Skip to content

Insights into Caldera Collapse Mechanics and Outstanding Questions from the 2018 Kīlauea Event

Paul Segall1, Taiyi Wang2,  Josh Crozier1, Mark Matthews3, Kyle Anderson4, Enrique del Castillo5

  • Affiliations: 1Geophysics Department, Stanford University, Stanford CA, USA; 2Geological and Planetary Sciences, Caltech, Pasadena CA, USA; 3Walden Consulting, Pittsfield MA, USA; 4U.S. Geological Survey Volcano Science Center, Moffett Field, CA, USA,  5Department of Civil and Environmental Engineering, Stanford University,  Stanford CA, USA

  • Presentation type: Poster

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

  • Poster Board Number: 49

  • Programme No: 3.11.9

  • Theme 3 > Session 11

Abstract

The 2018 Kīlauea eruption provided unique constraints on caldera collapse. High-rate effusion decreased pressure within the magma reservoir, causing the shear stress on the ring fault to increase. Eventually, frictional resistance was overcome, leading to collapse, re-pressurization the underlying magma. Continuum models of ring-fault development (using Smooth Particle Hydrodynamics) are consistent with the inferred magma pressure drop at the onset of collapse. Extra-caldera GNSS and tilt data are consistent with abrupt co-collapse pressure increases followed by exponential decay due to magma outflow, although the signature of faulting is difficult to isolate.  Significant uncertainties in the summit magma storage geometry, including the role of a distinct "south caldera reservoir remain.  * Volcano tectonic (VT) seismicity increased between collapses. GPS data from 2 sites (CALS, NPIT) within the caldera subsided proportional to cumulative seismicity counts.  This strongly suggest that most VT seismicity was driven by creep and/or distributed shear.  However, VT seismicity is not completely localized, and first-motions inconsistent with double-couple sources.  Seismicity was concentrated on newer ring-fault segments, suggesting fault roughness was important.  Seismicity rate, and the relative frequency of large VTs increased dramatically in the minutes before collapses. Inter-collapse subsidence at CALS & NPIT is consistent with rate-strengthening friction, but VTs and VLP nucleation require unstable friction.  This implies heterogeneities and high-speed fault weakening during collapses.  The final collapse cycle started unremarkably, although with a relatively low seismicity.  Whether the end of the eruption resulted from changes in the magmatic system, or the fault system is not settled.*