Coulomb stress calculations and the distribution of aftershocks indicate that the early December magnitude 7.0 earthquake likely stressed the northern end of the San Andreas, near the end of the 1906 San Francisco earthquake rupture.
By Volkan Sevilgen, Temblor, Inc., Shinji Toda , IRIDeS, Tōhoku University, Hector Gonzalez-Huizar, CICESE, and Ross S. Stein, Temblor, Inc.
Citation: Sevilgen, V., Toda, S., Gonzalez-Huizar, H., and Stein, R.S., 2024, Did the 2024 magnitude 7.0 Cape Mendocino earthquake trigger aftershocks on the San Andreas?, Temblor, http://doi.org/10.32858/temblor.355
On the morning of Thursday, Dec. 5, at 10:44 a.m. local time (Dec. 5, 18:44 UTC), a magnitude 7.0 earthquake struck offshore Northern California, about 100 kilometers (about 60 miles) west of the town of Ferndale, where a magnitude 6.4 event struck in 2022. According to the U.S. Geological Survey (USGS), the magnitude 7.0 strike-slip event occurred at a depth of about 10 kilometers (6 miles).
The National Weather Service issued a tsunami warning for parts of coastal Oregon and California, likely because of the initial magnitude of 7.3 (soon revised down to 7.0 as more data came in), shallow depth, and location of the event. The warning was broadcast as far south as Davenport, California (near Santa Cruz) and northward into Oregon, but was lifted about an hour later, when it became evident that no destructive tsunami was forthcoming.
Although the tsunami warning may have been a surprise, the earthquake itself was not. This particular region is among the most seismically active parts of the United States. In this preliminary analysis, we consider what happened and why and explore implications for future events. We also present evidence of dynamic triggering — a phenomenon by which seismic waves from one earthquake cause faults at a distance from the mainshock to rupture.
The collision of three tectonic plates
Just east of the earthquake’s epicenter is what’s called the Mendocino Triple Junction, where the Pacific, Juan de Fuca and North American plates — and the faults separating them — all meet in a chaotic crunch.
The San Andreas Fault, which separates the Pacific and North American plates through California, arrives at the Mendocino Triple Junction from the southeast. It takes a sharp left turn to meet the Mendocino Fault, a transform fault that stretches thousands of kilometers west, separating the Pacific Plate to the south from the Juan de Fuca Plate to the north. To the north, the Cascadia Subduction Zone accommodates the Juan de Fuca Plate sliding beneath North America, mostly in very large earthquakes every few hundred years (although just how Cascadia ruptures is still an active question for earthquake scientists).
At the Mendocino Triple Junction, the plates get distorted and misaligned, causing them to fracture, tear and deform in frequent earthquakes. Since 1980, this region has experienced six earthquakes of magnitude 7 or larger, including Thursday’s 7.0 shock. (For comparison, the destructive 1989 Loma Prieta earthquake was a magnitude 6.9 event.) Aftershocks and seismologic constraints suggest that Thursday’s quake occurred on the Mendocino Fault (Figure 1).
Seismic gap
A seismic gap is a section of an active fault that hasn’t experienced a significant earthquake in recent history, in between segments that have. Rollins and Stein (2010) identified a gap along the Mendocino Fault in between a magnitude 7.0 event in 1994 and a magnitude 6.1 event in 1983 (Figure 2). The December 2024 earthquake’s aftershock distribution suggests that the mainshock occurred in this gap, though some aftershocks extended beyond the previously identified gap boundaries toward the San Andreas Fault, possibly indicating stress transfer to adjacent fault segments (discussed further below).
Although the epicenter was about 50 kilometers offshore, aftershocks within the first 24 hours extended eastward to within a few kilometers from the coast. This suggests that the rupture propagated eastward from the epicenter toward the coastline (Figure 1).
Ground-shaking models
Because of the large magnitude and relatively shallow depth, the earthquake was felt as far north as Salem, Oregon, and throughout much of Northern California, including the San Francisco Bay Area and Central Valley. As of this writing, more than 16,000 people had submitted shaking reports through the USGS Did You Feel It? questionnaire.
Cape Mendocino — in line with the eastward-propagating rupture — recorded the highest ground shaking measured by seismic sensors, approximately 0.5 g, or 0.5 times the force of gravity. (This level of acceleration is similar to what passengers experience during emergency braking in an automobile.) Elsewhere, instruments and people alike reported mostly light shaking.
These shaking data provide an opportunity to assess how well Temblor’s global ground motion model (which uses location, magnitude, depth, and Temblor SiteAmp Vs30 data at 100-meter worldwide resolution) can predict shaking levels from new earthquakes. We directly compare the Temblor ground motion model predictions to observations — both felt reports and instrumental measurements (Figure 3). To do so, we downloaded the USGS Did You Feel It? report locations and corresponding shaking intensities. We also downloaded seismic station coordinates and measured shaking (Peak Ground Acceleration in g).
The Temblor ground motion model successfully predicts much of the observed shaking distribution of last Thursday’s earthquake. The Temblor ground motion model might slightly overestimate shaking at levels that won’t cause any damage far from the epicenter. We also note that at these distances, some of the observations are higher than our model. On balance, we consider our ground motion model to be consistent with observed shaking, especially at close distances to the rupture — where it matters the most.
Not a surprise
Out of anywhere in California, Cape Mendocino is the place that is forecast to experience the largest earthquake in any given 100-year period, per Temblor’s earthquake occurrence model (Figure 4). Looking at a specific magnitude, the model indicates that a magnitude 7.2 shock has a 1% per year likelihood. This may seem like a remote possibility, but it means that the likelihood of experiencing a magnitude 7.2 earthquake in 80 years — the average human lifetime — is 55%. In fact, this region was rocked by earthquakes in 1922, 1980, 1992 and 2005 that were magnitude 7.2 or larger, and several other recorded earthquakes have come close.
Impact on the San Andreas and Cascadia
The infamous 1906 San Francisco earthquake ruptured 477 kilometers (296 miles), from San Juan Bautista, which is southeast of the San Francisco Bay Area, to the Mendocino Triple Junction (see Prentice et al., 1999). An important question is how much stress was transferred to the northern San Andreas by the December 5 event. Because the Cascadia subduction zone also originates here, and is capable of megathrust events, we also consider whether stress was transferred to the subduction zone.
Most of the Mendocino Fault was free of any earthquakes in the seven days preceding the magnitude 7.0 Cape Mendocino mainshock (Figure 5a). However, two foreshocks occurred: a magnitude 4.4 event 7 hours prior (2024-12-04 11:36:36 UTC), and a magnitude 4.2 event 5.5 hours prior (2024-12-04 13:10:22 UTC).
In the seven days following Cape Mendocino mainshock, aftershocks were densely clustered near the epicenter, along the Mendocino Fault. Additional activity extended eastward and southward. Approximately 15 aftershocks were observed well east of the rupture zone. Among these, strings of aftershocks appear to extend along the northernmost San Andreas Fault and Whale Gulch-Bear Harbor fault zone located 5 kilometers inland, where the San Andreas and Mendocino Faults connect (Figure 5b).
Based on the USGS source model of the mainshock and the available focal mechanisms of the aftershocks and other, prior, earthquakes, we calculate that the northern tip of the San Andreas Fault — regardless of where its trace lies — was brought about 0.25 bar closer to failure by the magnitude 7.0 shock (Figure 6a). In support of this calculation, we see some aftershocks of the magnitude 7.0 event on or near the northern tip of the San Andreas Fault (Figure 5b). But, we don’t know if that section of the San Andreas Fault last ruptured in 1906 (in which case it has since accumulated a 3 meter slip deficit), or much earlier (in which case the slip deficit is at least 6 meters). The magnitude 7.0 event also brought the southern tip of the Cascadia megathrust fault 0.25 bar closer to failure, although over a more restricted section of the fault (Figure 6b).
Another way to assess the stress transfer from the magnitude 7.0 earthquake is to calculate the stress on surrounding focal mechanisms (the “beachballs”) for all available shocks since 1976 (Figure 7). Those mechanisms give the actual orientation of the faults that receive stress. If they turn red, slip on those faults has been promoted; if they turn blue, slip on those faults has been inhibited by the magnitude 7.0 mainshock.
We see that beachballs along the Mendocino Fault and extending to and onto the San Andreas Fault turned red, indicating promotion (Figure 7a). In contrast, no thrust mechanisms exist on the southern Cascadia megathrust, and so earthquake promotion there is equivocal. We made the same kind of calculation after the nearby 2022 magnitude 6.4 Ferndale earthquake (Stein et al., 2022), which is shown in Figure 7b. We also conducted a joint Coulomb stress change analysis of the two earthquakes (Figure 7c).
The comparison of the two quakes (Figure 7a and 7b) reveals two insights: First, the magnitude 6.4 shock did not bring the epicenter of the magnitude 7.0 shock closer to failure, so we see no simple chain reaction between these two shocks. Second, the magnitude 6.4 and magnitude 7.0 shocks each brought the northwest tip of the San Andreas Fault closer to failure by about 0.25 bar. Per Figure 7c, we see that a 40-kilometer-long (about 25-mile-long) fault section, where the San Andreas merges into the Mendocino, is 0.25-0.5 bar closer to failure than it was before 2022 — a significant amount, but certainly not so large as to suggest a San Andreas event is imminent.
Interestingly, the magnitude 6.4 shock also inhibited failure by about 0.25 bar on the northernmost part of the San Andreas Fault that remains onshore. This section of the fault may have returned to its stress state prior to the Ferndale earthquake.
Earthquake interactions
Three years earlier, on December 20, 2021, the magnitude 6.2 Petrolia earthquake occurred within the Gorda plate (Hellweg et al., 2024). Just 11 seconds prior, a magnitude 5.7 foreshock struck, and about 5.5 hours earlier, a magnitude 4.0 event occurred 7 kilometers away (Figure 8a; Yoon and Shelly, 2024). The two preceding events likely triggered each other, based on their similar focal mechanisms. Their locations align with the Mendocino Fault, unlike the Petrolia earthquake.
Although the Mendocino Fault showed little activity before the mainshock aside from these two events, the Petrolia earthquake’s aftershocks activated both the Mendocino fault and a large cluster of inland seismicity (Figure 8b). Are these aftershocks instead a result of the magnitude 5.7 earthquake? Although the Blaser et al. (2010) formula predicts an 8 kilometer rupture length for a magnitude 5.7 earthquake, the aftershock pattern along the Mendocino Fault extended 50 kilometers, comparable to the aftershock extent of a magnitude 7.0 event. This suggests that both the magnitude 6.2 and magnitude 5.7 events may have — together — triggered aftershocks on the Mendocino Fault.
Interestingly, the 2022 magnitude 6.4 Ferndale earthquake showed a different pattern. In the week before, the Mendocino Fault remained largely quiet (Figure 8c), with no aftershocks from the previous year’s events. The aftershock sequence of the Ferndale earthquake, observed over seven days, concentrated inland in a northeasterly direction (Figure 8d), diverging from any surface fault orientation. This likely reflects deformation within the Gorda Plate, as suggested by Yoon and Shelly (2024). Furthermore, few aftershocks of the 2022 Ferndale earthquake occurred on the Mendocino Fault, unlike the 2021 Petrolia events (Figure 8d).
Dynamically triggered seismicity?
Dynamic triggering occurs when seismic waves from a large earthquake temporarily alter stress conditions on distant faults, causing them to fail. Unlike aftershocks, which occur near the mainshock, dynamically triggered events can occur hundreds of kilometers away and are identified by their timing relative to the passing seismic waves and statistical comparison with background seismicity rates.
The seismic waves from the magnitude 7.0 earthquake appear to have triggered other small earthquakes far from its epicenter (Bradley and Hubbard, 2024). We observed at least three earthquakes — with epicenters between 253 and 314 kilometers (157 and 195 miles) to the southeast of the mainshock — that occurred a few minutes after the mainshock (Figure 9).
About 250 kilometers (155 miles) southeast, near Clear Lake, California, a magnitude 4.3 event, the Cobb earthquake, occurred 3 minutes after the magnitude 7.0 event (Figure 9). The region in which this event occurred has been seismically active for several years. However, an earthquake of this size has not occurred in nearly three months. This earthquake apparently triggered its own aftershocks, resulting in an evident increase in the seismicity rate in this region (outlined by the red-colored square in Figure 9).
The other two possibly triggered events occurred 9 and 13 minutes after the magnitude 7.0 earthquake, and both had a magnitude of 1.9. No previous events were reported for almost two days in the vicinity of these events. The most recent event was a magnitude 1.3 event that occurred 47 hours and 11 minutes before the magnitude 7.0 mainshock.
Although these (likely) triggered events are relatively small in magnitude, understanding and reporting such cases is important for scientists who are trying to understand this process. Indeed, dynamic triggering has been well-documented elsewhere (Gonzalez-Huizar, 2012; Gonzalez-Huizar and Toda, 2021). In particular, the greatest quakes have dynamically triggered large earthquakes around the world (Pollitz et al., 2012; Sevilgen et al., 2012).
A curious quake
At 3:08 p.m. local time (23:08 UTC) on Dec. 9, 2024, a magnitude 5.8 earthquake struck near Yerington, Nevada (Figure 1). Approximately 500 kilometers (310 miles) from the magnitude 7 mainshock and 70 kilometers (44 miles) east of Lake Tahoe, two notable seismic events occurred before the magnitude 5.8 Yerington earthquake. A magnitude 3.8 earthquake occurred 24 hours before the magnitude 7.0 Cape Mendocino earthquake. After the Cape Mendocino earthquake, a magnitude 2.8 event occurred in the same place just 44 minutes later.
Is there some relationship between the magnitude 5.8 Yerington earthquake that shook this part of Nevada on Dec. 9 and and the Cape Mendocino earthquake that struck on Dec. 5? The temporal and spatial clustering of these relatively uncommon events suggests possible delayed dynamic triggering, a phenomenon where seismic waves from a large earthquake can influence fault systems at considerable distances. Such a mechanism could initiate seismic activity hours or even days after the mainshock (Castro et al., 2015). However, such a relationship has not been demonstrated convincingly thus far.
A final word
The Dec. 5, 2024, magnitude 7.0 Cape Mendocino earthquake occurred in one of California’s most seismically active regions, and so, this event wasn’t a surprise. Thankfully, the earthquake’s impact on populated areas was relatively low due to its remote location. The event likely brought the northern terminus of the 1906 San Francisco earthquake slightly closer to failure.
Science editor: Dr. Alka Tripathy-Lang, Ph.D.
Reviewer: Chris Rollins, Ph.D.
References
Blaser, L., Krüger, F., Ohrnberger, M., & Scherbaum, F. (2010). Scaling relations of earthquake source parameter estimates with special focus on subduction environment. Bulletin of the Seismological Society of America, 100(6), 2914-2926.
Castro, R. R., González‐Huízar, H., Ramón Zúñiga, F., Wong, V. M., & Velasco, A. A. (2015). Delayed dynamic triggered seismicity in northern Baja California, México caused by large and remote earthquakes. Bulletin of the Seismological Society of America, 105(4), 1825-1835. doi:10.1785/0120140310.
Gonzalez-Huizar, H., Velasco, A. A., Peng, Z., & Castro, R. R. (2012). Remote triggered seismicity caused by the 2011, M9.0 Tohoku-Oki, Japan earthquake. Geophysical Research Letters, 39(10), L10302. https://doi.org/10.1029/2012GL051015
Gonzalez-Huizar, H., & Toda, S. (2021). New Zealand sees exotic earthquake sequence. Temblor. http://doi.org/10.32858/temblor.160
Hellweg, M., Dreger, D. S., Lomax, A., McPherson, R. C., & Dengler, L. (2024). The 2021 and 2022 North Coast California Earthquake Sequences and Fault Complexity in the Vicinity of the Mendocino Triple Junction. Bulletin of the Seismological Society of America. https://doi.org/10.1785/0120240023
Bradley, K., Hubbard, J. (2024). Magnitude 7.0 earthquake strikes offshore Cape Mendocino, California. Earthquake Insights. https://doi.org/10.62481/a3404512
Pollitz, F. F., Stein, R. S., Sevilgen, V. & Bürgmann, R. (2012). The 11 April 2012 east Indian Ocean earthquake triggered large aftershocks worldwide. Nature 490, 250–253.
Prentice, C. S., Merritts, D. J., Beutner, E. C., Bodin, P., Schill, A., & Muller, J. R. (1999). Northern San Andreas fault near Shelter Cove, California. Geological Society of America Bulletin, 111(4), 512-523.
Rollins, J. C., & Stein, R. S. (2010). Coulomb stress interactions among M ≥ 5.9 earthquakes in the Gorda deformation zone and on the Mendocino Fault Zone, Cascadia subduction zone, and northern San Andreas Fault. Journal of Geophysical Research, 115(B12), B12306. https://doi.org/10.1029/2009JB007117
Sevilgen, V., Stein, R. S., & Pollitz, F. F. (2012). Stress imparted by the great 2004 Sumatra earthquake shut down transforms and activated rifts up to 400 km away in the Andaman Sea. Proceedings of the National Academy of Sciences, 109(38), 15152-15156.
Stein, R. S., Toda, S., Rollins, C., and Sevilgen, V. (2023). December 2022 California earthquake ruptured unknown fault: an analysis. Temblor. http://doi.org/10.32858/temblor.294
Yeck, W. L., Shelly, D. R., Materna, K. Z., Goldberg, D. E., & Earle, P. S. (2023). Dense geophysical observations reveal a triggered, concurrent multi-fault rupture at the Mendocino Triple Junction. Communications Earth & Environment, 4(1), 94.
Yoon, C. E., and D. R. Shelly (2024). Distinct Yet Adjacent Earthquake Sequences near the Mendocino Triple Junction: 20 December 2021 Mw 6.1 and 6.0 Petrolia, and 20 December 2022 Mw 6.4 Ferndale, The Seismic Record, 4(1), 81–92, doi: 10.1785/0320230053.
Copyright
Text © 2024 Temblor. CC BY-NC-ND 4.0
We publish our work — articles and maps made by Temblor — under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license.
For more information, please see our Republishing Guidelines or reach out to news@temblor.net with any questions.