Dr. Harmony Colella has been awarded and NSF Earth Science Postdoctoral Fellowship to carry out a research and education plan at Miami University of Ohio. This study will employ complementary research approaches, observational and modeling, to constrain the parameters and processes that cause the variety of episodic tremor and slip (ETS) and earthquake phenomena observed in subduction zones. More specifically, the project will identify and locate tremor in south central Alaska (Alaska-Aleutian Subduction Zone) and the Oaxaca and Jalisco-Colima segments of the Middle America Subduction Zone. The spatial and temporal distribution of tremor and previously documented large and great earthquakes will be used to constrain fault simulations to explore the range of fault parameters that influence the frequency and segmentation of the variety of earthquake types observed in subduction zones and to develop robust statistical characterizations of great earthquakes and ETS events.
Subduction zones generate the world's largest earthquakes (M8-9+). These seismic releases of stored-up elastic energy result in intense ground shaking over large regions and often displace the sea floor, which generates tsunamis like those of the great 2004 Sumatra and 2011 Japan earthquakes. Recently discovered ETS events represents another way these faults can relieve the stored energy, but the slip occurs more slowly over weeks to months and is accompanied by weak tectonic (nonvolcanic) tremor. This research will specifically target the impact of ETS on hazardous great earthquakes, with particular interest in the potential for ETS events to trigger large or great earthquakes and understanding the properties that may be related to why each subduction zone demonstrates different behaviors. In addition, Dr. Colella will develop tools to assess the learning experience in undergraduate geohazards assignments and of undergraduate research success, and organize and lead graduate seminars that will assist graduate students in the natural sciences prepare for the transition into their careers.
The scientific goals of this project included an observational seismology and numerical modeling component, which aimed to investigate slow slip events and small, "regular" earthquakes adjacent to, and at the base of, the section of a subduction zones that generates Mw>7 earthquakes, occasionally resulting in devastating tsunamis. The observational seismology component of this project studied aftershock data from the 2012 Mw7.4 Oaxaca earthquake (Figure 1) was analyzed for detectable earthquakes (~45). The waveforms of the detected earthquakes were used to scan previously collected seismic data from the region for "matching" or "repeating" earthquakes; this is known as the template-matching method in seismology. Results show an increase in repeating earthquakes during slow slip events that occur in the coastal regions, near the base of the seismogenic zone (Figure 2). Tectonic tremor, a phenomenon sometimes associated with slow slip events, is observed only with slow slip events that occur further inland, or further away from the seismogenic zone (Figure 2). Furthermore, a slow slip event near the base of the seismogenic zone along with increased levels of repeater earthquakes were observed immediately prior to the Mw7.4 Oaxaca earthquake. This adds to evidence from other recent Mw>7 earthquakes that show precursory activity prior to the mainshock. Additionally an increase in the rate of repeating earthquakes was identified during 2 previously recorded slow slip events, which were geodetically located in the coastal region of Oaxaca, but did not culminate in a large-to-great earthquake. The numerical modeling component of this project (1) explored a variety of fault parameters (e.g., normal stress, fluids, frictional instability, convergence rates) to determine the controlling factor(s) of slow slip events, and (2) aimed to reproduce increased levels of seismicity associated with slow slip events. The first goal of the modeling component was completed with the assistance of an undergraduate student who worked on parts of the project as part of an Incorporated Research Institution of Seismology (IRIS) Research Experience for Undergraduates (REU), and completed the project as I mentored him through his senior thesis. Results indicate the effective normal stress (or increase in fluids) play a key role in the characteristics of slow slip events (Figure 3). These results were recently published in Geophysical Research Letters with my student as the first author. Preliminary models designed to reproduce increased levels of seismicity associated with slow slip events indicate that increased levels of seismicity are not associated with all slow slip events (Figure 4). Additionally, patches that generate earthquakes similar to repeating earthquakes need to be imbedded in areas that slip similar to slow slip events. More studies are planned understand all possible conditions that will generate increased levels of seismicity and to determine the likelihood a slow slip event and increased levels of seismicity will result in a Mw>7 earthquake. The broader impact component of this fellowship involved a qualitative assessment of the IRIS REU program. The goal of the assessments obtain specific examples (anecdotes) of (1) what attracts (or not) science, technology, engineering, and mathematics students to geosciences, (2) the qualities of good (or bad) mentors, (3) specific examples of what aspects of the IRIS REU program works (and does not work), and (4) any other experiences that may present themselves as influential to their decision-making process about geosciences and their futures. Pre-internship, post-internship, and post-conference presentation of their research interviews were completed with a sample of the 2013 and 2014 IRIS interns. The initial interviews were conducted and recorded immediately prior to the start of the internship orientation at New Mexico Tech to create a trust and comfortability with the interns. The second and third interviews were conducted via video chat, in which the audio was recorded. The transcripts are currently being evaluated and coded with the assistance of someone who specializes in qualitative assessments to obtain the best results.
