Skip to content

Semi-automated volcano monitoring using distributed MOLA seafloor sensor networks

Jens Karstens1, Fabian Steinmetz1, Everardo Gonzalez1, Swantje Wegehaupt1, Raphael Herges1, Axel Berger1, Thies Bartels1, Olaf Landsiedel2, Tayyaba Zainab1,2, Christian Renner3, Roberto Benavides1,  Christian Berndt1, Heidrun Kopp1

  • Affiliations: 1GEOMAR Helmholtz Centre for Ocean Research Kiel, Kiel, Germany; 2Kiel University, Kiel, Germany; 3Hamburg University of Technology, Hamburg, Germany

  • Presentation type: Poster

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

  • Poster Board Number: 167

  • Programme No: 2.2.31

  • Theme 2 > Session 2

Abstract

Marine volcanic eruptions, landslides and associated tsunamis pose a significant threat to coastal communities and seabed infrastructure. While most volcanic eruptions and landslides are preceded by precursory processes that provide effective early warning for terrestrial volcanoes, no dedicated tools and approaches exist for the vast majority of submarine volcanoes. Recent advances in embedded and distributed machine learning, and the availability of low-cost underwater communications technology, allow for the development of intelligent sensor platforms, such as the compact MOLA seafloor lander system[HK1] . The MOLA prototype can be equipped with a variety of sensors, including seismometers, hydrophones, inclinometers, accelerometers, pressure gauges, thermometers, cameras and chemical sensors, and consists of a multi-channel data recorder, an underwater acoustic communication device (ahoi modem) and a central analysis and computation unit. Multiple MOLA landers can form seafloor sensor networks that perform embedded and distributed data analysis, such as earthquake detection and localisation, or acoustic quantification of flow rates from hydrothermal vents. The network shares data and analysis results via acoustic communication, while on-node data analysis and event detection using thresholding or neural network-based data filtering significantly reduces data transfer requirements and enables (semi-)autonomous system assessments. These could, for example, indicate increased eruption probability or evidence of slope deformation, and could be uploaded to central observation facilities or early warning systems with little delay via gateway communication nodes at the sea surface. The MOLA system has already been tested offshore Etna, in the Santorini caldera and on Kolumbo volcano.