From quiescence to unrest: 20 years of satellite geodetic measurements at Santorini volcano, Greece

• Main Authors: Parks, Michelle M., Moore, James D. P., Papanikolaou, Xanthos, Biggs, Juliet, Mather, Tamsin A., Pyle, David M., Raptakis, Costas, Paradissis, Demitris, Hooper, Andrew, Parsons, Barry, Nomikou, Paraskevi
• Other Authors: Earth Observatory of Singapore
• Format: Article
• Language: English
• Published: 2017
• Subjects: InSAR; Joint Inversion
• Online Access: https://hdl.handle.net/10356/86277; http://hdl.handle.net/10220/43967
• Institution: Nanyang Technological University

Periods of unrest at caldera-forming volcanic systems characterized by increased rates of seismicity and deformation are well documented. Some can be linked to eventual eruptive activity, while others are followed by a return to quiescence. Here we use a 20 year record of interferometric synthetic aperture radar (InSAR) and GPS measurements from Santorini volcano to further our understanding of geodetic signals at a caldera-forming volcano during the periods of both quiescence and unrest, with measurements spanning a phase of quiescence and slow subsidence (1993–2010), followed by a phase of unrest (January 2011 to April 2012) with caldera-wide inflation and seismicity. Mean InSAR velocity maps from 1993–2010 indicate an average subsidence rate of ~6 mm/yr over the southern half of the intracaldera island Nea Kameni. This subsidence can be accounted for by a combination of thermal contraction of the 1866–1870 lava flows and load-induced relaxation of the substrate. For the period of unrest, we use a joint inversion technique to convert InSAR measurements from three separate satellite tracks and GPS observations from 10 continuous sites into a time series of subsurface volume change. The optimal location of the inflating source is consistent with previous studies, situated north of Nea Kameni at a depth of ~4 km. However, the time series reveals two distinct pressure pulses. The first pulse corresponds to a volume change (ΔV) within the shallow magma chamber of (11.56 ± 0.14) × 106 m3, and the second pulse has a ΔV of (9.73 ± 0.10) × 106 m3. The relationship between the timing of these pulses and microseismicity observations suggests that these pulses may be driven by two separate batches of magma supplied to a shallow reservoir. We find no evidence suggesting a change in source location between the two pulses. The decline in the rates of volume change at the end of both pulses and the observed lag of the deformation signal behind cumulative seismicity, suggest a viscoelastic response. We use a simple model to show that two separate pulses of magma intruding into a shallow magma chamber surrounded by a viscoelastic shell can account for the observed temporal variation in cumulative volume change and seismicity throughout the period of unrest. Given the similarities between the geodetic signals observed here and at other systems, this viscoelastic model has potential use for understanding behavior at other caldera systems.