Susan Schwartz

Profile | Research | Publications | Teaching

Costa Rica Seismogenic Zone Structure and Behavior

Subduction is a fundamental geological process generating and modifying continental crust and associated with severe natural hazards including earthquakes, volcanoes, and tsunami. The portion of a subduction zone capable of generating earthquakes (termed the seismogenic zone) is fairly restricted in depth but is responsible for releasing ~90% of the earth's seismic energy. Global observations at subduction zones reveal a large range in seismogenic behavior with variations in maximum earthquake magnitude, repeat time, and slip distributions. My research focuses on fundamental processes controlling earthquake occurrence and the large variations in earthquake behavior. Specifically, with many collaborators, I have been investigating the detailed seismic structure and seismic and geodetic behavior of the northern Costa Rica seismogenic zone. Using seismic tomography (DeShon et al., 2006), receiver function analysis (Linkimer et al., 2009) and integrated GPS modeling and seismicity studies (Newman et al., 2002; Bilek aet al., 2003; Norabuena et al. 2005; Deshon and Schwartz, 2004; Schwartz and DeShon, 2007; Outerbridge et al., 2010) we have learned that subduction of lithosphere with different genesis and therefore different geomorphic features, level of hydration and temperature profiles produce significant changes in seismic behavior.

Slow Slip and Tremor in the Northern Costa Rica Seismogenic Zone

It has been known for a long time that slip accompanying earthquakes accounts for only a fraction of plate tectonic displacements. Its most commonly been assumed that the rest of the motion occurs in continuous, aseismic slip. However, recently a fuller spectrum of strain release processes, including slow slip events, low and very low frequency earthquakes and seismic tremor has been observed (Schwartz, 2007; Schwartz and Rokosky, 2007). Slow slip events have the same fault motion as earthquakes but the slip occurs so slowly that it does not radiate seismic waves and therefore produces no ground shaking. Several episodes of slow slip and tremor have occurred at the plate boundary of northern Costa Rica between 2000 and 2009. The evidence for these events varies and consists of: 1) correlated fluid flow excursions and seismic tremor recorded on ocean bottom instruments in 2000 (Brown et al., 2005); 2) offsets in continuous GPS data in September 2003; 3) offsets in GPS data accompanied by seismic tremor in May 2007 (Outerbridge et al., 2010; Brown et al., 2009) and 4) strong prolonged seismic tremor in August 2008 and spring 2009. Modeling of the 2000 event suggested that slip occurred at shallow depth, between the surface and ~15 kilometers. The much better constrained slip distribution of the 2007 event consisted of 2 patches, the stronger centered at ~30 km depth, near the down dip transition from stick-slip to stable sliding, and the weaker patch located at ~6 km depth at the up dip edge of the shallow frictional transition. These results are significant in that they are the first to suggest that slow slip occurs at the up dip transition from stick-slip to stable sliding; locations of slow slip in other environments have been limited to the down dip frictional transition.

Due to the relatively small surface displacements (1-2 cm) associated with Costa Rica slow slip events, the coincident occurrence of seismic tremor is important for their detection and study. Similar to tremor observations in southwest Japan, some Costa Rica tremor consists of swarms of low-frequency earthquakes (LFEs) that occur as repetitive stick-slip motion on the plate interface (Brown et al, 2009), contains very low frequency earthquakes, with dominant energy between 20-50 s, and appears to be tidally modulated. In contrast to most other regions, Costa Rica tremor occurs within the seismogenic zone in freely slipping regions, adjacent to patches that are presently accumulating strain (Walter et al, 2010).

Glacial Seismology

Interdisciplinary studies are an important component of our seismology program. Graduate student Jake Walter, in collaboration with seismologists Emily Brodsky and Susan Schwartz and glaciologist Slawek Tulaczyk, has been working in the quickly emerging interdisciplinary field of glacier-generated seismicity. In particular, he has been investigating the bi-daily, tidally modulated stick-slip speed-ups of Whillans Ice Plain in West Antarctica. These events are important for understanding ice stream dynamics and instability and as a possible glacial analogue for tectonic slow slip phenomena. We operated a network of GPS receivers and on-ice broadband seismometers in the austral summers of 2007, 2008 and 2010 on Whillans Ice Stream. These data reveal a strong correlation between amplitude of seismic waves generated at the rupture front and total slip achieved over the duration of the slip events (~ 30 min), suggesting slip-predictable behavior. Also, we detect variable rupture speed that correlates with inter-event stress accumulation. We are studying these behaviors to reveal important information regarding mechanics and dynamics of ice stream beds at the scale of 10s to 100s of km (Walter et al., 2010).