Geology & Environmental Science Colloquium: Dr. Sophie Pailot-Bonnétat | Postdoctoral Research Associate, University of Pittsburgh
How can AI improve spaceborne thermal volcano monitoring?
Presented by Dr. Sophie Pailot-Bonnétat | Postdoctoral Research Associate, University of Pittsburgh
Talk Abstract:
Volcanic eruptions pose significant hazards to human populations and infrastructure, threatening millions of lives and billions of dollars in economic losses annually worldwide. Thermal precursors often accompany volcanic unrest, making spaceborne thermal infrared (TIR) remote sensing a critical tool for monitoring active and dormant volcanic systems, particularly in remote or inaccessible regions lacking ground-based instrumentation. Satellite TIR sensors can detect a diverse range of volcanic thermal manifestations, including summit crater lakes, lava lakes, active lava flows, dome effusion and extrusion, fumarolic fields, hot springs, diffuse degassing areas, and subtle ground-surface heating associated with hydrothermal alteration. Despite decades of satellite thermal monitoring, traditional detection algorithms often struggle with cloud contamination, atmospheric interference, low-magnitude thermal anomalies, and the sheer volume of data generated by modern sensors, limiting timely hazard assessment. Recent advances in machine learning and deep learning offer new opportunities to overcome these limitations by enabling automated, scalable, and more sensitive detection of thermal anomalies across large volcanic datasets. This presentation includes an overview of ground- and spaceborne unrest characterization at an island arc volcano with persistent hydrothermal activity, the generation of a global database of volcanogenic thermal anomalies using deep learning and the use of machine learning for underwater volcanogenic event tracking.
Biography:
Dr. Sophie Pailot-Bonnétat is a Postdoctoral Research Associate at the University of Pittsburgh’s Department of Geology and Environmental Science. She is part of the Image Visualization and Infrared Spectroscopy (IVIS) Laboratory which is directed by Michael Ramsey. She holds a mining engineering degree from the Polytechnique UniLasalle Institute (France) and a M.S. in Earth Science and Planets with a specialization on magmas and volcanoes from Clermont-Auvergne University (UCA, France), as well as a PhD in Volcanology from UCA.
The IVIS research team focuses on laboratory and remote sensing spectroscopy applied to Earth and planetary volcanological targets to better understand and quantify volcanic processes. Dr. Pailot-Bonnétat exploits thermal anomalies seen from space- and ground-based remote sensing sensors, to monitor volcanic activity and explore volcanogenic geothermal activity as well as mineral resources. She is developing and applying different machine learning and deep learning models to increase the accuracy of these detections. She is also taking part in the design and calibration of a multispectral thermal camera for multipurpose applications, including volcanic gas and pollutant monitoring.