Research
Introduction to Yifang and Yifang's research
Why small earthquakes?
"Although this interevent seismicity contributes only marginally to relative plate motion, it is symptomatic of processes underlying the earthquake cycle.
The wealth of data generated by frequent smaller events provides important clues to the seismotectonic fabric, kinematics, and state of stress within the brittle crust
And ultimately, to the seismogenic processes common to earthquakes of all sizes.
This persistent, inter-event seismicity also captures widespread popular interest because it includes most of the felt earthquakes in California (earthquakes of M a 3 can be locally felt), and the larger of these interevent earthquakes (M=6-7) can cause extensive damage and loss of life when they strike near major population centers."
—— Wallace, R. E. (1990). The San Andreas fault system, California. Department of the Interior, US Geological Survey.
Section 1: Volumetric fault zone structures and earthquake ruptures
Diverse volumetric faulting patterns in the San Jacinto fault zone
The standard model for earthquake and fault mechanics assumes that moderate and large earthquakes can be described to first order in terms of slip along a single surface in a continuum solid. However, our study on five moderate aftershock sequences in the San Jacinto Fault Zone shows diverse seismicity structures, large aftershock zones with little spatial overlap, long-duration sequences, and significant distance between main- shocks and aftershocks, pointing to complex interactions between the main faults and the surrounding damage zones. The results indicate that Describing deformation and rupture processes of San Jacinto Fault Zone requires volumetric frameworks and single fault surface assumption may obscure essential aspects of the dynamics.
Transient deepening of seismic-aseismic transition depth following moderate and large California earthquakes
The depth extent of seismogenic faults is essential for estimating the possible rupture area, magnitude and seismic hazard associated with large events. Previous modeling results show large earthquake may penetrate deeper than the locking depth and maximum depths of aftershocks are higher than the background events due to high strain rates. Here we estimate the bottom of seismogenic zone using D95 and find that large magnitude aftershock sequences have much deeper D95 compared with the D95 estimated by background seismicity. Our results are compatible with the results from modeling, suggesting deeper rupture and potentially increasing seismic hazard associated with large earthquake.
Variations of earthquake properties before, during, and after the 2019 M7.1 Ridgecrest earthquake
The Eastern California Shear Zone is one of the seismically most active regions in southern California and hosted in the last few decades several large earthquakes. The most recent of these is the July 5, 2019, Ridgecrest earthquake with magnitude 7.1, which was followed by a vigorous aftershock sequence. To clarify processes associated with the Ridgecrest earthquake sequence, we analyze properties of earthquakes before, during and after the 2019 Ridgecrest mainshock.
Key observations:
The fraction of normal faulting events in the area dropped in the past 20 yr from >25% to <10%, implying increasing shear stress.
The Mw6.4 and Mw7.1 events terminated at areas with abrupt changes of surface geology and geometrical fault complexities.
Aftershocks with diverse mechanisms produced significant potency in a 5-10 km wide zone, including deeper-than-usual early events
Isotropic source components of events in the 2019 Ridgecrest earthquake sequence
Earthquakes occur when rocks below the surface break and move rapidly. Deriving earthquake source mechanisms provides information on the involved physical processes. We examine source mechanisms of M ≥ 3.0 earthquakes in the 2019 Mw7.1 Ridgecrest sequence, using waveforms from 39 near-fault and regional stations.
Key observations:
50 out of 224 M>3 earthquakes show considerable isotropic components not resolved without near-fault data
Events with large isotropic components occurred early in the sequence near rupture ends and fault intersections
Rock fracturing in earthquake source volumes likely contributes significantly to the isotropic components
Section 2: Algorithms for solving small earthquake properties
An Automated Method for Developing a Catalog of Small Earthquakes Using Data of a Dense Seismic Array and Nearby Stations
We propose a new automated procedure for using continuous seismic waveforms recorded by a dense array and its nearby regional stations for P-wave arrival identification, location, and magnitude estimation of small earthquakes. The method utilizes the inter-station waveform similarity and relative arrival time differences across the dense array to achieve high accuracy, high performance earthquake P-wave arrival detection (about six times of the catalog events and 98% accuracy). By associating the detected dense array P-wave arrivals and the P- and S-wave arrivals from the surrounding stations using a 1D velocity model, ∼4 times of the catalog events have well-estimated locations and magnitudes. Combining the small spacing of the array and large aperture of the regional stations, the method achieves automated earthquake detection and location with high sensitivity in time and high resolution in space. Because no pre-knowledge of seismic-waveform features or local velocity model is required for the dense array, this automated algorithm can be robustly implemented in other locations.
A refined comprehensive earthquake focal mechanism catalog for Southern California derived with deep learning algorithms
Cheng et al., (2021; in prep)
Focal mechanism describes the style of faulting and fault orientation, which can be used to improve the understanding seismotectonic processes, facilitate potential seismic hazard estimation and help to determine the temporal and spatial variations of regional stress field orientation. Traditional focal mechanism determination uses first-motion polarities and arrivals determined by data analysts. Ross et al. [2018] trained a polarity picker with deep learning using millions of labeled seismograms and the picker can detect many more polarities than catalog with about 95% accuracy, resulting in more focal mechanisms. Here we are trying to update the focal mechanism catalog for the whole southern California using the polarities picked by picker in Ross et al. [2018] to obtain an updated focal mechanism catalog with both higher quality and quantity.
Spectra-semblance-based rupture directivity estimation and its application to small earthquakes in Southern California
Cheng et al., (2021; in prep)