tool to understand the eruption.
“Getting visible data helped us understand where it’s going,” Flynn said, but that data is two dimensional. “Now we can also generate flow thickness and understand how much material is coming out.” That information is key to understanding if an eruption has just begun or if it’s waning. It can also be analyzed in terms of the thermal trends to understand how the lava is cooling over time.
“One, if it’s still hot, it’s still a hazard. You don’t want someone walking along something that’s still degassing dangerous chemicals,” he said. And knowing when the lava cooled can help researchers more accurately analyze the lava’s composition.
When we search for active lava flows on other planets, knowing how long it takes for lava to cool on Earth will help us to better understand what’s happening if we see a hot flow on Venus,” he said. Depending on the environmental conditions, rates of cooling should be different. “Knowing how lava cools enables scientists to better constrain our models when we find active volcanoes on other planets.”
As more data becomes available, not only do Flynn and his colleagues continue to learn more about the Mauna Loa eruption, they learn more about the kinds of information they’ll need to know about other volcanoes. There won’t likely be a one-size-fits-all solution to predicting eruptions for all volcanoes, but there may be a way to find a unique solution for predicting eruptions at individual sites.
Mauna Loa may be the most active volcano in the world, but others can be just as—if not more—threatening to people living nearby. Each has its own personality, and each may need its own, tailor-made monitoring system.
Journal of Volcanology and Geothermal Research
10.1016/j.jvolgeores.2026.108603
Computational simulation/modeling
Not applicable
Satellite data synergy for volcano monitoring: The 2022 Mauna Loa eruption
22-Mar-2026