Seismic Waves and How They Appear on Seismograms

What a Seismogram Actually Records

A seismogram is a time series of ground motion measured at a specific location. It is not a picture of the fault breaking; it is the instrument’s record of how the ground at the station moved as different seismic waves arrived and passed. The horizontal axis is time, and the vertical axis is motion (commonly velocity in nm/s or mm/s, sometimes acceleration in m/s², or displacement in mm). The trace can be from one component (vertical Z, north-south N, east-west E) or three components recorded simultaneously.

Two ideas help you read any seismogram: (1) different wave types travel at different speeds, so they arrive in a predictable order; (2) the amplitude and frequency content change with distance, path, and site conditions, so the same earthquake can look different at different stations. Your job when reading a seismogram is to separate “arrival timing” (useful for locating and characterizing the event) from “waveform shape” (useful for understanding wave types, energy, and local effects).

Common measurement types: displacement, velocity, acceleration

Modern broadband seismometers often record ground velocity, while strong-motion accelerometers record acceleration during large shaking. The same physical motion can be represented in different ways: displacement emphasizes long-period motions (slow swaying), velocity is often a good compromise for many wave arrivals, and acceleration highlights high-frequency shaking (sharp jolts). If you compare seismograms from different instruments, confirm what is being plotted; otherwise, you may misinterpret which part of the record is “largest.”

Why three components matter

Seismic waves have particle motions (how the ground moves) that depend on wave type and direction of travel. A single vertical trace can show clear arrivals, but three components let you distinguish wave types more reliably. For example, Love waves are primarily horizontal and transverse (side-to-side), so they can be strong on horizontal components and weak on vertical. Rayleigh waves involve elliptical motion in a vertical plane, so they often appear on both vertical and radial horizontal components.

Seismic Wave Types You Will See on Seismograms

Body waves: P and S

Body waves travel through the Earth’s interior. They usually arrive before surface waves and are key for timing picks.

Continue in our app.

You can listen to the audiobook with the screen off, receive a free certificate for this course, and also have access to 5,000 other free online courses.

Or continue reading below...

Download the app

• P waves (primary, compressional): Particle motion is mostly parallel to the direction of travel (push-pull). P waves are typically the first clear arrival on a local or regional seismogram. They often have smaller amplitudes than later phases but can be sharp and high frequency, especially at short distances.
• S waves (secondary, shear): Particle motion is mostly perpendicular to the direction of travel. S waves arrive after P waves because they travel more slowly. They often have larger amplitudes and can dominate damaging shaking at local distances. S energy tends to be strong on horizontal components.

Surface waves: Love and Rayleigh

Surface waves travel along the Earth’s surface and usually arrive after body waves. They can have large amplitudes and longer durations, especially at regional and teleseismic distances. Because they are dispersive (different periods travel at different speeds), their waveforms can “spread out” in time, producing long wavetrains.

• Love waves: Horizontally polarized shear motion, typically strongest on the transverse horizontal component. They can be very damaging because they produce strong horizontal shaking.
• Rayleigh waves: Retrograde or prograde elliptical motion in a vertical plane. They often show as large, rolling oscillations on vertical and radial components and can last a long time.

Other phases you may encounter

Depending on distance and Earth structure, additional arrivals may appear. You do not need to identify every phase to read a seismogram effectively, but knowing they exist prevents confusion.

• Pg, Pn, Sg, Sn: Regional crustal and upper-mantle guided phases. They can create multiple “P-like” or “S-like” arrivals, especially between ~200–1500 km.
• Reflections and conversions: Waves can reflect off boundaries (like the Moho) or convert between P and S at interfaces, producing extra arrivals that may sit between the main P and S or after them.
• Coda: The trailing, gradually decaying wavetrain after the main arrivals, produced by scattering from heterogeneities. Coda length and decay can provide clues about attenuation and site effects.

How Wave Speed Controls What You See

Arrival order is governed by wave speeds. In typical crustal rocks, P waves travel roughly 5–7 km/s and S waves roughly 3–4 km/s (values vary by geology). Surface waves are often slower than S waves at short periods, but their speed depends strongly on period and structure.

Because P is faster than S, the time gap between P and S arrivals increases with distance. This gap is one of the most practical “rulers” on a seismogram: a larger P–S time difference usually means the earthquake is farther away (for local and regional events).

Practical rule-of-thumb for local distances

In many crustal settings, a rough approximation is that distance in kilometers is about 8 times the P–S time in seconds. This is not universal, but it is useful for quick estimates when you do not have a full velocity model. For example, a P–S time of 5 seconds suggests a distance on the order of ~40 km.

Recognizing P, S, and Surface Waves on a Seismogram

What P arrivals often look like

P arrivals are often the first noticeable deviation from background noise. On a filtered local seismogram (for example, 1–10 Hz), the P onset can be relatively sharp. On the vertical component, P can be especially clear. However, P can be emergent (gradual) if the path is complex, the event is small, or the station is noisy.

Practical cues for P:

• First consistent change from noise across multiple components.
• Often higher frequency than later surface waves.
• Often clearer on the vertical component for local events.

What S arrivals often look like

S arrivals typically show a stronger increase in amplitude than P and are often more prominent on horizontal components. The onset may be less sharp than P in some cases, but the jump in energy is usually obvious for nearby events.

Practical cues for S:

• Arrives after P with a noticeable time gap.
• Often larger amplitude than P at local distances.
• Often stronger on horizontal components.

What surface waves often look like

Surface waves can appear as large, slower oscillations that continue for many cycles. At regional distances, they can dominate the record. Their period content is often longer than the body waves, and their amplitude can build gradually.

Practical cues for surface waves:

• Arrive after S (often significantly later for regional/teleseismic events).
• Longer period, more “rolling” appearance.
• Long duration wavetrain; may show dispersion (period changes over time).

Step-by-Step: Picking P and S Arrivals by Hand

Manual picking is a foundational skill because it forces you to connect waveform features to physical arrivals. Even if software later refines picks, your initial picks guide quality control.

Step 1: Prepare the trace

• Confirm the component (Z, N, E) and units (velocity, acceleration, displacement).
• Choose a time window that includes at least a minute before the expected event and several minutes after.
• Apply a simple bandpass filter appropriate to the distance: for local events, try 1–10 Hz; for regional, 0.5–5 Hz; for teleseismic, 0.05–1 Hz. If you cannot filter, zoom and adjust scaling to make onsets visible.

Step 2: Estimate the noise level

Look at the pre-event portion. Note the typical amplitude and whether the noise is steady or spiky. A station near traffic or wind may have bursts that mimic arrivals. Your pick should be a sustained change, not a single spike.

Step 3: Identify the first plausible P onset

• Scan forward in time for the first sustained deviation from background noise.
• Check whether the change appears on more than one component at about the same time.
• Mark the earliest time where the signal consistently departs from noise. If the onset is emergent, pick the point where the waveform begins trending away from the baseline in a persistent way.

Step 4: Identify the S onset

• From the P pick, move forward by a few seconds to tens of seconds depending on expected distance.
• Look for a clear increase in amplitude and a change in waveform character (often more energetic, sometimes lower frequency than P).
• Use horizontal components to confirm: S often strengthens there.

Step 5: Quality-check with P–S time

Compute the P–S time difference. If it is extremely small (for example, less than ~1 second) for what you believe is a local earthquake, you may have picked noise or a later phase as P. If it is extremely large for a nearby event, you may have missed the true P onset. Use the rule-of-thumb distance estimate as a sanity check.

Step 6: Note surface-wave onset (optional but useful)

Mark where long-period, large-amplitude oscillations begin. This helps distinguish a local event (where surface waves may be modest) from a more distant event (where surface waves can dominate).

Practical Example Workflow: From One Station to a Distance Estimate

Suppose you have a three-component seismogram from a nearby station. You pick P at 12:00:10.2 and S at 12:00:16.0. The P–S time is 5.8 seconds. Using the rough multiplier of 8, the epicentral distance is about 46 km. This does not locate the earthquake uniquely (you need multiple stations for that), but it provides immediate situational awareness: the event is likely within the local region, and strong shaking may be possible depending on magnitude and site conditions.

Now compare with a second station where P–S is 12 seconds. That station is likely farther away (roughly ~96 km by the same rule). If the farther station shows much larger surface waves than the closer one, that may indicate path or site amplification differences, or simply that the instrument types differ. This is why you always check metadata and component orientation.

How Distance and Magnitude Change Seismogram Appearance

Distance effects

• Increasing P–S gap: P and S separate more with distance, making phase identification easier up to a point.
• Attenuation of high frequencies: As waves travel farther, high-frequency energy is absorbed and scattered, so distant records look smoother and more long-period.
• Surface-wave dominance: At regional and teleseismic distances, surface waves can become the largest part of the record.

Magnitude effects

• Amplitude: Larger events generally produce larger amplitudes, but instrument gain, distance, and site effects can overwhelm simple comparisons.
• Duration: Larger events tend to have longer rupture durations and longer-lasting coda, so the seismogram stays “busy” longer.
• Clipping: If the instrument saturates, the waveform tops flatten or distort. Strong-motion sensors are designed to avoid this for large nearby shaking, while sensitive broadband sensors may clip in extreme cases.

Frequency Content: Why Filtering Changes What You Think You See

A raw seismogram contains a mix of frequencies. Filtering is not “cheating”; it is a way to isolate features. But filtering can also shift apparent onsets or hide phases if used poorly.

• High-pass filtering (removing long periods) can make P and S onsets sharper and reduce microseism noise, but it can suppress surface waves.
• Low-pass filtering (removing high frequencies) can clarify long-period surface waves and teleseismic phases, but it can smear the onset of local P and S.
• Bandpass filtering is often best for picking: choose a band where the signal-to-noise ratio is highest.

Practical step: try two filters and compare picks. If your P pick changes drastically between filters, the onset is likely emergent or the trace is noisy; note lower confidence.

Polarization and Component Checks: A Practical Way to Confirm Wave Type

With three components, you can use simple polarization logic without advanced math.

Step-by-step component check

• Rotate mentally: think of motion as vertical vs horizontal. P energy often appears strongly on vertical for local events; S often appears strongly on horizontal.
• Compare N and E: if one horizontal component is much stronger during a phase, the wave motion may be oriented in a particular direction relative to the station.
• Look for phase-consistent behavior: a true arrival tends to show coherent oscillations across components, not random spikes.

If you have the ability to rotate horizontals into radial and transverse components (radial points to the source, transverse is perpendicular), Love waves tend to dominate transverse, while Rayleigh waves tend to show strong radial and vertical motion. Even without rotation, a strong late-arriving horizontal wavetrain with weak vertical energy is a hint of Love-wave dominance.

Common Pitfalls When Reading Seismograms

Confusing noise transients with arrivals

Short spikes from cultural noise, electrical glitches, or sensor bumps can look like a P onset. A real P arrival is usually followed by additional coherent motion, not a single isolated spike. Always check whether the “arrival” persists and whether it appears on multiple components.

Misidentifying S as P (or vice versa)

If you start looking at the record too late, you may miss the true P and label S as P. The result is an unrealistically small P–S time and a distance estimate that is too short. Prevent this by always scanning from well before the suspected event time and by checking the vertical component for an earlier, smaller onset.

Instrument response and scaling issues

Two stations can record the same event with very different amplitudes because of different instrument gains or because one trace is acceleration and the other is velocity. If you are comparing amplitudes, confirm that the data are corrected to the same physical quantity and that the plotting scales are comparable.

Clipping and baseline shifts

Clipping can hide true peak amplitudes and distort wave shapes, making phase identification harder. Baseline shifts (a sudden offset) can occur in strong-motion records during intense shaking. In such cases, arrival times may still be pickable, but amplitude-based interpretations should be cautious.

Reading Seismograms for Hazard Awareness: What to Pay Attention To

For practical hazard awareness, you often need quick, robust observations rather than perfect phase catalogs.

• How quickly did strong energy arrive after P? A short P–S time suggests the source is close, meaning strong shaking may follow quickly after the first detectable arrival.
• How long does strong shaking last? Duration on the seismogram can hint at event size and at whether surface waves are significant at your location.
• Which components are strongest? Strong horizontal energy is often more relevant for building response, while strong vertical spikes can affect certain structures and lifelines.
• Is the waveform dominated by long-period motion? Long-period surface waves can be especially important for tall buildings, bridges, and soft-soil basins.

Mini-Lab: Annotate a Seismogram Like a Practitioner

This exercise can be done with any plotted seismogram (printed or on screen). The goal is to build a repeatable annotation habit.

Step 1: Create an annotation legend

• Use “P?” for uncertain P, “P” for confident P, similarly “S?” and “S”.
• Mark “Surf” where surface-wave energy begins.
• Mark “Coda” where the signal begins a steady decay toward noise.

Step 2: Mark noise characteristics

• Write a short note: “steady low noise,” “spiky noise,” “microseism visible,” or “cultural noise bursts.”

Step 3: Pick P and S on each component

• Pick P on Z, then check N and E for consistency.
• Pick S on N and E, then check Z.
• If picks differ by more than a second between components for a local event, label them uncertain and re-check filtering or scaling.

Step 4: Compute P–S and estimate distance

• Compute P–S in seconds.
• Multiply by ~8 for a quick distance estimate in km (local/regional rough use).
• Write the estimate next to the trace: “Dist ~ ___ km (rough).”

Step 5: Describe the waveform in plain language

Write two sentences: one about onset sharpness (impulsive vs emergent) and one about dominant motion (short-period shaking vs long-period rolling). This habit helps you communicate observations to non-specialists during an event.

Seismogram Appearance Across Scales: Local, Regional, Teleseismic

Local (within tens of km)

• P and S arrivals are close together in time.
• High-frequency energy is strong; waveforms look “busy.”
• Surface waves may be present but not always dominant.

Regional (hundreds of km)

• P–S gap is larger; multiple regional phases may appear.
• High frequencies are reduced; S and surface waves can be prominent.
• Surface-wave dispersion may be noticeable as changing period with time.

Teleseismic (thousands of km)

• Arrivals are more separated; long-period energy is often emphasized.
• Surface waves can be very large and long lasting.
• Filtering to low frequencies is often necessary to see clear phases.

Quick Reference: Visual Cues for Wave Identification

• P wave: first arrival; often sharper; often clearer on vertical; smaller amplitude than S at local distances.
• S wave: arrives after P; larger amplitude; stronger on horizontals; often marks onset of stronger shaking.
• Surface waves: arrive later; longer period; large rolling oscillations; long duration; can show dispersion.
• Coda: trailing decay; scattered energy; not a distinct single phase.