Skip to content

Are there thermal precursors to eruptions detectable by satellite? Evaluating 22 years of global medium resolution satellite thermal observations at 195 erupting subaerial volcanoes

Matthew E Pritchard1, Andrea Gomez-Patron 1, Hugh Peng2, Abigail Downes2, Chloe Crothers3, Jiada Valenza4, Austin Wada5,6, Lin Way7 and Laura Zapata8

  • Affiliations: 1Cornell University, Earth and Atmospheric Sciences, Ithaca, NY, United States, 2Department of Earth & Atmospheric Sciences, Cornell University, Ithaca, United States, 3Lund University, Lund, Sweden, 4Hunter College, City University of New York, New York City, United States, 5Cornell University, Ithaca, United States, 6San Francisco State University, San Francisco, United States, 7School of Earth Sciences, University of Bristol, Bristol, United Kingdom, 8University of California Los Angeles, Los Angeles, United States

  • Presentation type: Poster

  • Presentation time: Monday 16:30 - 18:30, Room Poster Hall

  • Poster Board Number: 265

  • Programme No: 2.4.39

  • Theme 2 > Session 4

Abstract

Temperature changes at volcanoes can reveal trends that could anticipate eruptions. Using nighttime Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) observations with medium spatial resolution (90 m/pixel), we manually collect thermal measurements for 195 subaerial volcanoes with 810 eruptions from 2000-2022 to evaluate thermal precursors. This is a larger sample of volcanoes than possible with ground-based thermal sensors (55 volcanoes) or detected by lower spatial resolution satellite imagery. Roughly 22% of the eruptions have sufficient cloud-free observations to construct statistically significant time series. On average, each volcano has 3 thermal features, so there are 532 individual time series.  We use our Deviation From Thermal Time Series Baseline (DTTSB) program to identify temperatures that deviate from previous values and compare them to eruption times from the Global Volcanism Program to find precursors. There are 99 instances (19%) of increase in temperature before eruption, 27 instances (5%) of decrease in temperature, 58 instances (11%) of increase then decrease trends, and 348 instances (65%) where there is sufficient data, but no precursor to eruption. Thermal precursors start between one week and five years before an eruption; however, most precursors last less than a year.  We find that all kinds of thermal precursors are present in all volcanic and eruptive characteristics. We also find that the size and duration of the precursor do not indicate size of the subsequent eruption. Finally, we show several examples where more temporally dense cloud-free satellite observations at spatial resolution of 90 m/pixel or better are needed.