Imagine an earthquake that releases as much energy as a magnitude 7 — and nobody feels a thing. No shaking, no rattling windows, no sirens. Just a few centimetres of ground quietly sliding, deep underground, over a period of weeks. That is a slow-slip earthquake, and it happens regularly beneath the Pacific Northwest, Japan, Mexico and New Zealand.
Slow-slip events — also called slow earthquakes, or episodic tremor and slip (ETS) when they come with a faint seismic hum — are one of the most important discoveries in modern seismology. They show that faults do not have just two settings, locked or breaking. There is a middle setting: creeping, trembling, releasing energy in slow motion.
A regular earthquake is fast. A fault that has been locked for decades or centuries suddenly slips — metres of movement in seconds, sometimes a minute or two for the very largest events. The sudden jolt is what shakes buildings and, on the seafloor, can launch a tsunami.
A slow-slip event does the same kind of thing, only stretched out. The same plate boundary slips by a few centimetres, but over days to weeks, and sometimes months. Because the energy is released so gradually, there is no sharp jolt for anyone to feel. EarthScope, the consortium that runs much of North America's geophysical monitoring, puts it plainly: during a Cascadia slow-slip episode, the plate boundary slides only a few centimetres to relieve stress in that part of the fault — and people nearby do not feel the tremors at all.
That energy comparison is why scientists take slow slip seriously. It is not a curiosity at the edge of seismology. It is a major part of how subduction zones spend their energy budget between big earthquakes.
Slow slip was invisible until instruments got good enough to see it. In the early 2000s, scientists working in Japan noticed a strange, noise-like signal repeating in seismic data — a faint, continuous rumbling they called tremor. At almost the same time, researchers in Cascadia spotted something equally odd in GPS data: stations that normally drifted steadily in one direction as the Juan de Fuca Plate pushed beneath North America would periodically reverse course and drift back the other way for days to weeks.
Put the two observations together and the picture clicked. The tremor and the GPS reversal were the same event, seen two ways: a stretch of the plate boundary, some 25 to 40 kilometres below the surface, was slipping slowly instead of staying locked. Because the slip and the tremor recurred together, researchers named the phenomenon episodic tremor and slip — ETS for short.
Today the Pacific Northwest Seismic Network keeps a continuously updated tremor catalogue for the whole Cascadia margin, and GPS stations across the region record every episode. Slow slip went from undetectable to routinely monitored in about two decades — a reminder of how much of earthquake science depends on simply watching the ground closely enough.
The best-studied slow-slip zone is Cascadia, the subduction zone running roughly 1,000 kilometres from northern Vancouver Island to Cape Mendocino in northern California, where the Juan de Fuca Plate slides beneath the North American Plate. But Cascadia is not unique. Slow-slip events have been found in Japan's Nankai Trough, in Mexico, in New Zealand's Hikurangi margin, and in other subduction zones around the world.
In Cascadia, the episodes do not happen everywhere at once, and they do not recur on a single neat schedule. The interval varies along the margin — roughly every 10 to 24 months depending on where you look, about every 14 months in the Seattle–Vancouver segment. That segmentation itself is a clue: different stretches of the same plate boundary behave differently, which matters for understanding where stress is building and where it is being released.
Here is the part that keeps seismologists up at night. A subduction zone has a locked section — the shallow part of the fault that is stuck and accumulating stress — and, deeper down, a section that slips freely. Slow slip happens in the transition zone between them.
When slow slip relieves stress in the transition zone, it can add stress to the locked section next door, nudging it closer to failure. Every slow-slip event, in other words, slightly changes the odds on the locked megathrust zone — the section that, in Cascadia, last ruptured in 1700 in a magnitude ~9 earthquake. That does not mean a slow-slip event causes the next megathrust quake. Most slow-slip episodes pass with no large earthquake following at all. But the stress transfer is real, and it is one more variable in a system scientists are still learning to model.
It also raises a question the public asks every time a slow-slip episode makes the news: is this a warning? The honest answer, from the scientists who monitor these events, is no — not a reliable one. Slow slip is not a prediction tool. Some large earthquakes have been preceded by slow slip, but most slow-slip events are not followed by a large earthquake, and treating every episode as a forecast would mean crying wolf on a regular schedule. What slow slip gives scientists is not a warning siren. It gives them a live reading of a fault's behaviour between earthquakes — and that, over years, sharpens hazard maps, building codes and preparedness planning far more than any single alarm could.
For most of history, the earthquake cycle looked simple: stress builds, the fault breaks, the ground shakes, repeat. Slow slip adds a quieter chapter to that story. Between the dramatic breaks, faults also sigh, creep and tremble — releasing energy in instalments too slow for human senses, but large enough to matter.
Next time you hear that a slow-slip event is under way beneath the Pacific Northwest, there is no need to panic — and no reason to ignore it either. Nobody will feel it. But deep below the coast, a fault is moving, instruments are watching, and our picture of how the next big earthquake builds is getting a little sharper.
Sources: EarthScope Consortium, "The Cascadia Slow Slip Phenomenon"; Pacific Northwest Seismic Network (PNSN) episodic tremor and slip summaries; Temblor, "Slow earthquakes rumble through the Pacific Northwest" (2022); Dragert et al. (2001) on Cascadia GPS-detected slow slip.