Why Does an Earthquake Shake Twice, a Small Jolt First and a Bigger Shake Later?

1. An Earthquake Sends Out Two Kinds of Waves, and One Is Faster

When rock slips underground, the energy spreads out as waves. Seismologists (scientists who study earthquakes) give the two main ones plain names. The first to arrive is the P wave, for "primary." The one that follows is the S wave, for "secondary" [2].

The U.S. Geological Survey (USGS) gives rough speeds: P waves travel about 3.7 miles per second (roughly 6 km per second), and S waves about 2.5 miles per second (roughly 4 km per second). That makes the P wave about one and a half times as fast [2]. A USGS page on the 1906 San Francisco earthquake says the same thing another way: S waves generally travel about 60% as fast as P waves, and the S wave always arrives after the P wave [3].

The two waves also shake the ground differently. In a P wave the rock vibrates back and forth along the direction the wave is traveling. In an S wave the rock moves side to side, across that direction [3]. A simple way to picture it: a P wave is like pushing and pulling one end of a spring, and an S wave is like flicking a rope up and down and watching a bump run along it. That comparison is only a picture, not something the sources say.

The USGS describes P waves as generally not producing strong shaking, while the slower waves that follow are generally the more damaging ones [2]. So the first small jolt is usually the fast wave arriving first, and the stronger shaking that follows often comes from the slower waves. These speeds are averages. Waves move at different speeds in different layers of rock, so real numbers vary from place to place.

2. Why the Wait Between the Jolt and the Shake Gets Longer With Distance

Think of two runners starting together, one faster than the other. After 100 meters the faster one is only a little ahead. After a kilometer the lead is much bigger, because the speed difference has had longer to add up. Seismic waves do the same thing.

The USGS publishes travel-time tables for a quake 33 km deep, using a standard model of Earth. Distances in the table are measured in degrees around the globe, where one degree is about 111 km. Here are a few rows [1]:

The kilometer values are my conversion from degrees. The pattern is the point: the farther you are, the longer the wait between P and S. Geologists call this wait the S-minus-P time, and the USGS notes that it grows steadily with distance, which is why it can be used to estimate how far away an earthquake happened [1].

3. How Scientists Turn the Wait Into a Distance

A seismograph records the ground's motion over time. On the recording, a quiet line suddenly starts to wiggle slightly (P wave), and a little later the wiggles grow bigger (S wave). Measure the time between those two moments and you have the S-minus-P time. The USGS says its models identify the P and S waves in a recording and use the time between their arrivals to estimate the distance from the station to the epicenter, the spot on the surface above where the quake began [1].

A rough rule of thumb comes from the table above. Dividing distance by S-minus-P time gives about 8 km per second near the quake (111 km ÷ 13.5 s) and about 10 km per second at 1,110 km (1,110 km ÷ 111.9 s). I calculated these myself from the table. So, as a very rough guide, each second of gap means somewhere around 8 to 10 km of distance. Real earthquakes can differ because the speeds depend on the rock and on how deep the quake was.

One station gives the distance to the earthquake's epicenter but not its direction: the epicenter could be anywhere on a circle around the station. With distances from three or more stations, scientists can find where the circles meet and estimate the epicenter. This is the same idea as finding where you are from three radio towers.

4. How the Missing S Wave Showed That Earth's Outer Core Is Liquid

P waves can travel through solid rock and through liquids. S waves cannot cross a liquid: a liquid does not hold the sideways push and pull that an S wave needs [4]. That difference let scientists learn about places no one can dig to.

In 1906 the British geologist Richard Dixon Oldham plotted P and S travel times from many earthquakes recorded at distances between about 20° and nearly 160°. He noticed that S waves were only seen out to about 120°; beyond that, any later arrivals came 10 or more minutes late. He proposed a central core that behaved like a liquid. Berkeley's course notes add a caution: his identifications of the waves and his arithmetic were approximate [4].

Later work sharpened the picture. Beyond about 105° from an earthquake, there is an S-wave shadow zone: seismographs do not detect S waves there, and P waves are weak too. The notes say the core boundary lies about 2,900 km down (found by Beno Gutenberg in 1914), and that scientists are now very confident the outer core is liquid. Inge Lehmann found in 1936 that there is a solid inner core inside it [4].

So the S wave's absence is evidence in itself. The missing S waves helped scientists infer that Earth's outer core is liquid.

5. How Earthquake Alerts Use the Gap, and Why Some Places Get No Warning

An earthquake early warning system turns the same gap into a head start. Sensors close to the quake pick up the first, gentler P waves. A computer estimates where the quake was and how big it was, and sends an alert. Alerts travel as electronic signals, which are far faster than the waves, so they can arrive before the slower, stronger S waves do. The USGS says this works because the data processing and communication systems are faster than the seismic waves that cause the shaking [2]. An alert is not a prediction. It reports a quake that has already started.

The same page notes that people farther from the epicenter have more time. It also means the time is short even at best: the USGS describes alerts coming seconds to a few tens of seconds before the damaging waves arrive [2].

There is a limit too. For every quake there is a zone close to the epicenter where people feel the shaking before an alert reaches them, because the system needs time to collect data, calculate, and deliver the message [2]. Right on top of a quake the P and S waves arrive almost together, and there is no gap left to use.

6. Try It: Count the Gap Yourself, and Where to Read More

You do not need an earthquake to practice the idea. In the next thunderstorm, count the seconds between a flash and the thunder. You see lightning before you hear thunder because light travels much faster than sound, and the longer the wait, the farther away the lightning was. (Stay indoors while you do it.) It works for the same reason the P and S waves do.

If you want the real thing, the USGS publishes recordings and travel-time tables, and you can try the S-minus-P calculation with the three rows above. To dig deeper, the USGS earthquake early warning overview [2] and the Berkeley notes on Earth's interior [4] are good places to start. If you ever feel shaking yourself, the best response is the one your local emergency authority recommends. Knowing about the two waves can help you remember that the first jolt is not the strongest part.

Sources

Revision history

October 23, 2026: First published.