The advances draw on a combination of publicly and privately sourced satellite data together with machine learning algorithms developed at Pitt in collaboration with a researcher in Italy. The findings appear in the Journal of Volcanology and Geothermal Research.
Ian Flynn, research assistant professor in the Department of Geology and Environmental Science in Pitt's Kenneth P. Dietrich School of Arts and Sciences, was working in the lab of Professor Michael Ramsey during the 13-day eruption. As privately launched satellite data was becoming increasingly accessible to researchers, Ramsey asked whether those new sources could be combined with traditional government satellite feeds to improve eruption monitoring. Flynn took on the challenge and was able to track the lava flow front in real time as it advanced toward 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."
To improve early warning capability, Flynn collaborated with Dr. Claudia Corradino of the Italian National Institute of Geophysics and Volcanology (INGV) to apply a machine learning algorithm to thermal satellite data from Mauna Loa. The algorithm identified a measurable thermal increase one month before the eruption began. Although this signal was recognized retrospectively, the methodology establishes a framework for detecting similar precursory signals ahead of future eruptions.
"Every volcano has its own personality," Flynn said. "Yes, it's cheesy, but it's the truth. They're all different."
Flynn then sought to extract additional information from the satellite record, specifically the thickness of the lava flow itself. He reached out to Dr. Shashank Bhushan at NASA's Goddard Space Flight Center, who had developed analogous techniques for measuring glacier thickness. "I reached out and asked, 'can we use this methodology that you apply to glaciers and adapt it to lava flows?'" Flynn said. "He said, 'I don't know. Let's try.'"
The adaptation proved successful, providing a three-dimensional view of the eruption to complement the two-dimensional spatial tracking. "Getting visible data helped us understand where it's going," Flynn said. "Now we can also generate flow thickness and understand how much material is coming out." That volume data is critical for determining whether an eruption is intensifying or winding down.
The thermal component of the data serves additional purposes. Active flows that remain hot continue to emit hazardous gases, posing risks to anyone who might approach. Knowing precisely when lava cooled also allows researchers to more accurately characterize its chemical composition, since compositional analysis depends on the thermal history of the material.
The research also has direct implications for planetary science. Venus, which hosts numerous volcanic features, is of particular interest. Understanding how quickly lava cools under Earth's atmospheric conditions gives scientists a baseline for interpreting thermal signatures detected on other worlds. "Knowing how lava cools enables scientists to better constrain our models when we find active volcanoes on other planets," Flynn said.
The team expects that as additional satellite data continues to accumulate, understanding of the 2022 Mauna Loa event will deepen further, and lessons learned will inform monitoring strategies at other volcanoes globally. Because each volcano exhibits distinct behavior, there is unlikely to be a universal predictive model. Instead, Flynn envisions tailored monitoring systems built around the unique eruptive history and thermal characteristics of individual sites. Mauna Loa may be among the most active volcanoes on Earth, but many others pose comparable or greater risks to nearby populations and stand to benefit from similar data-fusion approaches.
Research Report:Satellite data synergy for volcano monitoring: The 2022 Mauna Loa eruption
Related Links
University of Pittsburgh
Bringing Order To A World Of Disasters
When the Earth Quakes
A world of storm and tempest
| Subscribe Free To Our Daily Newsletters |
| Subscribe Free To Our Daily Newsletters |
