Subduction megathrust faults, which develop along the interface between the downgoing and overriding plates at convergent margins, are responsible for over 90% of the world’s largest earthquakes (M_w > 8).  Thus these regions are excellent locations to explore several key questions about how fault conditions and processes affect earthquake rupture behavior.  This study leveraged past NSF and internationally funded experiments conducted along the Central America subduction margin by using large datasets of earthquakes recorded locally with temporary land and ocean bottom seismometers.  We explored connections between fault conditions and earthquake rupture along this complex margin with a focus on the megathrust offshore Nicaragua and Costa Rica.  This region has hosted a slow tsunami earthquake (an earthquake that generates a larger tsunami than expected for an event of its size) and other slow slip and tremor processes, as well as “normal” earthquakes, thus we could look at along-strike transitions in fault properties that may lead to the slow slip processes. The targeted earthquakes span a range of magnitudes, locations, and mechanisms to address a key question: How do fault conditions impact earthquake rupture?  This question is a very important subset of one of the “Seismological Grand Challenges” [Lay et al., 2009]. 

The study was conducted in two parts.  First, we used a rich earthquake dataset from the NSF-funded Costa Rica Seismogenic Zone Experiment and subsequent networks and German-funded SFB 574 project to characterize earthquake source characteristics over a range of magnitudes and compare these parameters to specific tectonic and geologic variations. Second, we attempted to calculate near earthquake source compressional-to-shear velocity ratios (Vp/Vs) to complement already existing coarser scale velocity tomography to provide more detailed information on fault conditions.  The source parameter study was a success but the calculation of local Vp/Vs did not work due to unforeseen data and technique limitations.  Earthquake source parameters were therefore compared to high-resolution seismic velocity images.  We found significantly different earthquake source spectra for earthquakes occurring within the Nicaragua tsunami earthquake rupture zone relative to Costa Rica, with lower stress drops occurring offshore Nicaragua. Small magnitude earthquakes associated with reduced P-wave velocities along the plate interface extend much closer to the subduction trench, consistent with the generation of a tsunami earthquake. Stress drop was not significantly different for earthquakes located in slow slip and tremor zones along Costa Rica, however, and there was only weak evidence for major changes in P-wave or Vp/Vs variations associated with tremor and slow slip regions.  Along Costa Rica, most microseismicity is confined to a low Vp and moderately low Vp/Vs region that we hypothesize reflects changes in overpressure along the subduction megathrust.  The aftershock sequence of the 2012 Nicoya, Costa Rica, generated events with both high and low relative stress drop, even within a small segment of the subduction thrust fault.  The variations in stress drop were correlated with variations in main slip, with higher mean stress drop correlated with areas of higher mainshock slip.  Velocity images pre-dated the 2012 earthquake and showed mainshock rupture confined to the low Vp and low Vp/Vs features that also contain the interseismic microseismicity.   The results of this study illuminate the heterogeneity of complex subduction margins but suggest that detailed earthquake source studies combined with velocity imaging, geodetic work, and ongoing geophysical monitoring can be used better define the seismic and tsunami hazard of subd...

Please report errors in award information by writing to: awardsearch@nsf.gov.