Research

As scientists, we use simplified models to try to understand the earthquake rupture process and the variability of observations of earthquake frequencies and ground motions. In reality, though, earthquakes are quite complex - they come in all shapes and sizes, and new events continue to surprise the earthquake science community. To understand source physics and rupture variability, we modeled static stress drop (i.e., stress released in each earthquake assuming they are simple circular sources) and found significant variability along a fault plane activated during the Mogul earthquake swarm. The plot above shows temporal variation (foreshocks vs. aftershocks) and spatial variation (the patterns in each plot) in stress drops of the larger events in the seismic sequence. We calculate independent stress drop results for both P waves (left column) and S waves (right column), and their consistency suggests that the variation we measured is real and robust. See Ruhl et al. (2016b) and Ruhl et al. (2017b) for more information.

Abundant microseismicity is very common in diffusely deforming regions and in developing fault systems such as the transtensional Walker Lane in the Western US. Large, characteristic faults typically host moderate to large earthquakes, but are quiescent most of the time. Instead of on the major throughgoing faults, microseismicity tends to concentrate on the ends of normal faults and in the transition zones between en echelon faults. In Ruhl et al. 2016a, we relocated 15 years of seismicity in the Reno-Lake Tahoe area of California and Nevada and investigate spatial patterns to understand their relationship to active tectonics. The figure above shows smoothed seismogenic depths from our original study (left) over the mountainous topography in the area. The right figure is showing the crustal thickness (colors=Moho depth) from geophysical studies as well as heat flow measurements (labeled). As expected, the crustal thickness and seismogenic depths increase towards the west as the crust thickens under the Sierra Nevada mountains. Conversely, heat flow increases towards the east as the crust thins and gets hotter in the Basin and Range.

Typical earthquake sequences occur with a mainshock first (the largest event) and are followed by aftershocks that decrease in intensity (number and magnitude of events) with time after the mainshock. Sequences that begin with smaller earthquakes before the mainshock (foreshocks) or never have a mainshock at all (swarms) are an unusual and poorly understood phenomena. By carefully relocating over 7,000 earthquakes and investigating their fault kinematics, we learned about the earthquake triggering process of the 2008 Mogul, NV earthquake swarm. The figures on the left (a and b) show earthquakes from the swarm plotted by distance from the first event as a function of time. The curves and lines are showing maximum diffusion and/or linear migration rates that might explain the spatiotemporal evolution of the events. See Ruhl et al. (2016b) to read more about this fascinating (and unusually shallow) sequence of earthquakes.

INTERESTED IN WORKING WITH THIS DATASET?

A relocated catalog of ~7,500 earthqukes is available, as well as ~1000 first-motion focal mechanisms and 11 moment tensor solutions. If you are interested in working with this unique sequence, please email me for more information.

Figure: Spatiotemporal evolution of the 2008 Mogul, NV earthquake swarm. (a) Two-months of foreshock diffusion leading up to M4.9; (b) Entire five-month spatiotemporal evolution of the swarm; (c)-(f) migration of individual subclusters of events found using statistical seismology.

Peak Ground Displacements (PGD) can be used to accurately estimate earthquake magnitudes for M>6 events. This figure shows the maximum displacement measured from data recorded on geodetic (GPS & GNSS) instruments during 29 large earthquakes occurring all over the world. Notice that PGD scales with magnitude (increasing magnitudes colored from blue to red) and with distance away from the earthquake. Our new scaling law more accurately estimates magnitudes than previous relationships developed using smaller datasets. See Ruhl et al. (2018) for more information.

LOOKING FOR EARTHQUAKE DATA?

Seismic and geodetic data compiled for these studies is available for download here.

Earthquake Early Warning (EEW) aims to detect and characterize earthquakes after they begin in order to warn ahead of the arrival of strong ground shaking. To acheive this goal, shaking intensity (MMI) predictions must be both accurate and fast. Geodetic EEW algorithms predict unsaturated magnitudes and finite-fault solutions that lead to better MMI accuracy than seismic point-source systems. It continues to be debated whether geodetic EEW algorithms are really fast enough to provide alerts, however. In our upcoming paper (see AGU 2018 abstract here), we show quantitatively that geodetic algorithms are, indeed, fast enough and provide significantly more accurate results than seismic-only systems with long warning times. The figure above shows our results for one of the closest stations to the 2011 M9.0 Tohoku, Japan earthquake. The multi-colored line shows the data, the blue results show the seismic predictions, and the red and purple results show the geodetic solution. The improvement in MMI accuracy is significant and clear. See my AGU 2018 poster here for more information.