Search for a project

Observing and modelling transient slip events on creeping faults

Dr Jessica Hawthorne (SEE), Dr John Elliott (SEE), Dr Richard Walters (Durham), Prof Tim Wright (SEE)

Contact email: J.C.Hawthorne@leeds.ac.uk

Creeping faults accommodate most of their slip aseismically, often at average rates of a few cm per year, comparable to the speeds of plate motion.  Friction on creeping faults is thought to be broadly velocity-strengthening---resistant to large increases in slip speed.  And indeed, creeping faults are so far observed to host few to no large earthquakes  (e.g., Johanson and Bürgmann, 2005; Lienkaemper et al, 2012). From a hazard perspective, creeping faults would then seem ideal, as they gradually slip to allow for large-scale plate motion.

However, there is at least one important aspect of creeping faults that we do not understand: transient slip events, or surface creep events (e.g., Gladwin et al, 1994; Rousset et al, 2016).  In these intervals, the slip rates on portions of creeping faults increases by a factor of 10 to 1000 and remains high for a few minutes to a few days before decreasing back to rates smaller than the rate of plate motion.  For instance, on the northern portion of the creeping section of the San Andreas Fault, highly periodic several-hour long surface creep occur every few months.  These events account for most of the slip accumulated at this location, as seen in Figure 1 (e.g., Gladwin et al, 1994).

Figure 1: Creep events recorded at USGS creepmeter CWN, which records the surface displacement across the fault.  CWN is located near the northern end of the creeping section of the San Andreas Fault.

Figure 2: The creeping (red) and locked (orange) portions of the central San Andreas Fault.  The creeping section hosts numerous small earthquakes and creep events but few large earthquakes.

Despite accounting for large fractions of the total slip, creep events have sometimes been considered relatively unimportant in the scheme of large-scale slip on creeping faults.  Individual creep events are mostly small, and they occur at the surface, so one hypothesis is that creep events are just a small-scale phenomenon occurring in response to rainfall or atmospheric pressure variations.  However, recent observations have shown that many creep events extend 4 or 5 km along the strike of the fault and to depths of 4 or 5 km---well into the seismogenic zone (e.g., Mencin, 2016).  The apparently large extent of repeated creep events also forces us to reconsider physical models of their behaviour.  What frictional property allows portions of the fault to accelerate to speeds well above plate rate but then stall long before into an earthquake?  This question is a topic of vigorous debate in the context of slow slip events---much larger (several 100-km-long) creep events that occur at depth, on the plate interface below the seismogenic zone.  A range of possible physical mechanisms have been put forward, including changes in pore pressure that restrict acceleration or constraints associated with the limited size of slipping patches (e.g., Segall et al, 2010; Wei et al, 2013).  However, it remains unknown whether the physical processes allowing for transient slip events are the same for deep slow slip as for shallow creep events.

Creep events are of particular interest today because of the increasing availability of high-resolution GPS and InSAR data, which make it possible to identify, observe, and understand the behaviour of creep events.

Objectives

In this project you will

  1. Explore several possible models of creep events,
  2. Test and develop the models with existing creepmeter data, and
  3. Constrain and compare more predicted properties of creep events using InSAR data.

Numerical and Conceptual Modelling

You will consider several frictional models of the evolution of slip in creep events.  You may begin with a “standard” frictional model, often used to reproduce steady creep or earthquakes.  However, the relatively slow slip rates in creep events suggest that you may quickly decide that a more complex frictional model is necessary.  For instance, it has been hypothesized that fault zones expand as they shear.  This expansion could reduce the pore pressure in the fault zone, which in turn can increase the effective normal stress and clamp the fault together, keeping it from slipping further (e.g., Segall et al, 2010).

Of particular importance in the models is the importance of asperities: portions of the fault that may be especially prone to episodic creep.  We want to know whether the entire fault is somewhat unstable, or whether it's just a few locations that initiate the episodic slip.  To test the importance of asperities, you may consider a range of questions, including:

  1.  whether creep events start quickly and decay, or grow at a steady rate,
  2.  whether the slip rate in creep events is simply related to the propagation rate,
  3.  whether events persistently start at the same place,
  4.  how often creep events are influenced by external loading.

Observations

Some of these questions can be addressed by comparing the models' propagation and slip rates with creepmeter and strainmeter observations. However, creepmeters measure slip only at a single location, so they provide limited information about the spatial extent of creep events.  To better constrain the spatial extent and depth of creep events, you may use InSAR data to see the distributed deformation produced by the slip.  Better constraints on the depth extent can feed back into the models and allow you to determine how much of the large-scale fault slip is influenced by creep events.  In addition, the observations maypermit you to identify creep events at locations where creepmeters are unavailable, and thus to constrain the behaviour of a larger range of faults.

Training

In addition to specialist training, through Leeds and COMET you would be able to interact with a large number of scientists with expertise in a range of aspects of fault mechanics, including modelling and geophysical and geological observations.  You will also have access to courses organized by the faculty and university (http://www.emeskillstraining.leeds.ac.uk).    

Student profile

This project would be suited for a student with a background in geology, physical sciences, computer science, or engineering. 

References

  • Johanson, I. A., and R. Bürgmann. 2005. “Creep and Quakes on the Northern Transition Zone of the San Andreas Fault from GPS and InSAR Data.” Geophysical Research Letters 32 (July): 14306.
  • Lienkaemper, James J, Forrest S McFarland, Robert W Simpson, Roger G Bilham, David A Ponce, John J Boatwright, and S. John Caskey. 2012. “Long-Term Creep Rates on the Hayward Fault: Evidence for Controls on the Size and Frequency of Large Earthquakes.” Bulletin of the Seismological Society of America 102 (1): 31–41. doi:10.1785/0120110033.
  • Mencin. 2016. “Shallow and deep creep events observed and quantified with strainmeters along the SAF in Parkfield and the NAF in the Maramara.” FaultLab workshop.
  • Rousset, Baptiste, Romain Jolivet, Mark Simons, Cécile Lasserre, Bryan Riel, Pietro Milillo, Ziyadin Çakir, and François Renard. 2016. “An Aseismic Slip Transient on the North Anatolian Fault.” Geophysical Research Letters 43 (7): 2016GL068250. doi:10.1002/2016GL068250.
  • Segall, Paul, Allan M. Rubin, Andrew M. Bradley, and James R. Rice. 2010. “Dilatant Strengthening as a Mechanism for Slow Slip Events.” Journal of Geophysical Research 115 (December): B12305. doi:201010.1029/2010JB007449.
  • Wei, Meng, Yoshihiro Kaneko, Yajing Liu, and Jeffrey J. McGuire. 2013. “Episodic Fault Creep Events in California Controlled by Shallow Frictional Heterogeneity.” Nature Geoscience 6 (7): 566–70. doi:10.1038/ngeo1835.

Related undergraduate subjects:

  • Computer science
  • Engineering
  • Geology
  • Geophysics
  • Physics