Aftershocks, Liquefaction and Other Earthquake Hazards Explained
A large earthquake is the start of a sequence, not a single event. Here is how aftershocks behave, and why the damage a quake does depends as much on the ground as on its magnitude.
What Aftershocks Are and How Omori’s Law Describes Them
Earthquakes cluster in time and space. Within a cluster, seismologists call the largest event the mainshock, the earthquakes before it foreshocks, and those after it aftershocks (Jones and Reasenberg 1996). As a rule of thumb, an earthquake counts as an aftershock if it lies within one or two fault-rupture lengths of the mainshock. That distance varies widely: the rupture was about 15 kilometers long in the 1994 Northridge earthquake and 430 kilometers long in the 1906 San Francisco earthquake (Jones and Reasenberg 1996).
The Japanese seismologist Fusakichi Omori published the first quantitative description of aftershock decay in 1894, based on the aftershocks of the 1891 Nōbi earthquake in central Japan. His formula states that the aftershock rate falls roughly in proportion to the inverse of the time since the mainshock. Tokuji Utsu later added an exponent, p, that lets the decay run slightly faster or slower than that, and a review published a century after Omori’s paper concluded that the modified formula generally gives an appropriate description of how aftershock activity changes over time (Utsu et al. 1995).
In practical terms, the second day after a mainshock brings about half as many aftershocks as the first, and the tenth day about a tenth as many (Jones and Reasenberg 1996). Aftershocks of all sizes decline at the same rate, so a sequence does not run out of large events first (Jones and Reasenberg 1996). Seismologists in southern California noted that aftershocks of the 1992 Landers earthquake became less frequent with time rather than smaller, and that magnitude 5 events could still occur late in the sequence, only spaced farther apart (SCEDC 1999).
How Big Can an Aftershock Be? Båth’s Law
Larger mainshocks tend to produce larger aftershocks (Jones and Reasenberg 1996). The Swedish seismologist Markus Båth gave the relationship a number. Båth’s law holds that the largest aftershock is, on average, about 1.2 magnitude units smaller than the mainshock, independent of the mainshock’s magnitude (Båth 1965; Helmstetter and Sornette 2003). By that rule, a magnitude 7.0 mainshock would typically be followed by a largest aftershock near magnitude 5.8.
The value is an average across many sequences, and individual sequences depart from it in both directions. Helmstetter and Sornette (2003) showed that the law can be reproduced from two other statistical properties of earthquakes, the magnitude distribution described below and the way aftershocks trigger further aftershocks. They argued that its origin lies partly in how mainshocks and aftershocks are defined, rather than in a separate physical mechanism.
The Gutenberg–Richter Law and What the b-Value Means
Small earthquakes vastly outnumber large ones, and the ratio is regular. Beno Gutenberg and Charles Richter described the relationship in a 1944 study of California earthquakes (Gutenberg and Richter 1944). It is usually written log N = a − bM, where N is the number of earthquakes of magnitude M or larger, a reflects the overall level of activity, and b sets how quickly the count falls as magnitude rises.
The b-value is close to 1 in most global catalogues, a result well enough established that a 1993 review was subtitled “much ado about 1.0” (Frohlich and Davis 1993). With b equal to 1, each step of one magnitude unit cuts the number of earthquakes by a factor of ten. A region that records 100 earthquakes of magnitude 5 or larger in a given period would, on average, record about 10 of magnitude 6 or larger and about one of magnitude 7 or larger.
The value is not fixed. Global catalogues give b-values from roughly 0.72 to 1.34, and part of that spread reflects errors in detection, location and magnitude rather than real differences (Frohlich and Davis 1993). Some variation is genuine. Schorlemmer et al. (2005) found that b is lowest for thrust faulting and highest for normal faulting, and interpreted the difference as a sign that thrust faults operate under higher stress. In California, b decreases as earthquakes get deeper (Mori and Abercrombie 1997). A lower b-value means that large earthquakes make up a larger share of the total, which is why the parameter matters in hazard estimates.
How Long Do Aftershocks Last?
An earthquake counts as an aftershock for as long as the rate of earthquakes in the area stays above the rate before the mainshock. How long that takes depends on the size of the mainshock and on how active the region was beforehand, since aftershocks stand out longer against a quiet background (Jones and Reasenberg 1996). Eight and a half years after the 1992 Landers earthquake, aftershocks had become few and far between but still occurred clearly above the background rate (SCEDC 2001).
Toda and Stein (2018) found that the duration of a sequence depends less on the mainshock’s size than on how fast the fault slips. Studying the 2011 Tohoku-Oki earthquake and eight other large Japanese mainshocks, they found sequences that lasted about a month on the fastest-slipping faults and projected that sequences on the slowest faults would persist for more than 2,000 years. In slowly deforming regions far from plate boundaries, long-lived aftershock sequences can therefore be mistaken for ordinary background activity, which can inflate estimates of future hazard (Toda and Stein 2018).
What Makes Earthquake Shaking Damaging: Distance, Depth, Soil and Duration
Magnitude measures the size of the earthquake source, not the shaking at a particular place. The U.S. Geological Survey (USGS) lists the type of earthquake, its depth, the distance from the fault rupture and the ground material, from soft soil to bedrock, among the factors that control what people feel, along with building construction (USGS n.d.a). Shaking is generally strongest near the rupturing fault and weakens with distance. A deep earthquake places that distance between the source and everything on the surface above it.
Local ground conditions can outweigh distance. Research by the Southern California Earthquake Center (SCEC) and USGS found that in a single earthquake, shaking at one site can easily be ten times stronger than at a neighboring site the same distance from the fault. The most important local factors were the softness of the rock or soil near the surface and the thickness of sediments above hard bedrock (SCEC n.d.; Field et al. 2000).
Mexico City in 1985 is the standard example. The magnitude 8.0 earthquake of 19 September 1985 struck off the Pacific coast of Michoacán, roughly 350 to 400 kilometers from the capital (Furumura and Kennett 1998). Peak ground accelerations in the epicentral region were low, yet the damage in Mexico City was severe (Anderson et al. 1986). Central parts of the city stand on soft, water-saturated clay deposited in a former lake. Measured against a site in the city’s hill zone, the lake-bed sites amplified ground motion by factors of 8 to 56 at their natural periods (Singh et al. 1988). Duration added to the effect: strong shaking in the lake-bed zone lasted more than three minutes (Furumura and Kennett 1998). Estimates of the number of deaths vary widely; the figure attributed to USGS is at least 9,500, and other estimates are considerably higher (Live Science n.d.).
Liquefaction and Earthquake-Triggered Landslides
Liquefaction occurs when strong shaking strikes loose sand and silt saturated with water and the solid ground begins to behave like a liquid (USGS n.d.b). The shaking raises the water pressure between grains until they lose contact with one another. The ground can then lose its ability to support buildings, flow down even gentle slopes and erupt at the surface as sand boils (USGS n.d.c).
The magnitude 7.5 Niigata earthquake of 16 June 1964 caused extensive damage to engineering structures through liquefaction of sandy soil (Seed and Idriss 1967). At Kawagishi-cho, near the Shinano River, apartment buildings suffered bearing-capacity failures and tilted severely while remaining structurally almost intact (University of Washington n.d.). The analysis that Seed and Idriss (1967) published after the earthquake became a foundation of modern liquefaction assessment.
The 2010–2011 Canterbury earthquake sequence in New Zealand repeated the lesson in a modern city. The magnitude 6.2 earthquake of 22 February 2011 killed 185 people (Canterbury Earthquakes Royal Commission n.d.). Liquefaction in native soils was greater in both extent and severity than in the magnitude 7.1 Darfield earthquake of September 2010. It was most severe in residential areas east of the central business district, where strong shaking near the fault, a water table about one meter below the surface and highly susceptible soils combined. In the city center, liquefaction caused buildings to settle and tilt (Cubrinovski et al. 2011).
Landslides are another common secondary hazard. A study of 40 historical earthquakes found that rock falls, disrupted soil slides and rock slides were the most abundant types and were triggered by the weakest shaking. The maximum area likely to be affected grows from essentially zero at about magnitude 4.0 to some 500,000 square kilometers at magnitude 9.2 (Keefer 1984). The magnitude 7.9 Wenchuan earthquake in China on 12 May 2008 triggered more than 60,000 landslides (Zhang et al. 2016). Estimates of the deaths they caused range from about 20,000 (Yin et al. 2009) to 25,000–30,000 (Zhang et al. 2016).
Tsunamis form when shallow faults slip during large earthquakes and rapidly displace the seafloor vertically, most often at subduction zones (USGS n.d.d). How they travel and what to do is explained in the Tsunamis guide in the Science section.
Induced Earthquakes: Wastewater Injection in Oklahoma
Human activity can also trigger earthquakes. Pumping fluid deep underground raises the fluid pressure within faults, which can allow a stressed fault to slip (Ellsworth 2013). Hydraulic fracturing, or fracking, directly causes only a small percentage of felt induced earthquakes. Disposal of wastewater in deep wells carries a higher risk, because it injects larger volumes over longer periods (Ellsworth 2013; Rubinstein and Mahani 2015). Most wells cause no felt earthquakes: of roughly 35,000 active wastewater disposal wells and tens of thousands of other injection and fracking wells in the United States, only a few dozen were known to have induced felt earthquakes, and induced seismicity can occur 10 kilometers or more from the injection point (Rubinstein and Mahani 2015).
Oklahoma became the clearest case. According to USGS, most of the state’s earthquakes are caused by wastewater disposal, and more than 90 percent of the injected wastewater is a byproduct of oil extraction rather than spent fracking fluid. From 2014 through 2017 the state’s rate of magnitude 3 and larger earthquakes exceeded California’s (USGS n.d.e). The Oklahoma Geological Survey counted about one such earthquake per year from 1976 to 2007 and a record 903 in 2015. After state regulators ordered some wells closed and injection volumes cut in others from mid-2015, the count fell to 623 in 2016, 302 in 2017 and 196 in 2018 (Associated Press 2019).
The largest earthquake recorded in Oklahoma struck near Pawnee on 3 September 2016. The magnitude 5.8 event ruptured a previously unmapped fault in the crystalline basement rock. Yeck et al. (2017) noted that if it was induced, it would be the largest earthquake known to have been induced by fluid injection, and that it occurred after volume reductions had begun. They concluded that the long-lasting effects of injected fluid spreading through the rock make induced hazard difficult to reduce quickly (Yeck et al. 2017).
Frequently asked questions
- How long do aftershocks last after a big earthquake?
- Aftershocks continue for as long as the earthquake rate in the area stays above its level before the mainshock. On fast-slipping faults a sequence can fade within about a month, while on slowly slipping faults far from plate boundaries it may persist for centuries or longer. The rate falls quickly at first: the tenth day usually brings about a tenth as many aftershocks as the first.
- Can an aftershock be bigger than the main earthquake?
- By definition, no. The mainshock is the largest earthquake in a cluster, so if a later earthquake is larger, the first is reclassified as a foreshock. On average the largest aftershock is about 1.2 magnitude units smaller than the mainshock, a rule known as Båth’s law.
- What is the Gutenberg–Richter law?
- It describes how the number of earthquakes falls as magnitude rises. The slope, called the b-value, is close to 1 in most global catalogues, which means about ten times as many magnitude 5 earthquakes as magnitude 6 earthquakes. The b-value varies with faulting style and depth.
- Why was Mexico City so badly damaged in 1985 when the epicenter was far away?
- Central Mexico City is built on soft, water-saturated clay from a former lake. Those sediments amplified the shaking by factors of 8 to 56 compared with firmer ground in the city’s hills, and the strong motion lasted more than three minutes, even though the epicenter lay roughly 350 to 400 kilometers away.
- What is liquefaction in an earthquake?
- Liquefaction happens when shaking raises the water pressure in loose, water-saturated sand or silt until the grains lose contact and the ground behaves like a liquid. Buildings can sink or tilt, as they did in Niigata in 1964 and Christchurch in 2011.
- Does fracking cause earthquakes in Oklahoma?
- Mostly not. USGS attributes most of Oklahoma’s earthquakes to deep disposal of wastewater, most of which is a byproduct of oil extraction rather than fracking fluid. Hydraulic fracturing directly causes only a small percentage of felt induced earthquakes.
References
- Anderson, J. G., P. Bodin, J. N. Brune, J. Prince, S. K. Singh, R. Quaas, and M. Onate. 1986. "Strong Ground Motion from the Michoacan, Mexico, Earthquake." Science 233 (4768): 1043–49.
- Associated Press. 2019. "For Third Straight Year, Oklahoma Sees Decrease in Quakes." Insurance Journal, January 2, 2019. www.insurancejournal.com/news/southcentral/2019/01/02/513352.htm.
- Båth, Markus. 1965. "Lateral Inhomogeneities of the Upper Mantle." Tectonophysics 2 (6): 483–514. doi.org/10.1016/0040-1951(65)90003-X.
- Canterbury Earthquakes Royal Commission. n.d. "Canterbury Earthquakes Royal Commission." canterbury.royalcommission.govt.nz/.
- Cubrinovski, Misko, Brendon Bradley, Liam Wotherspoon, Russell Green, Jonathan Bray, Clint Wood, Michael Pender, et al. 2011. "Geotechnical Aspects of the 22 February 2011 Christchurch Earthquake." Bulletin of the New Zealand Society for Earthquake Engineering 44 (4): 205–26. doi.org/10.5459/bnzsee.44.4.205-226.
- Ellsworth, William L. 2013. "Injection-Induced Earthquakes." Science 341 (6142). doi.org/10.1126/science.1225942.
- Field, Edward H., and the SCEC Phase III Working Group. 2000. "Accounting for Site Effects in Probabilistic Seismic Hazard Analyses of Southern California: Overview of the SCEC Phase III Report." Bulletin of the Seismological Society of America 90 (6B): S1–31. doi.org/10.1785/0120000512.
- Frohlich, Cliff, and Scott D. Davis. 1993. "Teleseismic b Values; or, Much Ado about 1.0." Journal of Geophysical Research 98 (B1): 631–44. doi.org/10.1029/92JB01891.
- Furumura, T., and B. L. N. Kennett. 1998. "On the Nature of Regional Seismic Phases—III. The Influence of Crustal Heterogeneity on the Wavefield for Subduction Earthquakes: The 1985 Michoacan and 1995 Copala, Guerrero, Mexico Earthquakes." Geophysical Journal International 135: 1060–84.
- Gutenberg, B., and C. F. Richter. 1944. "Frequency of Earthquakes in California." Bulletin of the Seismological Society of America 34 (4): 185–88. doi.org/10.1785/BSSA0340040185.
- Helmstetter, Agnès, and Didier Sornette. 2003. "Båth’s Law Derived from the Gutenberg-Richter Law and from Aftershock Properties." Geophysical Research Letters 30 (20): 2069. doi.org/10.1029/2003GL018186.
- Jones, Lucile M., and Paul A. Reasenberg. 1996. "Some Facts about Aftershocks to Large Earthquakes in California." U.S. Geological Survey Open-File Report 96-266. doi.org/10.3133/ofr96266.
- Keefer, David K. 1984. "Landslides Caused by Earthquakes." Geological Society of America Bulletin 95 (4): 406–21. doi.org/10.1130/0016-7606(1984)95<406:LCBE>2.0.CO;2.
- Live Science. n.d. "Seismic History: The Deadly 1985 Mexico City Earthquake." www.livescience.com/30791-deadly-1985-mexico-city-earthquake.html.
- Mori, Jim, and Rachel E. Abercrombie. 1997. "Depth Dependence of Earthquake Frequency-Magnitude Distributions in California: Implications for Rupture Initiation." Journal of Geophysical Research 102 (B7): 15081–90. doi.org/10.1029/97JB01356.
- Rubinstein, Justin L., and Alireza Babaie Mahani. 2015. "Myths and Facts on Wastewater Injection, Hydraulic Fracturing, Enhanced Oil Recovery, and Induced Seismicity." Seismological Research Letters 86 (4): 1060–67. doi.org/10.1785/0220150067.
- SCEC (Southern California Earthquake Center). n.d. "Phase III Fact Sheet: Finding the Hotspots." ceo.scec.org/phase3/factsheet.html.
- SCEDC (Southern California Earthquake Data Center). 1999. "Earthquake Commentary for May, 1999." service.scedc.caltech.edu/ftp/ca.earthquakes/1999/May1999.html.
- SCEDC. 2001. "Earthquake Commentary for November, 2001." service.scedc.caltech.edu/ftp/ca.earthquakes/2001/November2001.html.
- Seed, H. Bolton, and I. M. Idriss. 1967. "Analysis of Soil Liquefaction: Niigata Earthquake." Journal of the Soil Mechanics and Foundations Division 93 (3): 83–108. doi.org/10.1061/JSFEAQ.0000981.
- Singh, S. K., J. Lermo, T. Domínguez, M. Ordaz, J. M. Espinosa, E. Mena, and R. Quaas. 1988. "The Mexico Earthquake of September 19, 1985—A Study of Amplification of Seismic Waves in the Valley of Mexico with Respect to a Hill Zone Site." Earthquake Spectra 4 (4): 653–73. doi.org/10.1193/1.1585496.
- Toda, Shinji, and Ross S. Stein. 2018. "Why Aftershock Duration Matters for Probabilistic Seismic Hazard Assessment." Bulletin of the Seismological Society of America 108 (3A): 1414–26. doi.org/10.1785/0120170270.
- University of Washington. n.d. "Niigata, Japan Earthquake." Soil Liquefaction Web Site, Department of Civil and Environmental Engineering. depts.washington.edu/liquefy/html/quakes/niigata/niigata.html.
- USGS (U.S. Geological Survey). n.d.a. "Earthquakes: What You Feel." www.usgs.gov/media/images/earthquakes-what-you-feel.
- USGS. n.d.b. "EarthWord: Liquefaction." www.usgs.gov/news/earthword-liquefaction.
- USGS. n.d.c. "San Francisco Bay Region Geology and Geologic Hazards: About Liquefaction." Archived copy. webarchive.library.unt.edu/eot2008/20081108052822/http://geomaps.wr.usgs.gov/sfgeo/liquefaction/aboutliq.html.
- USGS. n.d.d. "Tsunamis." Subduction Zone Science. www.usgs.gov/special-topics/subduction-zone-science/science/tsunamis.
- USGS. n.d.e. "Oklahoma Has Had a Surge of Earthquakes since 2009. Are They Due to Fracking?" www.usgs.gov/faqs/oklahoma-has-had-surge-earthquakes-2009-are-they-due-fracking.
- Utsu, Tokuji, Yosihiko Ogata, and Ritsuko S. Matsu’ura. 1995. "The Centenary of the Omori Formula for a Decay Law of Aftershock Activity." Journal of Physics of the Earth 43 (1): 1–33. doi.org/10.4294/jpe1952.43.1.
- Yeck, W. L., G. P. Hayes, D. E. McNamara, J. L. Rubinstein, W. D. Barnhart, P. S. Earle, and H. M. Benz. 2017. "Oklahoma Experiences Largest Earthquake during Ongoing Regional Wastewater Injection Hazard Mitigation Efforts." Geophysical Research Letters 44: 711–17. doi.org/10.1002/2016GL071685.
- Yin, Yueping, Fawu Wang, and Peng Sun. 2009. "Landslide Hazards Triggered by the 2008 Wenchuan Earthquake, Sichuan, China." Landslides 6 (2): 139–52. doi.org/10.1007/s10346-009-0148-5.
- Zhang, Shuai, Limin Zhang, Suzanne Lacasse, and Farrokh Nadim. 2016. "Evolution of Mass Movements near Epicentre of Wenchuan Earthquake, the First Eight Years." Scientific Reports 6: 36154. doi.org/10.1038/srep36154.