Could several California faults rupture together as in the M=7.8 New Zealand earthquake?

By Ross Stein and David Jacobson, Temblor

Check your earthquake risk

Century City lies at the northern tip of the Newport-Inglewood Fault in Los Angeles. The base of the Santa Monica mountains trace the Santa Monica Fault.
Century City lies at the northern tip of the Newport-Inglewood Fault in Los Angeles. The base of the Santa Monica Mountains right behind the buildings, traces the active Santa Monica Fault.

 

For several decades, geologists assumed that faults more than 5 km (3 mi) apart could not rupture together. This inference was based on examination the surface ruptures of all M≥7 shocks around the world by geologists such as Steve Wesnousky (Univ. Nevada Reno), as well as by numerical modeling of rupture propagation by Ruth Harris (USGS) and Steve Day (San Diego State Univ.). If you asked earthquake researchers how to interpret this work, they would say that for faults separated by less than a kilometer, through-going ruptures are possible; if separated by more than a few kilometers, rupture is rare; and if more than about 5 km, jumping from one fault to another becomes extremely unlikely. Further, the width of rupture along the fault can be as narrow as several centimeters (an inch), and rarely exceeds several kilometers. The caveat is that faults are three-dimensional, and that we are condemned to live on the surface of the Earth, and so faults widely separated at the Earth’s surface could be connected at depth.

 

Kaikoura changed everything

But in one fell swoop, the Kaikoura earthquake chucked this hallowed view out the window—or at least showed that there are glaring exceptions.

In the earth sciences, good data always trumps good theory. Seismology is an experimental science in which you cannot plan your own experiments: Mother Nature plans them for you. Each large, well recorded event furnishes not only a chance to learn, but also the prospect to falsify the current dogma or consensus. What a gift this is for young scientists, because evidence, not eminence, rules.

 

kaikoura-new-zealand-earthquake-rupture-map-los-angeles-fault-map
Left-hand panel shows in blue the fault rupture in the M=7.8 Kaikoura earthquake as mapped by Litchfield al. (2017), with their approximate locations digitized by Temblor, and the active faults from GNS Science in red. Compare this to the discontinuous segments of the Newport-Inglewood Fault extending south from Los Angeles: All those segments are closer together than the faults that ruptured in the New Zealand earthquake. The blue pin lies at Century City.

 

In the New Zealand event, faults as much as 20 km (12 mi) apart ruptured together, and the width of the faulting along parallel faults was as much as 40 km (25 mi). Was this unprecedented? Probably not. Rather, thanks to New Zealand geologists and geophysicists, the fault slip was very well recorded and extensively documented. Publications have just begun to emerge about this quake, and they make two things clear:

• Mapped active faults did only a fair job of foretelling where the faults would rupture

• Faults 25 km apart can indeed rupture together

 

The faults unzipped at 3300 mph (5300 kph), pausing briefly before the jumps

The story at the Earth’s surface is also fairly consistent with the slip inferred at depth by Anna Kaiser et al. (2017) and independently by James Hollingsworth et al. (2017): They both show fault ruptures separated by ~20 km. Here, below, the energy recorded at seismometers is projected backwards to the source, a technique used to watch ruptures propagate. The earthquake nucleates at the star and unzips northeast for 20 seconds (purple to blue colored dots), then it jumps ~15 km (9 mi) to the north and ruptures for another 20 seconds (blue to turquoise colored dots). Then it jumps north for the third time and ruptures for another 40 seconds (green to yellow dots). The entire leaping rupture process lasted an astonishing 2 minutes, the same amount of time that it took the 2011 Tohoku M=9.0 earthquake, which was 40 times larger.

 

kaikoura-earthquake-rupture
Rupture propagation (colored dots) and energy release (grey to black dots) from “back projection” of seismic recordings of the New Zealand earthquake by Kaiser et al. (2017) show that even at depth, the rupture jumped 20 km three times during the earthquake, each time hanging for about 5-10 sec before making the leap.

 

What does this mean for the seismic hazards in LA and SF?

Through the prism of Kaikoura, two fault systems jump out that deserve another look. In southern California, the Newport-Inglewood Fault has always been an oddity, because it is straight but highly discontinuous, with as much as 10 km between mapped active faults. But jumping from one to another now looks much more possible than it did before.

In the San Francisco Bay Area, the northern tip of the Calaveras Fault is separated by 10 km from the southern tip of the Concord Fault, and 15 km from the parallel Greenville Fault. Two, or perhaps all of these could rupture together, creating a much larger earthquake.

 

new-zealand-earthquake-rupture-map-san-francisco-bay-area-fault-map
Left-hand panel shows in blue the fault rupture in the M=7.8 Kaikoura shock as mapped by Litchfield et al. (2017). Compare this to the Concord, Calaveras, and Greenville Faults east of San Francisco and San José; they are closer together than those in the New Zealand earthquake. If Bay Area faults ruptured together a M≥7.5 shock would be possible.

 

The Newport-Inglewood Fault, with a 1 mm/yr slip rate, has 1.0-1.5% per year chance of rupturing in a M≥6.7 shock in the next 30 years. If that sounds unlikely, think again: It means there is 2 chances out of 3 that it will strike in a typical lifetime, and the consequences of such a quake would be immense, tearing through the densely populated south central Los Angeles corridor. Such a rupture was included in current USGS model (Field et al., 2014), and that judgement now seems validated.

The Calaveras and Concord Faults have slip rates of 3.5-6.0 mm/yr, and about a 1% per year chance of rupturing. But the combined ruptures were not considered in the 2014 USGS hazard model, and so much larger, longer duration ruptures now seem possible.

 

What can you do about it?

If you live in an older (pre-1977) home, retrofit it. A $5,000 retrofit can likely save $100,000 in damage in a M~7 earthquake. Second, consider earthquake insurance. Because the insurance companies are using the older, 2008 USGS model, they are not considering any ruptures that can jump from one fault to another, regardless of distance. The possibilities of the long jumps seen in Kaikoura mean that companies are probably selling insurance for less than its true risk-based price, so you will be effectively getting a discount.

 

References

Edward H. Field, Glenn P. Biasi, Peter Bird, Timothy E. Dawson, Karen R. Felzer, David D. Jackson, Kaj M. Johnson, Thomas H. Jordan, Christopher Madden, Andrew J. Michael, Kevin R. Milner, Morgan T. Page, Tom Parsons, Peter M. Powers, Bruce E. Shaw, Wayne R. Thatcher, Ray J. Weldon II, and Yuehua Zeng (2015), Long-Term Time-Dependent Probabilities for the Third Uniform California Earthquake Rupture Forecast (UCERF3), Bull. Seismol. Soc. Amer., 105, 511–543, doi: 10.1785/0120140093

Ruth A. Harris and Steven M. Day (1993). Dynamics of fault interaction–parallel strike-slip faults, J. Geophys. Res. 98, 4461–4472.

James Hollinsworth, Lingling Yec, Jean-Philippe Avouac, Dynamically triggered slip on a splay fault in the Mw 7.8, 2016 Kaikoura (New Zealand) earthquake (2017), Geophys. Res. Letts., doi: 10.1002/2016GL072228.

Anna Kaiser, N. Balfour, B. Fry, C. Holden, N. Litchfield, M. Gerstenberger, E. D’Anastasio, N. Horspool, G. McVerry, J. Ristau, S. Bannister, A. Christophersen, K. Clark, W. Power, D. Rhoades, C. Massey, I. Hamling, L. Wallace, J. Mountjoy, Y. Kaneko, R. Benites, C. Van Houtte, S. Dellow, L. Wotherspoon, K. Elwood, and K. Gledhill (2017), The 2016 Kaikoura, New Zealand, Earthquake: Preliminary Seismological Report, Seismol. Res. Letts., 88, doi: 10.1785/0220170018.

Steven G. Wesnousky, (2006). Predicting the endpoints of earthquake ruptures, Nature 444, 358–360, doi: 10.1038/nature05275.

Nicola J. Litchfield, Benson A, Bischoff A, Hatem A, Barrier A, Nicol A, Wandres A, Lukovic B, Hall B, Gasston C, Asher C, Grimshaw C, Madugo C, Fenton C, Hale D, Barell DJA, Heron DW, Strong DT, Townsend DB, Noble D, Howarth JD, Pettinga J, Kearse J, Williams J, Manousakis J, Mountjoy J, Rowland J, Clark KJ, Pedley K, Sauer K, Berryman KR, Hemphill-Haley M, Stirling MW, Villeneuve M, Cockcroft M, Khajavi N, Barnes P, Villamor P, Carne R, Langridge RM, Zinke R, Van Dissen RJ, McColl S, Cox SC, Lawson S, Little T, Stahl T, Cochran UA, Toy V, Ries WF, Juniper Z. 2017. 14th November 2016 M7.8 Kaikoura Earthquake. Summary surface fault rupture traces and displacement measurements. GNS Science. http://dx.doi.org/10.21420/G2

  • Edwin Nissen

    At the risk of shameless self-promotion, this 20 km jump does not come as a total surprise. In fact we imaged a fault-to-fault jump in seismic rupture of ~50 km in an obscure Mw 7 earthquake in Pakistan in 1997. It was listed in all catalogues as a single event and it required InSAR + back-projection to reveal that two separate faults had ruptured. Details here:
    http://www.nature.com/ngeo/journal/v9/n4/full/ngeo2653.html

    • Ross Stein

      Thank you, Edwin. That’s not self promotion; that’s evidence. Thank you for pointing this out.