When the fault rings twice: repeated ruptures on the same fault stretch

The recent publication of a paper on the Weitin Thrust (Papua New Guinea) by Chen, Milliner and Avouac (Fig. 1) gave me the opportunity to dig out and look back to some notes I wrote few months ago. Chen et al. use optical image correlation to document coseismic surface ruptures along the Weitin Thrust occurred in a Mw 8.0 event in 2000 and in a Mw 7.7 event in 2019. The ruptures overlap along a 20-km long portion, with 3-4 m of slip (Fig. 2).

Figure 1
Figure 2: Comparison of surface fault slip for the 2019 Mw 7.7 and 2000 Mw 8.0 earthquakes.

I’m particularly interested in re-rupturing: I’m Italian and the first ground-rupturing event I studied in the field was the 2016 Central Italy seismic sequence, when the same fault stretch ruptured twice on August 24th and October 30th, 2016 (Fig. 3). If I have to recall a single thing that surprised the scientific community, including researchers much more experienced than me, that is re-rupture.

Figure 3: the same spot along the road to Castelluccio di Norcia ruptured in the August 24th 2016 earthquake and in the October 30th earthquake.

I don’t want to get lost in defining how close in time ruptures should be, let’s simplify things saying “somehow below the resolution of dating techniques, so that they cannot be recognized by paleoseismic investigations” – after all, I’m writing on paleoseismicity.org 😊 

So, I started to look for other case histories and here’s what I found (some are well-constrained, other less):

  • Ridgecrest (USA): Mw 6.4 on July 4th and Mw 7.1 on July 5th, 2019;
  • Weitin Thrust (PNG): Mw 8.0 in 2000 and Mw 7.7 in 2019;
  • Lake Miur (Australia): Mw 5.3 on September 16th and Mw 5.2 on November 8th, 2018;
  • 2017 Hojdek triplet (Iran);
  • Mw 6.5 and Mw 7.3 Kumamoto (Japan) earthquakes in 2016;
  • Central Italy: Mw 6.0 and Mw 6.5 in 2016, three months apart;
  • Mw 7.4 Izmit and Mw 7.1 Duzce (Turkey) earthquakes in 1999, 3 months apart;
  • 1981 Sirch and 1998 Fandoqa (Iran) earthquakes;
  • Mw 7.1 Dasht-e-Bayaz 1968 and Mw 7.1 Khuli-Buniabad 1979 (Iran);
  • Ms 7.0 and 7.1 in 1957 and 1967 at the western end of the North Anatolian Fault (Turkey);
  • 1936, 1979 and Mw 7.2 1997 Zirkuh (Iran);
  • Mw 7.0 1940 El Centro earthquake and Imperial Valley 1979;
  • 1944 and 1970 earthquakes in Gediz (Turkey).

It seems re-rupture is the new black

I can’t see a common pattern in the above-listed case histories: they are different in terms of faulting style, plate motion rates, slip distribution, range of magnitudes and evolutive pattern (sometime the strongest shock came first, sometime not). The list is obviously far from complete and I would love to collect other case histories!

Re-rupture is challenging because it forces to rethink – or even abandon? – some assumptions widely-adopted in earthquake geology:

  • Paleoseismology: in most cases, dating techniques will never be able to disentangle events occurred few weeks-years apart. The slip we see in a trench is due to a single event or to multiple events?
  • To make things even more complex, distributed faulting lies at the opposite end of the spectrum of surface faulting behaviors: slip may occur on faults other than the seismogenic source. How to recognize in a trench if slip is due to distributed faulting?
  • Recurrence models and hazard assessment: re-rupture totally alters the estimation of recurrence intervals and slip per event. To which extent it changes the output of hazard codes or earthquake forecasting?
  • Coulomb stress transfer: it is usually assumed that all the accumulated stress is released during an earthquake and the stress goes to zero; is this assumption still valid?

Sometimes I have the impression that my research is guided by the drunkard’s search principle, the observational bias that occurs when people only search for something where it is easiest to look. Our observations are still too limited and shed a faint light on the earthquake process; analyzing re-rupturing events may improve our understanding and ultimately help in coping with seismic risk.


Chen, K., Milliner, C., & Avouac, J.‐P. (2019). The Weitin Fault, Papua New Guinea, Ruptured Twice by Mw 8.0 and Mw 7.7 Earthquakes in 2000 and 2019. Geophysical Research Letters, 46. https://doi.org/10.1029/2019GL084645

submit to reddit

Francesca Ferrario

I work at the Insubria University in Como, Italy (http://perigeo.uninsubria.it/) I'm interested in active tectonics and late Quaternary environmental evolution aimed at mitigatin natural hazards.

See all posts Francesca Ferrario


  • Luke Wedmore | 2019-11-28|17:44 (UTC)

    Great stuff Francesca – but don’t forget that during the 2016 central Italy sequence, the northern end of the Vettore fault near Mt Bove ruptured 4 days apart in the Visso and Norcia events (see Walters et al., 2018 and Wedmore et al., 2019). Helpfully Eutizio Vittori alerted us to the initial rupture following the Visso earthquake and we were able to visit the fault prior to the Norcia event!

  • David P. Schwartz | 2019-11-29|17:06 (UTC)

    This is an interesting issue. You have some examples that I wasn’t aware of. I touch on this in a 2018 paper in the Bulletin of the Seismological Society of America : David P Schwartz, Review: Past and Future Fault Rupture Lengths in Seismic Source Characterization—the Long and Short of It. V108, p. 2493-2520. See table 3: List of Repeat Historical Ruptures along the Same Fault Section.

    Maybe we can put our heads together on this.


  • David Schwartz | 2019-11-29|17:25 (UTC)

    Interesting list. You found some I wasn’t aware of. See Table 3 of Schwartz.David P., 2018, Review:: Past and Future Rupture Lengths in Seismic Source Characterization—The Long and Short of It, Bulletin of the Seismological Society of America, v. 108(5A), 2493-2520.

    We should put our heads together!

Leave a Comment