Volcano Monitoring with High-resolution satellite SAR

  • Mark J W Bemelmans

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)

Abstract

Volcanoes can present a danger to nearby people and infrastructure through eruptions, ash,
landslides, lahars, and pyroclastic density currents. Sufficiently large eruptions can even
impact the global economy. As such, monitoring for any unrest signs becomes important
to limit their negative impact. Surface motion is one of the most frequently observed precursors
to volcanic eruptions and can be monitored effectively using space-borne Synthetic Aperture
Radar (SAR) sensors. SAR volcano monitoring has grown substantially over the last 2-3 decades.
However, high-resolution SAR data are not commonly used. These data are crucial in detecting
and understanding small-footprint deformation linked to (often shallow) volcanic processes. In
this thesis, I use high-resolution SAR data to investigate two small-footprint and shallow volcano
deformation signals, 1) intra-crater deformation at Agung in 2017, and 2) flank motion at Merapi
during 2020-2021. At Agung, I find that the deformation is linked to the pressurisation of the
shallow (≤150 m depth) hydrothermal system in response to a deeper dyke intrusion. At Merapi,
displacements reach >25 m and cannot be estimated reliably with Interferometric SAR (InSAR).
Therefore, I use (Sub-)Pixel Offset Tracking (SPOT) to retrieve the large-magnitude displacements.
I improve upon the basic SPOT method by applying outlier detection and removal with an absolute
median difference filter and averaging over offset estimates from multiple window sizes to create
precise and robust displacement estimations. At Merapi, the measured displacements show sliding
of several stacked planes on the western and northwestern flank of the volcano. The displacement
rate is modulated by volcanic and seismic activity. I find that shallow (precursory) processes,
invisible to medium- or low-resolution data can provide insights into shallow volcanic unrest and
other volcanic processes at depth, thus showing the effectiveness of high-resolution SAR data. I,
therefore, argue for an increased uptake in high-resolution SAR usage for volcano monitoring.
Date of Award8 Jan 2025
Original languageEnglish
Awarding Institution
  • University of Bristol
SupervisorJuliet J Biggs (Supervisor), James M Wookey (Supervisor) & Michael P. Poland (Supervisor)

Cite this

'