After a big earthquake, why do aftershocks go on for days?
― The rock that didn't break is slowly finding its balance again
Even after the big shaking stops, small tremors can keep coming for days, sometimes months. This isn't because the ground is "still on the verge of breaking." It's the process of rock that couldn't break all at once gradually letting go of the force pressing on it. And the way that rate falls off follows a surprisingly clear pattern.
You slowly bend a hard bar of chocolate with both hands. At some point, it snaps with a loud crack.
But that's not the end. After it breaks, small cracking sounds can keep coming from your hands. That's because parts around the main break didn't fully snap.
At first the sounds come one after another, then the gaps between them grow longer. What happens underground is a lot like this.
Only two reasons aftershocks keep coming
The spot that slipped in the mainshock releases its force, but that force gets pushed onto the ends that didn't slip. The surrounding rock ends up under more pressure than before.
When a pressured spot finally breaks depends on how sturdy it is. Brittle spots go quickly; tougher ones hold out for days. That's why the shaking drags on.
The mainshock isn't an "ending" — for the surrounding rock, it's a "beginning." Let's look at this in order.
The mainshock doesn't erase force
Underground rock is slowly bent by the movement of tectonic plates over decades and centuries. When the built-up strain passes its limit, the rock slips along a surface called a fault. That's an earthquake.
At that moment, the force that had built up along the slipped surface is released all at once. But the force itself doesn't vanish. A kind of step forms at the boundary between the part that slipped and the part that hasn't. That boundary ends up under stronger force than before.
Think of the chocolate bar again: it's the tip of the crack. Force concentrates most at the leading edge of a crack. It's the same with earthquakes — force gets pushed onto the outer edge of the area that slipped in the mainshock, and along its extension. Look at the left side of Figure 1.
Why doesn't it just break all at once?
If force is being pushed onto a spot, you might expect it to break all at once, right there. But that's not how rock breaks.
Inside rock, sturdy patches and brittle patches are mixed together unevenly. Even under the same force, the time it takes to break varies widely. Brittle spots break within seconds; tougher spots take days; the toughest spots can take months. This spread in "waiting time before breaking" is the real reason aftershocks drag on so long.
There's another mechanism that buys time. Water seeps into the gaps in cracked rock, and its pressure affects how the rock moves. When the mainshock loosens the rock, water gradually migrates from high-pressure areas to low-pressure ones. Wherever the water reaches, the rock becomes easier to break — this is thought to be another source of delay.
And each aftershock triggers its own new shift of force. Aftershocks beget aftershocks. So the rate doesn't stop abruptly — it trails off in a long tail.
An aftershock refers to shaking smaller than the mainshock. But if a later tremor turns out to be bigger, the earlier one gets relabeled a "foreshock." In other words, no one can know at the time whether the first tremor was the mainshock. This is said to be why the Japan Meteorological Agency has moved away from the term "aftershock" and now speaks of "seismic activity" instead.
Empirically, the largest aftershock is often about one magnitude smaller than the mainshock. But this is only an average tendency — there are known cases where shaking on a similar scale to the mainshock has occurred again afterward. This isn't a number to rely on for thinking "nothing bigger is coming."
So what should you actually do?
- When it shakes, protect your head firstGet under a table, or cover your head with a bag or cushion, and stay low. Move only after the shaking stops.
- After a big quake, expect more shakingStay away from furniture that might topple and places with broken glass. Choose spots where nothing can fall on you.
- Stay away from cracked buildings and wallsA building weakened by the first shock can collapse in the next one. Avoid walking past it even if you don't need to.
After a big quake, act on the assumption that shaking of similar size could happen again. If a building is cracked, leaning, or its doors won't close, don't force yourself to stay inside — move to a safe place such as a shelter. If you're near the coast and feel shaking, head for higher ground immediately. If someone is injured or there's a fire, call 119. Before rescuing anyone trapped under fallen furniture, make sure the surroundings are safe, and if it feels dangerous, wait for the fire department or police. Follow evacuation and access guidance from your local municipality and the Japan Meteorological Agency.
Summary
The mainshock doesn't erase built-up force — it pushes it onto the parts that didn't break. Those pressured spots then break, one after another, over widely varying timescales depending on how sturdy they are. That's why shaking drags on, and why its decline follows a regular pattern.
Aftershocks aren't the sound of the ground still breaking.
They're the sound of unslipped rock finding its balance again, one piece at a time.
The difference between the sharp "rattle" and the rolling "sway" you feel in an earthquake is explained in "Why does the earthquake's sharp "rattle" arrive before the rolling "sway"?" How the ground turns soft after shaking is covered in "Why does liquefaction turn solid ground into something like water?", and the ongoing movement of plates underground is covered in "How did seashell fossils end up on top of a mountain?"
- Slowly bend a dry bar of chocolate with both hands. Keep applying force until it snaps with a loud crack.
- After it breaks, keep holding it without relaxing your grip. You may notice small sounds continuing here and there.
- Count each sound with your finger as it happens. You'll notice they come rapidly at first, then space out more and more over time.
The same thing happens with cookies or biscuits. Real ground is far more complex than this, but the pattern — a big break followed by a trailing series of small ones — is the same.
Want to know more? ― Terms, formulas, and how this connects to textbooksLabels show whether each part is middle-school-level or university-specialist-level
- MSCovered in middle-school science
- HSCovered in high-school Earth Science Basics
- HS+High-school enrichment, or textbook sidebar content
- Univ.Not taught in high school — university-level specialist content (seismology, rock mechanics)
- ResearchNot yet settled even at university level — an active research question
MSTerms: this phenomenon has names
- Mainshock: the largest tremor in a sequence of seismic activity. It's only determined after the fact.
- Foreshock: a tremor smaller than the mainshock that occurred before it. Telling it apart while it's happening is said to be difficult.
- Stress: the strength of the forces pushing and pulling on a surface inside rock. Expressed as force per unit area.
- Ōmori's law: the empirical rule that the number of aftershocks decreases in inverse proportion to time since the mainshock. Shown by Fusakichi Ōmori in 1894.
MSHSWorking it out with a formula: how fast do aftershocks decline?
The simplest form of Ōmori's law says: "the number per day is inversely proportional to the number of days since the mainshock." Here we'll consider a case with 100 aftershocks on the first day. All you need is division and subtraction.
| In symbols | n = K ÷ t (the modified form is n = K ÷ ( t + c )^p) |
| In words | Aftershocks per day = a constant fixed for that earthquake ÷ days elapsed since the mainshock |
| Where it comes from | In 1894, Fusakichi Ōmori found this empirical rule (Ōmori's law) from records of aftershocks following the Nōbi earthquake. Tokuji Utsu later refined it into the Modified Ōmori Law, adding an exponent p and a small time offset c |
| Aftershocks on day 1 | 100 (an assumed value for this example) |
| Rule for the decline | the count is taken to be inversely proportional to the number of days |
| Counting interval | every 1 day |
| Symbol | Meaning and unit |
| Count | number of aftershocks on that day [count] |
| Days | days elapsed since the mainshock [days] |
| Inverse proportion | relationship where doubling the days halves the count [no unit] |
| Approx. count on day 2 | 100 ÷ 2 = 50 |
| Approx. count on day 10 | 100 ÷ 10 = 10 |
| Approx. count on day 100 | 100 ÷ 100 = 1 |
| Drop from day 1 to day 10 | 100 − 10 = 90 |
Most of the overall activity is used up in the first day, and by day 10 the daily count is down to a tenth. But dropping by another tenth from there takes another 90 days. It falls sharply, yet never quite reaches zero — that's what "trailing off for a long time" means. The line on the right of Figure 1 traces exactly this shape.
HSHS+How does the force get shifted?
HSWhen a fault slips, the force along that surface is released, but at the edges of the slipped area, the rock deforms as if being torn. The greater the deformation, the stronger the concentration of force. This concentration is said to explain why aftershocks are common just beyond the edges of the mainshock's slipped area.
HS+In seismology, this change in force is expressed as a quantity called the Coulomb stress change: the force driving slip on a given surface, minus the effect of the force clamping that surface shut. Many earthquakes have shown increased aftershock activity where this quantity rises, and quieter seismic activity where it falls.
Univ."Delayed breaking" from a rock-mechanics perspective
Rock doesn't necessarily break the instant it reaches its breaking limit. Even under a force slightly below that limit, it can still break given enough time. This is called subcritical crack growth: it happens because the rate at which bonds snap one by one at a crack tip depends strongly on force, temperature, and the presence of water. Because that rate changes sharply with force, even a small difference in force can change the time to failure by orders of magnitude. Averaged across an entire rock mass, this property has been shown to produce a decline close to Ōmori's law, with aftershock counts falling off inversely with time. Rate-and-state friction laws, which describe fault friction in terms of slip rate and contact history, yield a similar pattern of decline.
📖 For the derivation of the formula and further reading: Ōmori's law (Japanese Wikipedia) ／ Aftershock (Japanese Wikipedia)
ResearchWhat's still not fully understood
- Telling foreshocks from aftershocks. There's still no established way to judge, while it's happening, whether a tremor is a sign of a bigger one to come.
- What mainly causes the delay. The properties of the rock itself, the movement of water through gaps, and chains of aftershocks triggering aftershocks — how much each contributes is still debated, and varies by region.
- How long it lasts. There are cases thought to show aftershock activity continuing for decades after a big earthquake, and where to draw the line on what counts as "one sequence" is unsettled.
- Shaking triggered far away. Cases have been reported where activity picks up hundreds of kilometers away, triggered by the passing of seismic waves, and how this happens is still being studied.
In other words, even this article only explains things as far as they're currently understood. That's exactly why it makes sense to move to a safe place soon after a big quake.
How this connects to textbooks, by level
| Level | Subject/Unit | Where in this article |
|---|---|---|
| MS | Science: Changes in the Earth / Direct and inverse proportion | Faults and slip, calculating the decline in count |
| HS | Earth Science Basics (earthquakes and faults, plates) | The flow of force building up, releasing, and shifting |
| HS+ | Enrichment/sidebar (change in stress) | Coulomb stress change and aftershock distribution |
| Univ. | Seismology / rock mechanics (subcritical crack growth, friction laws) | Mechanism of delayed breaking, derivation of Ōmori's law |
| Research | Statistical seismology / earthquake forecasting (ongoing) | Distinguishing foreshocks, triggered activity |
| ― | Connection to daily life | How to spend time after a big quake, judging when to avoid damaged buildings |
- Japan Meteorological Agency, "About Earthquakes" and explanatory materials on seismic activity following major earthquakes.
- Tokuji Utsu, "Seismology" (Kyoritsu Shuppan / 共立出版), the chapter covering aftershock statistics and Ōmori's law.
- High-school "Earth Science Basics" (地学基礎) textbook, the section on earthquakes and faults.
- Scholz, C. H., "The Mechanics of Earthquakes and Faulting", Cambridge University Press. (A standard textbook covering fault friction and aftershocks)
※This article is a general-audience science explainer. Any figures given are approximations meant to help explain the underlying mechanism. For actual evacuation and building-safety decisions, follow the instructions of your municipality, fire department, local authorities, and the Japan Meteorological Agency.