Noise taxonomy

Physics prerequisites for acquisition

Learning objectives

Identify five canonical noise families: thermal, ambient ocean, ground roll, cultural, shipping/harmonic
Recognise each family's time-domain signature and frequency-domain signature
Predict which family will dominate at a given site (marine, land, urban, remote)
Use the noise model to inform receiver choice and array design

“Noise” is a single word for a dozen different physical phenomena. Knowing which noise family dominates at your site is how you pick receivers, design arrays, and set expectations for the data.

The five families

Thermal (white Gaussian) — noise from the sensor itself and its electronics. Flat spectrum, independent across traces. Fundamental floor.
Ambient ocean (red/pink) — microseisms from wind and swell, dominant at 1–8 Hz. The main low-frequency enemy in marine surveys.
Ground roll — Rayleigh-wave surface noise on land. Strong, low-frequency (4–15 Hz), slow (300–800 m/s), dispersive. Coherent across traces, which is how f-k filtering catches it.
Cultural (impulsive) — traffic, hammering, nearby construction. Bursts rather than stationary; show up as random spikes in time and broad energy in frequency.
Shipping (harmonic) — narrow lines at the propeller shaft frequency and overtones. Notable in marine; hit it with a narrow-band reject filter.

Why this shapes acquisition

A coiled-geophone array is useless against thermal noise (random per-sensor) but kills ground roll (coherent, wavelength < array length). Dual-sensor streamers cancel sea-surface ghost noise. An OBN node on the sea floor dodges swell noise entirely — hence OBN’s celebrated low-frequency signal. Every receiver choice is a bet against a specific noise family; you have to know the zoo.

References

Sheriff, R. E., Geldart, L. P. (1995). Exploration Seismology (2nd ed.). Cambridge University Press.
Yilmaz, Ö. (2001). Seismic Data Analysis: Processing, Inversion, and Interpretation of Seismic Data (2 vols.). SEG Investigations in Geophysics 10.
Pritchett, W. C. (1990). Acquiring Better Seismic Data. Chapman & Hall.