An earthquake swarm is a sequence of mostly small earthquakes clustered in one area with no single, dominant mainshock. Unlike a typical earthquake sequence where one large quake strikes first and smaller aftershocks follow, a swarm produces many events of similar size over days, weeks, or sometimes months. Swarms are most commonly linked to underground fluid movement and geothermal activity.
How Swarms Differ From Aftershock Sequences
In a standard earthquake sequence, the pattern is predictable: a large mainshock ruptures a fault, then aftershocks follow in decreasing frequency and magnitude. The aftershocks are part of the crust readjusting after that main rupture, and they taper off over time in a well-understood decay pattern that can last weeks to decades.
Swarms break this pattern. Instead of one big event followed by declining activity, a swarm produces a cluster of earthquakes that stay relatively steady or even increase in rate before eventually stopping. No single event stands out as “the big one.” The U.S. Geological Survey applies the swarm label when many earthquakes occur in a small area and simply don’t fit the mainshock-aftershock template. Most individual quakes in a swarm are too small to feel, though some swarms produce events large enough to be noticed at the surface.
What Causes Earthquake Swarms
Most swarms are driven by processes happening deep in the crust that don’t involve the sudden, large-scale fault rupture of a typical earthquake. The most common trigger is underground fluid movement. When water, gas, or other fluids migrate through rock, they increase the pressure inside tiny pore spaces along faults. This pressure buildup weakens faults that are already under stress, causing them to slip in many small events rather than one large break.
In volcanic and geothermal areas, the picture gets more complex. Magma pushing through rock, pockets of volcanic gas rising toward the surface, and hot water circulating through fracture networks can all generate swarms. Research at Yellowstone has shown that the interplay between slowly diffusing hydrothermal fluids and sudden bursts of fluid injection controls when swarms happen. Fluids may diffuse slowly over years through rock, explaining quiet pauses between swarm episodes. Then, when a permeability seal breaks (essentially a barrier in the rock that had been blocking fluid flow), pressure can spike rapidly, triggering a burst of earthquakes lasting a few weeks.
Tectonic forces also play a role. In some settings, slow, steady creep along a fault generates swarms without any fluid involvement. And in extensional zones where the crust is being pulled apart, swarms can accompany the formation of new fractures as rock stretches and breaks along multiple small faults rather than one large one.
Volcanic Swarms as Warning Signals
When magma forces its way toward the surface as a vertical sheet called a dike, it cracks rock as it moves. This produces a distinctive swarm pattern: the earthquakes migrate through space over time, tracing the path of the advancing magma. During a recent volcanic crisis near Santorini, Greece, scientists tracked thousands of earthquakes that migrated along a clear path, accompanied by measurable ground deformation detected by GPS and satellite radar. Tremor signals containing bursts of tiny quakes coincided with the most intense periods of underground rupture.
These migrating swarms look fundamentally different from stationary clusters. The earthquakes shift location in a way that maps the underground plumbing. Scientists use this migration pattern, combined with surface deformation measurements, to distinguish magma intrusion from ordinary tectonic activity. Similar patterns have been observed during intrusions at volcanoes in Iceland, Hawaii, and Japan.
Human Activities That Trigger Swarms
Not all swarms are natural. Injecting large volumes of wastewater underground, a common practice in oil and gas operations, has caused a significant increase in swarm-like seismicity in parts of the United States. The mechanism is the same as natural fluid-driven swarms: injected fluid raises pore pressure along pre-existing faults, pushing them past their breaking point.
Enhanced geothermal energy projects have run into similar problems. At a geothermal site in Pohang, South Korea, hydraulic stimulation generated a swarm that escalated to a magnitude 5.5 earthquake in 2017, two months after pumping had stopped. A geothermal project in Basel, Switzerland, was permanently shut down after stimulation triggered a magnitude 3.4 event in 2006. These cases illustrate a troubling pattern with injection-induced swarms: pore pressure increase and earthquake interactions can weaken faults progressively, sometimes triggering larger events later in the sequence rather than earlier.
How Long Swarms Typically Last
Most swarms are short-lived, fading within days to weeks. Some persist for months, and in geothermally active areas, swarms often recur in the same locations over years or decades. At Yellowstone, the three largest modern swarms occurred in 1985, 2008 to 2009, and 2010, all within or near the caldera. A fourth major sequence near Maple Creek in 2017 to 2018 produced over 3,000 recorded earthquakes. That particular sequence turned out to be partly long-lived aftershocks from the magnitude 7.2 Hebgen Lake earthquake of 1959, mixed with some influence from magmatic fluids, showing that the line between swarms and aftershock sequences isn’t always clean.
Do Swarms Lead to Bigger Earthquakes?
This is the question most people really want answered, and the data is reassuring. Worldwide, the probability that any given earthquake will be followed within three days by a larger earthquake nearby is just over 6%. In California, that number is about the same. This means roughly 94% of earthquakes, including those in swarms, are not foreshocks to something bigger.
That said, about half of California’s largest historical earthquakes were preceded by foreshocks, while the other half struck without warning. The challenge is that there’s no reliable way to tell in real time whether a swarm is building toward a larger event or simply releasing energy gradually. Scientists treat swarms as situations worth monitoring closely, but a swarm in progress is not, by itself, a prediction of a major earthquake.
How Scientists Monitor Swarms
Networks of seismometers continuously record ground motion in seismically active regions. When a swarm begins, the sheer volume of small, overlapping earthquakes creates a processing challenge. Events can be so close together in time that their signals overlap on the seismograph, making it hard to measure each one individually.
Machine learning is increasingly helping solve this problem. At Yellowstone, scientists have trained AI models to estimate earthquake magnitudes from short windows of seismometer data, which works even when events are stacked on top of each other. Each seismic station gets its own trained model, calibrated to local conditions. Deeper learning models that process raw ground motion data are also being used to refine earthquake locations, giving scientists a more precise three-dimensional picture of where swarm activity is concentrated underground. That spatial picture is critical: a swarm that stays in one place suggests fluid diffusion, while one that migrates may indicate magma on the move.