A New Perspective on Earthquake Triggering of Volcanic Unrest
Stephanie Prejean1 , Joan Gomberg2, Paul Bodin3, Jeremy Pesicek4
Affiliations: 1. USGS-USAID Volcano Disaster Assistance Program, Anchorage Alaska, USA 2. U.S. Geological Survey, Seattle Washington, USA 3. Univ. of Washington, Seattle Washington, USA 4. USGS-USAID Volcano Diasater Assistance Program, Vancouver Washington, USA
Presentation type: Talk
Presentation time: Friday 16:00 - 16:15, Room S160
Programme No: 2.1.13
Abstract
The scientific literature contains many suggested cases of volcanic eruptions triggered by large regional earthquakes. These studies generally fall into two categories: 1) case studies examining the effect of a single earthquake on a specific volcano or 2) global "earthquake-centric" studies that analyze changes in eruption rates at volcanoes in some region around a large earthquake. These studies are generally limited to earthquakes of >=M6 and to eruptions of >=VEI 2. They also neglect crustal relaxation and the cumulative effect of multiple earthquakes. Here, we present a new "volcano-centric" methodology to address this critical issue. We calculate both the static and dynamic strain changes that a volcano "feels" from all earthquakes, regardless of magnitude. Thus, we generate time series of earthquake-induced strain changes at specific volcanoes that can be compared to measurable volcanic unrest. We apply our method to three volcanoes with differing tectonic settings and eruptive styles: Mt. Redoubt (USA), Turrialba volcano (Costa Rica), and the Taupo volcanic zone (New Zealand). We find no clear relationship between earthquake-induced strain changes and eruptions at Redoubt and Turrialba. At Taupo we find that the cumulative effect of earthquakes helps explain a potential earthquake-induced intrusion. Finally, we also find that nearby earthquakes <M6 may impart a larger strain change on the volcanoes than more distant larger earthquakes, and that the cumulative strain on volcanoes from more frequent smaller earthquakes can exceed the strain from individual large earthquakes. Thus, prior studies on this topic may be using an incomplete dataset.