This three-year project aimed to improve our understanding of earthquakes that occur at a depth of 70-300 km below the Earth's surface.  These are referred to as "intermediate-depth" earthquakes, and they are interesting because they occur at depths where laboratory experiments indicate that earthquakes should not occur.  They clearly do occur, however, and it's important to understand why and how that happens.  There are two competing ideas to explain their occurrence.  The first is that water released by dehydration reactions of minerals effectively lubricates faults and allows them to slip.  The second is that the heating that accompanies localized deformation causes weakening, which further localizes deformation as slip onto faults and creates more heat through friction.  This positive feedback is known as a thermal runaway.  Our results are consistent with the second of these ideas, the thermal runaway hypothesis.  Our findings fall into three categories: 

(1) We found that small, intermediate-depth earthquakes are very inefficient at generating seismic waves.  This observation is consisent with the thermal runaway hypothesis because it indicates that for small earthquakes, much of the energy is taken up near the fault, possibly through melting of rock.  For larger earthquakes, this would be a negligible effect.

(2) We found a relative scarcity of small, intermediate-depth earthquakes.  We consider this to be consistent with the thermal runaway hypothesis because small earthquakes would have a bias to grow large (and not be small anymore) due to the thermal runaway process.  Although it sounds simple, documenting that small earthquakes are "missing" requires establishing sensitive measures to detect as many small earthquakes as possible, accurate magnitude measures for small earthquakes, and a quantitative assessment of the expected detection threshold such that the observations of a lack of small earthquakes aren't simply an artifact of an incomplete earthquake catalog.  After taking these steps, we confirmed the scarcity of small earthquakes in the Bucaramanga Nest.

(3) We discovered two populations of earthquakes that occur in close proximity to one another, yet with an opposite sense of slip.  After relocating these earthquakes as accurately as possible, we were able to infer that they are consistent in their geometry with features that form in shallower rocks when they are squeezed and stretched ductilely - that is - without fracturing brittlely.  This suggests that the earthquakes we observe are occurring at very near the brittle-ductile transition, such that temperature perturbations would have a strong effect on their behavior.  This too supports the thermal runaway hypothesis, though not as directly as points (1) and (2).

While our work under this project strongly suggests that thermal runaway plays a key role in intermediate-depth earthquakes, our findings strictly only apply to the Bucaramanga Nest under Colombia.  The next, obvious step will be to apply our techniques to intermediate-depth earthquakes around the world to see whether our conclusions generalize.  This will be key to establishing the broader scientific impacts.  Other, immediate broader impacts are that the empirical subspace detection method we developed to detect small earthquakes is now being used to detect small tectonic earthquakes in Europe and to detect injection-induced earthquakes in the central US.  Another broader impact is that this grant funded the PhD research of Sarah Barrett, a woman in a STEM field in which females remain under-represented.  Dr. Barrett is now working in the private sector for reinsurance company SwissRe in Armonk, New York, where she is using her background to help insurance companies redistribute their risk from natural hazards.

Last Modified: 11/21/2016
Modified by: Gregory Beroza