When the ground along the Salton Sea began shaking on June 5 and didn’t stop, scientists thought it was another swarm. But was it?
By Debi Kilb, Ph.D., Scripps Institution of Oceanography, University of California, San Diego (@Kinect_with_Sci), Wenyuan Fan, Ph.D., Scripps Institution of Oceanography, University of California, San Diego, and Gabi Laske, Ph.D., Scripps Institution of Oceanography, University of California, San Diego
Citation: Kilb, D., Fan, W., Laske, G., 2021, Semantics at the Salton Sea: When does a swarm become a sequence?, Temblor, http://doi.org/10.32858/temblor.181
On June 5 at 1 a.m. local time, a series of small to medium earthquakes started shaking the southeast shore of the Salton Sea in Southern California. The largest event so far has been a magnitude-5.3 strike-slip earthquake about 11 hours after the series started. Over the last few days, more than 1,000 quakes have occurred in this area.
At first glance, this seemed like a fairly standard earthquake swarm — a common occurrence for the Salton Sea region. Upon a deeper dive into the earthquake data, however, the latest series of quakes may actually be a mainshock/aftershock sequence. What that means for the residents of this shaken region doesn’t change: The chance of an earthquake exceeding magnitude-7 within the following month remains less than 1% (see the U.S. Geological Survey’s Aftershock Forecast). For researchers, though, these events offer a new perspective on the source physics of these earthquakes.
Earthquakes vs. swarms
If you look at a map of the world’s earthquakes, they are not random like paint splatters in a Jackson Pollock painting. Instead, quakes tend to occur in patterns along the edges of tectonic plates. When you zoom in to take a closer look, you start to see somewhat fuzzy earthquake patterns, where many events participate in space/time clusters as part of earthquake swarms or mainshock/aftershock sequences. An earthquake swarm is a series of earthquakes that do not have an obvious single mainshock, and later quakes in the sequence may migrate geographically over time. A mainshock/aftershock sequence includes a large earthquake followed by many smaller aftershocks.
We think of mainshock/aftershock sequences as following a well-defined causal relationship: The largest aftershock is typically about one magnitude unit smaller than the mainshock (Kilb & Vernon, 2020), and it is possible to forecast how long the aftershock sequence might last and what size aftershocks to expect (Michael et al., 2020). In typical aftershock sequences, the magnitude and the number of events follow an exponential decline — Omori’s law. Swarms, on the other hand, are not well behaved. Sometimes they contain many mid-size quakes; sometimes swarms are composed of only smaller quakes. Swarm durations are variable and can span days, weeks or even years (Ross et al., 2020). This complexity makes earthquake swarms hard to forecast but also fascinating to study.
The mechanisms that cause swarm earthquakes are thought to differ from those causing tectonic earthquakes. The latter are triggered by changes in the forces, or stresses, that ultimately cause the movement of tectonic plates. One of the main driving mechanisms of swarms is the migration of fluids belowground. As fluids diffuse through fault systems, they can trigger quakes that are similar in size. Swarms also have been associated with hydraulic fracturing (i.e., fracking) and often occur in geothermal regions (Hill et al., 1975; Ellsworth, 2013).
Salton Sea swarms
Swarms are a common occurrence near the Salton Sea in Southern California, where they have been observed for more than two decades (Figure 1).
Some Salton Sea swarms are more vigorous than others in terms of their temporal evolution, spatial footprint and earthquake sizes. For example, the rates of seismic activity in the 2000 and 2016 swarms were rather wimpy, but the current swarm (if it is that) is notable for its high rate of seismic activity (see Table 1 and Figure 2). There is not a one-size-fits-all aspect to swarms even when limited to a single location.
The southern Salton Sea is located in a complex tectonic area. The San Andreas Fault to the east marks the transform boundary along which the Pacific Plate slides northward past the North American Plate. To the south, plate divergence causes the Gulf of California to widen. The geometry of the tugging stresses is complicated, leading to many long and short faults, some parallel to the San Andreas, others perpendicular to it. An earthquake on one fault may then increase the stresses on another fault, triggering an earthquake on the neighboring fault.
There are geothermal production sites within the Salton Sea region, which can influence how fluids are transported within the fault system, often dictating fluid migration. Given this, many researchers have hypothesized that fluids that migrate through the ground are the primary driving mechanisms for swarms in the southern Salton Sea region, although these have not caused large earthquakes here so far. It is possible, however, that a combination of fluid migration and plate tectonic movement drives seismicity, in which case a group of events can look like both a swarm and a mainshock/aftershock sequence.
We propose that the 2021 swarm (as well as a previous one in 2012) is not a swarm at all and instead is more akin to a mainshock/aftershock sequence (referred to as “sequence” from now on). Our first clue that this might be true is that the decay rates of the earthquakes look more like a sequence, differing substantially from the decay behavior of a typical swarm (Figure 3).
The spatial patterns of these datasets show different generalized spatial footprints for different years. The 2000, 2005, 2016 and 2020 swarms mapped swaths of seismicity along a single fault that trends northeast-to-southwest, perpendicular to the San Andreas Fault. The 2012 and 2021 data differ, instead delineating two faults of different orientations (Figure 4). This is intriguing and begs for an explanation. These observations are consistent with our conjecture that the 2012 and 2021 swarms were actually sequences, not swarms, or at minimum some kind of hybrid swarm/sequence.
Collectively, the two decades of Salton Sea earthquake data offer only a blurred view of the seismicity patterns, too fuzzy to explore intricacies within the fault system. Additional research is also needed to better differentiate a swarm from a sequence and to understand how a combination of mechanisms makes a series of events adopt some features of both a swarm and a sequence. The necessary next step is to refine the location of all these events. This step often leads from an initially fuzzy cluster of events to multiple delineated faults of various orientations and depths.
In other parts of Southern California, it has been proposed that fault interconnectivity can explain spatial segmentations among swarms (Ross et al., 2020). For example, some fault patches may experience more total slip than others and therefore be more damaged and permeable, which allows fluids to migrate more easily within these damaged zones. It will be interesting to eventually unveil the interconnectivity of Salton Sea faults and how they interact with each other over different spatial and temporal scales.
What does this mean?
Do these Salton Sea earthquake swarms/sequences indicate that a large San Andreas earthquake rupture is soon to occur? To answer this question, scientists use past observations to forecast what may come in the future. Here, we have presented results for six swarms or sequences near the Salton Sea that occurred over approximately 20 years. These data show no evidence that swarms or sequences trigger large earthquakes on the San Andreas Fault. But in all honesty, we have too few observations to make any long-range future claims, just as a survey of six individuals on a topic does not properly represent what the population might think.
The take-home message here is that we live in earthquake country. We need to be prepared and have a protection and response plan in place. Building codes protect our structures, and regular quake drills sharpen our awareness. Be prepared, not scared.
Ellsworth, W. L. (2013). Injection-induced earthquakes. Science, 341.
Hill, D. P., Mowinckel, P. & Peake, L. G. (1975). Earthquakes, active faults, and geothermal areas in the Imperial Valley, California. Science, 188, 1306-1308.
Kilb, D. and Vernon, F. (2020), Southern California jolted by moderate but intense quake, Temblor, http://doi.org/10.32858/temblor.084.
Michael, A. J., McBride, S. K., Hardebeck, J. L., Barall, M., Martinez, E., Page, M. T., van der Elst, N., Field, E.H., Milner, K.R. & Wein, A. M. (2020). Statistical seismology and communication of the USGS operational aftershock forecasts for the 30 November 2018 M w 7.1 Anchorage, Alaska, Earthquake. Seismological Research Letters, 91, 153-173.
Ross, Z. E., Cochran, E. S., Trugman, D. T. & Smith, J. D. (2020). 3D fault architecture controls the dynamism of earthquake swarms. Science, 368, 1357-1361.