New Satellite Integration and Machine Learning Improve Volcanic Eruption Predictions for Mauna Loa and Beyond

University of Pittsburgh researchers use machine learning and satellite data to track lava flows and predict eruptions, with new implications for Venus.

By: AXL Media

Published: Apr 28, 2026, 5:23 AM EDT

Source: Information for this report was sourced from EurekAlert!

The Technological Leap in Hawaiian Lava Tracking

The 2022 eruption of Mauna Loa presented a significant threat to Hawaii’s infrastructure, specifically the Daniel K. Inouye State Highway 200, which serves as a vital artery for residents. According to research assistant professor Ian Flynn, the uncertainty surrounding whether lava would sever this critical route highlighted the urgent need for more precise monitoring. By integrating data from private Planet SuperDoves satellites with traditional government systems like Landsat 8 and Sentinel 2, researchers successfully tracked the advancing flow front as it moved toward the highway. The development eventually stalled 1.5 miles from the road, but the event served as a catalyst for a more sophisticated, data-driven approach to volcanic safety.

Thermal Signals and the Power of Algorithmic Foresight

A collaborative effort between the University of Pittsburgh and the Italian National Institute of Geophysics and Volcanology has introduced machine learning as a predictive tool for geological hazards. By deploying an algorithm developed by Dr. Claudia Corradino, the team retroactively identified a distinct thermal increase 30 days prior to the first signs of the 2022 eruption. While this signal was confirmed after the event, the ability to isolate such subtle shifts in temperature amidst vast datasets offers a new methodology for identifying the "personality" of individual volcanoes. This early detection capability aims to move beyond general indicators of seismic activity toward site-specific warning windows.

Expanding the Dimension of Volcanic Data Streams

Traditional satellite monitoring often focuses on two-dimensional surface mapping, which provides limited information on the actual volume of an eruption. Flynn collaborated with Dr. Shashank Bhushan of NASA’s Goddard Space Flight Center to adapt techniques originally used for measuring glacial thickness to volcanic lava flows. This transition allows scientists to calculate flow thickness and the rate of material discharge, which are critical metrics for determining if an eruption is intensifying or losing momentum. Understanding the mass of the material provides a clearer picture of the hazard level and the potential duration of the threat to surrounding communities.