Supplementary Tables 1-5, and Supplementary Text S1 The 2015 VLFE spectra as a sum of smaller events It is possible to model the 2015 M W 3.8 VLFE as the sum of about 10 M W 3.2 earthquakes occurring over the course of about 10 seconds, following the approach Gomberg et al. (2016) used to model VLFEs as sums of much smaller LFEs. To see this, imagine the VLFE is the sum of N subevents, all of which have moment M 0s and corner frequency f cs , and source spectra

S U M M A R YWe searched for earthquake swarms in South America between 1973 and 2009 using the global Preliminary Determination of Epicenters (PDE) catalogue. Seismicity rates vary greatly over the South American continent, so we employ a manual search approach that aims to be insensitive to spatial and temporal scales or to the number of earthquakes in a potential swarm. We identify 29 possible swarms involving 5-180 earthquakes each (with total swarm moment magnitudes between 4.7 and 6.9) within a range of tectonic and volcanic locations. Some of the earthquake swarms on the subduction megathrust occur as foreshocks and delineate the limits of main shock rupture propagation for large earthquakes, including the 2010 M w 8.8 Maule, Chile and 2007 M w 8.1 Pisco, Peru earthquakes. Also, subduction megathrust swarms commonly occur at the location of subduction of aseismic ridges, including areas of long-standing seismic gaps in Peru and Ecuador. The magnitude-frequency relationship of swarms we observe appears to agree with previously determined magnitude-frequency scaling for swarms in Japan. We examine geodetic data covering five of the swarms to search for an aseismic component. Only two of these swarms (at Copiapó, Chile, in 2006 and near Ticsani Volcano, Peru, in 2005) have suitable satellite-based Interferometric Synthetic Aperture Radar (InSAR) observations. We invert the InSAR geodetic signal and find that the ground deformation associated with these swarms does not require a significant component of aseismic fault slip or magmatic intrusion. Three swarms in the vicinity of the volcanic arc in southern Peru appear to be triggered by the M w = 8.5 2001 Peru earthquake, but predicted static Coulomb stress changes due to the main shock were very small at the swarm locations, suggesting that dynamic triggering processes may have had a role in their occurrence. Although we identified few swarms in volcanic regions, we suggest that particularly large volcanic swarms (those that could be detected using the PDE catalogue) occur in areas of infrequent eruption and may be related to large regional fault zones.

No abstract

[1] Continuous GPS stations in the Pacific Northwest Geodetic Array network clearly record subduction-related strain accumulation and slow slip episodes along the Cascadia convergent margin. Many of the slow slip episodes have been correlated in time and space with seismic evidence for nonvolcanic tremor, leading to the previous discovery of episodic tremor and slip (ETS). In this study, we use a hyperbolic tangent curve fitting technique for the identification of slow slip times and displacement magnitudes within the GPS time series, independent of seismic tremor data. We then apply this technique to study the patterns of strain accumulation and release associated with ETS events and characterize patterns of coupling associated with the locked and transition zones of the plate interface. We demonstrate the effectiveness of this automated technique for both identification of slow slip observations and calculation of slow slip displacements. Recurrence patterns in the distribution of GPS observations demonstrate coherence among neighboring stations over time and apparent along-strike segmentation of the subduction interface. When slow slip events are removed from the time series, we can estimate the total site velocities between slow slip events. These velocities decay as depth to the subduction interface increases, but they diverge from the long-term trends expected from the interseismic cycle at $30-60 km above the interface, consistent with the location where slow slip displacements occur. Forward modeling of coupling on the plate interface reveals that in between slow slip events there is a patch of at least 30% coupling from 20 to 35 km depth, which is needed to produce the observed back slip displacements. Intriguingly, our best fitting models have a decrease in coupling down to $30% at $20 km depth followed by a peak of greater than 80% coupling at $30-35 km depth, suggesting the source zone for ETS events acts as a distinct locking zone that releases strain more frequently than the updip seismogenic locked zone, although a zone of constant $30% coupling cannot be ruled out with this data set. Such a scenario indicates that frictional behavior with depth follows a more complex model than a simple temperature controlled transition. We propose that coupling initially decreases with depth due to a decrease in strength of the overriding lower crust, but then coupling increases again when the subducting plate comes in contact with the stronger overriding mantle.Citation: Holtkamp, S., and M. R. Brudzinski (2010), Determination of slow slip episodes and strain accumulation along the Cascadia margin,