From source to subsurface image: how the seismic method actually works

A recorded wavefield, not a photograph

Seismic exploration is often described as imaging the Earth with sound. The analogy is useful, but incomplete. A seismic section is a model-dependent reconstruction produced by recording elastic-wave energy, processing the measurements, and mapping observations into the subsurface using assumptions about wave propagation.

A survey begins with an energy source. On land this may be a vibrator or a controlled explosive source; offshore surveys commonly use compressed-air arrays. Energy propagates as elastic waves and is reflected, refracted, converted, or scattered when material properties change. Receivers record the returning wavefield through time, and one activation produces a shot gather: a family of traces across source–receiver offsets.

Geometry determines the information budget

Source spacing, receiver spacing, aperture, azimuthal coverage, maximum offset, and bandwidth govern illumination and the range of spatial wavelengths that can be recovered. Processing can organize and enhance recorded information, but it cannot recreate wavefield content that was never measured.

This is why acquisition design and quality control belong to interpretation. A beautifully processed line may still have weak sensitivity beneath an obstruction, at the edge of the aperture, or below a strongly attenuating overburden.

Processing should preserve meaning

Raw records contain reflections together with ground roll, multiples, ambient noise, coupling variation, source effects, and near-surface distortion. Early processing verifies geometry and attenuates unwanted energy, but the objective is not merely a clean-looking section. Filters must preserve the phase, timing, and—where needed—relative amplitude of signals carrying geological information.

Near-surface velocity variation is especially important on land. If weathering and elevation effects are handled poorly, deeper events can be misplaced during imaging. Every correction is therefore both a numerical operation and a geological assumption.

Velocity turns time into depth

Seismic traces are recorded in time, while most geological decisions are made in space. Converting one to the other requires a velocity model built from moveout, tomography, refraction information, migration updates, full-waveform inversion, or some combination of these.

Velocity and structure are coupled. A slow or fast model can move a reflector to the wrong depth, create false relief, defocus events, or degrade fault positioning. In complex geology, the velocity model can be as consequential as the final migration.

Imaging and interpretation remain inverse problems

Migration relocates energy to plausible subsurface origins. Ray-based methods can be efficient; wave-equation methods better represent complex propagation and multipathing. Neither removes limited aperture, finite bandwidth, or incomplete illumination.

Interpreters then combine seismic character with wells, rock physics, regional geology, and structural reasoning. Multiple geological models can explain the same response. A bright amplitude, for example, is not uniquely diagnostic of hydrocarbons: lithology, tuning, pressure, processing, and fluids can produce similar observations.

• What wavefield was generated and recorded?
• Which areas and wavelengths were adequately illuminated?
• How was the velocity model constructed and tested?
• Which conclusions are observations, and which are interpretations?