Mw4.7 earthquake strikes within Death Valley

We propose a potential new strike-slip fault beneath the valley floor

Earthquake Insights is an ad-free newsletter written by two earthquake scientists. Our posts are written for a general audience, with some advanced science thrown in! To get these posts delivered by email, become a free subscriber. If you would like to support our work here, please also consider a paid subscription.

People in eastern California and parts of Nevada felt a rattle around 1 AM on October 25, 2024. One person in Las Vegas described it as “very light rolling. Enough to set the ceiling fan off balance and make the walls creak.” The USGS PAGER estimates that more than 3 million people may have experienced weak shaking, although presumably most of them were asleep. Nearly three hundred people have reported feeling shaking so far.

The culprit? A M4.7 earthquake in Death Valley.

Figure 1: Shaking reported to the USGS for the October 25 M4.7 earthquake. Source: https://earthquake.usgs.gov/earthquakes/eventpage/ci40964128/dyfi/intensity

But actually, that’s not the only earthquake that was reported around that time: 24 seconds after the M4.7, another earthquake was reported: a M4.6, about 120 km to the south. This earthquake did not actually happen, and was eventually deleted from the catalog, about six hours later.

Figure 2: Map of earthquakes from 2024-10-24 12:00 UTC to 2024-10-25 14:00 UTC. Earthquakes are projected to the right onto a timeline. Red lines are faults from the Quaternary Fault and Fold Database of the United States. Colored lines are contours of shaking intensity estimated by the USGS for the M4.7 earthquake in Death Valley.

Phantom earthquakes reports can happen when earthquake waves caused by one earthquake are misidentified as a different earthquake in a different location. In this case, we are told by Dr. Allen Husker at Caltech that the system identified the S waves of the M4.7 as the P waves of a second earthquake, something that is known as a split event. These types of misreports are usually, but not always, caught by the system — and are generally limited to areas where the network is sparse, like in and around Death Valley.

During the six hours between when the second M4.6 appeared in the list and when it was removed, we spent some time looking at seismograms to try to figure out if it was real. You can find seismograms for the M4.7 here, and for the (now deleted) M4.6 here. If you look at stations close to the two earthquakes you will see that the second earthquake does not fit the waveforms.

Errors like this are rare, and are typically corrected within minutes to hours as scientists manually review the automated picks (as we saw with this event). In a time when people are desperately trying to automate everything in sight, it is comforting to know that experienced seismologists are still at the helm of earthquake detection.

You might have noticed something else in the map and timeline above: a swarm of seismicity prior to the M4.7 in Death Valley. Indeed, the earthquake was preceded by foreshocks starting with a M4.2 about an hour and thirteen minutes prior to the M4.7.

Those foreshocks seem to just be the latest iteration of a longer seismic swarm that started on October 9th with a M2.9 earthquake. The swarm seems to have been located just to the north of the latest foreshock-mainshock sequence.

There have also been some aftershocks since the M4.7, including a M4.3 a few kilometers to the southeast.

Figure 3: Seismicity around the M4.6 in Death Valley, since September 1 2024, showing that seismicity started here on October 9. Earthquakes are projected to the right onto a timeline. The M4.3 to the southeast occurred ~7 hours after the mainshock.

Death Valley faults and earthquakes

Today’s Mw4.7 earthquake is one of the largest earthquakes to have been instrumentally recorded in Death Valley. The very northern end of the valley experienced a M4.7 event in 1945, and a M4.7 event also occurred on the eastern side of the valley, at the same latitude as today’s event, in 1992. While our map doesn’t include the southern half of Death Valley, no M4.5+ earthquakes have ever been recorded there, either.

This lack of large recorded earthquakes might be surprising to those of us who have spent much time in the valley, often examining extremely well exposed, large, and apparently young fault surfaces!

Figure 4: Earthquake locations are from Southern California Earthquake Data Center (squares) and ANSS/ComCat (circles). Focal mechanisms include low-magnitude solutions for Southern California calculated by Cheng et al. (2023).

So, which fault ruptured during this event? That is surprisingly hard to answer. Because small to moderate sized earthquakes rarely cause surface ruptures or significant displacement of the surface, we have to rely on other, less direct indicators.

The focal mechanisms of the mainshock and foreshocks give us two options. The fault is either a northeast-southwest oriented, left-lateral strike-slip fault, or it is a northwest-southeast oriented, right-lateral strike-slip fault.

The foreshock and aftershock distribution isn’t conclusive. Sometimes the pattern of small earthquakes “lights up” a single fault. In this case, there are two trends. A NE-SW oriented trend is defined by the foreshocks, and a NW-SE trend is defined by many of the aftershocks.

Figure 5: The foreshocks and aftershocks of the M4.7 earthquake appear to define two trends: one toward the northeast, one toward the northwest.

Let’s take a look at the mapped faults in the surrounding area, and see if any of them are a good fit. These faults are known because they appear at the surface, usually as breaks cutting across young sediments and landforms. The map below shows fault traces from the USGS Quaternary Fault and Fold database. We have labeled some of the main faults.

Figure 6: Known active fault traces in the Death Valley region, from the USGS Quaternary Fault and Fold database.

Option 1: The Northern Death Valley Fault Zone

The biggest nearby strike-slip fault is the Northern Death Valley Fault Zone (NDVFZ), which cuts an almost straight line across the landscape along the eastern valley edge:

Figure 7: A perspective view of the NDVFZ in Google Earth. The fault cuts across alluvial fans along the eastern edge of Death Valley. To the right, older fans are being eroded; to the left, new fans are being deposited. The fault trace is marked by white triangles.

The NDVFZ has plenty of evidence for Quaternary activity. It is a strike-slip fault with the correct slip sense to match today’s earthquake, but has little to no recorded seismicity along its trace, as you can see on the map above.

Could the NDVFZ be responsible for the recent earthquakes, and past earthquakes along the same trend? For shallow earthquakes below the center of Death Valley to have actually occurred on the NDVFZ, there would have to be a ~7km systematic mis-location of both the earthquakes and the focal mechanisms. For regions with very sparse networks it is possible to get that kind of mislocation. Here, however, that would translate to about a 1-2 second difference in expected arrival time for P-waves at nearby stations. That seems pretty unlikely, given the density of seismic stations in the area, on all sides. The reported location uncertainty for the Mw4.7 earthquake is only ~0.3 km.

Figure 8: The distribution of Southern California Seismic Network stations (triangles) around Death Valley. In most of the world, this would be considered a dense network. In Southern California, this is considered a sparse part of the network.

However, it is important to remember that when faults aren’t vertical, earthquakes on them can appear to be offset in map view. Here, it is possible that the NDVFZ actually dips westward, so that earthquakes happening on its deeper parts would appear farther west on a map. Can we match up the earthquakes with the fault that way?

We drew some cross-sections across Death Valley, to try to see how the earthquakes in the valley center line up with a potentially west-dipping fault (the NDVFZ, or the Black Mountains Fault Zone farther south - on the D-D’ section).

Figure 9: Four cross-sections across Death Valley, showing the locations of recorded earthquakes. The tall black arrows show the location of major mapped fault traces. Red lines indicate the apparent dip (in cross section) of selected focal mechanisms. A planar fault would intersect the surface where the red lines would intersect the surface.

The black inverted triangles show where the surface faulting is seen at the eastern valley edge. We also drew in red lines on the profiles to show the dip of the steeper nodal plane (fault surface) for the focal mechanisms, as cut by the profile. Do the nodal planes align with the surface trace? No: it’s pretty clear that the aligned seismicity and the focal mechanisms both disagree with a west-dipping fault that crops out at the black triangle. Instead, it looks like a more vertical fault exists, farther west.

To conclude, we don’t think the NDVFZ is a viable host fault for this earthquake, or any of the recorded seismicity, unless something is extremely wrong with earthquake locations in Death Valley.

Option 2: The Towne Pass Fault

What about a more east-west oriented fault? Since earthquake focal mechanisms are ambiguous — presenting two possible fault planes — this is within the realm of possibility, on the basis of the seismicity.

The northwest side of the Panamint Range has a spectacular fault system that is exposed at the surface, which also lines up with the focal mechanism of the Mw4.7 earthquake. The contact between the alluvial fans (material eroded off the Panamint Mountains and dumped at the rangefront) and the basement rock is clearly faulted. In the image below, the location of the earthquake is approximately shown with a red star.

Figure 10: A perspective view in Google Earth of the northern end of the Panamint Range, showing the surface expression of the Towne Pass Fault. Mosaic Canyon is the big canyon at top center. The location of the October 25 earthquake is schematically shown as a red star.

This is the Towne Pass Fault. Many thousands of tourists, and numerous geology students, have crossed this fault on their way to Mosaic Canyon, located near the top of the image above. Note: If you look at the image, you can see that the darker alluvial fan in the foreground is cut by the fault, but the lighter-colored fan in the background is not. Because some rocks in the desert will acquire a dark “varnish” slowly over time, the darker fan can be inferred to be older.

However, there are some serious problems with this fault being the host fault. First, the Towne Pass Fault is reported to have normal-sense offset, with no known strike-slip component. Second, it is not known to be active based on surface geology (edit: Comments from Tim Dawson and Nick van Buer clarify that despite not having definitive evidence of recent activity, all signs are there that the Towne Pass Fault is probably active enough to host future large earthquake - and we agree). It is possible that this fault, or a similar structure, hosted the small-magnitude foreshock swarm.

Option 3: An unmapped, hidden strike-slip fault below the valley floor

The recorded seismicity in Death Valley shows a northwest-to-southeast alignment of earthquakes beneath the relatively flat valley floor. When focal mechanisms can be determined, they are typically strike-slip events on steep faults, with one nodal plane that aligns with this larger trend. This suggests to us that there is actually a big right-lateral strike-slip fault hidden below the salt pans and sand dunes, which is the most seismically active structure at present.

However, a strike-slip fault of this location and orientation isn’t present in the USGS Quaternary Fault and Fold database, or on the Southern California Earthquake Data Center fault map.

Why would a fault like this exist beneath the floor of Death Valley, where we can’t see it? There is a good geological reason.

The huge hole in the ground that is Death Valley was formed by extreme crustal extension, but this extension was fundamentally linked to strike-slip faulting.

Death Valley is the canonical pull-apart basin, first described by Burchfiel and Stewart (1966):

Figure 11: The proposal of Burchfiel and Stewart (1966) that Death Valley formed as a pull-apart.

This type of structure forms when two strike-slip faults come close together, but don’t actually line up. The area between the faults is forced to stretch out over time, forming a new sedimentary basin trapped between active normal faults (if the faults stepped over the other way, there would instead be a zone of compression).

However, Death Valley does not have any recorded normal-type earthquakes on its boundary faults, and it’s not clear how much extension is currently happening. Indeed, we see mostly strike-slip earthquakes within the valley itself.

As the total displacement across a pull-apart basin increases, it gets harder and harder to keep stretching the crust in between. Eventually, a strike-slip fault can develop that cuts across the valley, killing off the normal faults as it takes up the slip. This is called a cross-basin fault zone. These faults are often deeply buried by sediment, and they can be hard to discover.

The images below show how cross-basin faulting forms in an analog model - a glorified sandbox (Rahe et al., 1998):

Figure 12: Analog model showing how cross-faults can develop within a pull-apart basin. Source: Figure 7 of Rahe et al. (1998).

So, this earthquake has us wondering: is there a buried strike-slip fault beneath Death Valley? It is notable that the figure above comes from a paper that discusses Death Valley specifically, commenting: “There is no strong evidence of cross-basin faulting in Death Valley,” and “the estimated 3 km of alluvial fill in Death Valley is probably not enough to effectively conceal a large cross-basin fault at the surface.”

Do the earthquakes recorded since publication of that paper in 1998 suggest otherwise? Maybe it’s a bit audacious for us to propose the existence of an entirely new fault on the basis of the quick evaluation in this post, but that’s what it looks like to us. This would indicate that Death Valley, like a Pokémon, has indeed reached the next step of its evolution.

Figure 13: Our proposed strike-slip fault beneath the flat floor of Death Valley. Is it - or isn’t it?

We welcome any input from the many people who have more knowledge of this iconic area, or who may be aware of prior proposals for a similar structure that we didn’t find in our literature search! And many thanks to the seismologists who helped us understand the origin of the phantom-quake!

References:

Burchfiel, B.C. and Stewart, J.H., 1966. “Pull-apart” origin of the central segment of Death Valley, California. Geological Society of America Bulletin, 77(4), pp.439-442. https://doi.org/10.1130/0016-7606(1966)77[439:POOTCS]2.0.CO;2

Cheng, Y., Hauksson, E. and Ben‐Zion, Y., 2023. Refined earthquake focal mechanism catalog for southern California derived with deep learning algorithms. Journal of Geophysical Research: Solid Earth, 128(2), p.e2022JB025975. https://doi.org/10.1029/2022JB025975

Rahe, B., Ferrill, D.A. and Morris, A.P., 1998. Physical analog modeling of pull-apart basin evolution. Tectonophysics, 285(1-2), pp.21-40. https://doi.org/10.1016/S0040-1951(97)00193-5

U.S. Geological Survey, 2020, Quaternary Fault and Fold Database for the Nation, accessed [10/25/2024], at https://doi.org/10.5066/P9BCVRCK