Search

We implement an algorithm to automatically detect migrations of low frequency earthquakes at time scales between 30 min and 32 h during the 2003, 2004 and 2005 slow slip events in Cascadia. We interpret these migrations of seismicity as a passive manifestation of secondary slip fronts (SSFs) that propagate faster than the main front. We identify the dominant features of 383 SSFs, including time, location, duration, area, propagation velocity and estimate: their moment, stress drop, slip, and slip rate. We apply the same algorithm to continuous tremor detection in Cascadia between 2009 and 2015 and characterize 693 SSFs at time scales between 4 h and 32 h. We identify — to our knowledge for the first time — numerous 11–22.5 h long SSFs that propagate at velocities intermediate between slow slip events and previously reported SSFs. The systematic detection of SSFs fills a gap between seismically and geodetically detectable slow earthquake processes. Analysis of SSF basic features indicates a wide range of stress drops and slip rates (with medians of 5.8 kPa and 1.1 mm/h) as well as an intriguing relationship between SSF direction and duration that was observed in other contexts and could potentially help discriminate between the different physical models proposed to explain slow slip phenomena. [ABSTRACT FROM AUTHOR]

• Creation of 3D velocity-corrected discontinuity model for the Cascadia margin. • Forearc crustal thickness and seismic structure highlights zones of subcretion. • Slab morphology, fluid release, and subcretion contribute to significant margin variations. • Subcretion is the primary control on Cascadia forearc differential topography. A number of seismic and other geophysical phenomena exhibit significant lateral heterogeneity along the strike of the Cascadia forearc. Both the overriding and downgoing plate have been invoked to play the dominant role in controlling along-strike correlations between seismogenic behavior, potential field measurements, morphological/tectonic characteristics, and seismic structure; however, significant feedbacks likely exist between the two. In this study, we apply a 3D velocity correction to receiver function data and interpret the resulting discontinuity model alongside a recently published shear-wave velocity model to understand the possible causative relationships between correlative along-strike variations. Our discontinuity model indicates that the forearc crust thickens as it approaches the mantle wedge corner before progressively thinning toward the magmatic arc, likely due to the basal accretion of material from the downgoing plate to the overriding plate. In the northern and southern portions of the forearc, this "subcreted" material is characterized by thick (∼10 km) anomalously low shear-wave velocity zones. The thickness, high internal reflectivity, and low Bouguer gravity signatures associated with the low velocity zones likely indicate that this subcreted material is composed of dominantly (meta)sedimentary material that has been emplaced through successive subcretion events over geologic timescales. Furthermore, the anomalously low velocities and spatial correlation with high non-volcanic tremor (NVT) density and short slow slip recurrence intervals indicate that these regions are fluid-rich. While first-order variations in the fluids that control NVT and slow slip likely result from differences in the permeability of the downgoing slab as inferred from its stress state and the distribution of intraslab seismicity, these subcreted packages likely represent thick, vertically-impermeable regions in the lower crust that further accentuate this correlation. Variability in the amount of subcretion explains patterns of exhumation and uplift along the Cascadia margin and the resulting forearc topography over geologic timescales, and is likely controlled by some combination plate interface geometry/rheology and overriding plate architecture. [ABSTRACT FROM AUTHOR]

• We invert GPS, strong motion, and tide gauge records to determine the slip distribution of the 2018 May 4 Hawai'i event. • This earthquake likely ruptured through an adjacent region that regularly hosts slow slip. • Effective stress levels can control the extent to which coseismic slip can penetrate into regions hosting slow slip. Earthquakes and slow slip are different faulting processes that release accumulated stress on different time scales. Despite nearly two decades of study, it is still unclear whether the same section of fault is capable of hosting both slow, aseismic and fast, seismic slip. Here, we jointly invert GPS, strong motion records, and empirically corrected tsunami waveforms for the 2018 M7.1 Hawai'i earthquake. Our inversion results suggest the earthquake had a slow rupture speed of 1.2 km/s with a maximum slip of ∼3 m South-East of the hypocenter extending to the South-West through an area known to regularly host slow slip events. After exploring how the choice of fault geometry and regularization parameters affect the slip distributions of the earthquake and slow slip events we find strong evidence of overlap between the two. Additionally, we perform numerical modeling to simulate fast and slow slip behavior. Due to the homogeneity of the Hawaiian décollement differences in effective stress are likely an important factor that causes the diversity of slip. We find that an earthquake can completely penetrate into the slow slip zone provided the effective stress differences between fast and slow slip zones is large enough to facilitate this. Our results reinforce the idea that an individual section of fault is capable of hosting a variety of distinct slip behaviors. This is important for estimating earthquake rupture extent and for better ground motion and tsunami hazard assessment. [ABSTRACT FROM AUTHOR]