Lecture 13: Seismic reflection data

Seismic Reflection Data

The lecture explains how seismic reflection surveys create images of the subsurface, how those images are processed, and what limits their resolution.

Reflection and Acoustic Impedance

Seismic reflection is analogous to hearing an echo: a wave reflects when it encounters a boundary between materials with different properties.

A material’s acoustic impedance is its density multiplied by the sound-wave velocity in it:

[ Z = \rho V ]

The amount of energy reflected at an interface depends on the acoustic impedances on either side. A strong contrast produces a stronger reflection; where the contrast is small, most of the energy passes through and only a small amount reflects.

Seismic surveys primarily measure compressional, or P-waves. The lecture gives typical rock velocities of about 2.5–6.8 km/s, with most sedimentary rocks below 6 km/s.

Seismic Sections Show Time, Not Depth

A survey records the time between sending a wave and receiving its echo. This is two-way travel time: the wave travels down to a reflector and back.

The recorded time depends on both the distance traveled and the velocity of the rocks. Therefore, converting a seismic section from time to depth requires knowing the subsurface velocities.

Lateral velocity changes can create apparent structures. For example, a reflector may appear to step down beneath a slower-velocity basin compared with faster basement. This is a velocity pull-down—an imaging artifact rather than a real step in the reflector.

Vertical Exaggeration and Interpretation

The lecture demonstrates converting a time section to depth using an assumed constant P-wave velocity of 3 km/s. At that velocity, one second of two-way travel time corresponds to 1.5 km of depth.

In the example, the displayed section has a vertical exaggeration of roughly 3:1. Reducing its vertical dimension to one-third removes that exaggeration.

Because the example assumes a constant velocity, it is only an approximation. The broader lesson is that a seismic section displayed in time is not automatically a geological cross-section; velocity and display scaling affect its apparent shape.

Building a Clearer Image: CMP Processing and Stacking

Most interfaces produce weak reflections, so surveys record subsurface points multiple times to improve the signal relative to noise.

The processing sequence is:

1. Collect traces into a CMP/CDP gather. As the source moves along the survey line, different source–receiver pairs record reflections passing through the same common midpoint (CMP), also called a common depth point (CDP).
2. Correct for different travel times. Rays in a gather travel different distances, so their reflections arrive at different times. A normal moveout (NMO) correction shifts the traces into alignment. The NMO velocity is the velocity that best aligns the traces in the gather.
3. Stack the aligned traces. Once aligned, the traces are added together. The coherent reflection signal builds up while noise is reduced. This is repeated for each gather. The lecture’s example notes that with 96 receivers, a midpoint may be sampled by 48 traces.

Migration and Dipping Reflectors

Basic CMP processing places reflection points as if they were directly beneath the surface location, which works well for horizontal layers.

For dipping layers, the reflection point is not directly below the receiver position. The unmigrated image can therefore shift dipping reflectors down-dip and make them appear less steep than they are.

Migration corrects this positional distortion and also removes diffraction effects. Migrated sections may show broad migration “smiles,” particularly near the bottom and edges of an image.

Seismic Resolution

Resolution depends on wavelength, which is velocity divided by frequency. With increasing depth, velocity generally increases and frequency decreases, so wavelength grows and resolution tends to decline.

• Vertical resolution: The thinnest separable features are approximately one-quarter of the dominant wavelength. In the example, a 50 Hz signal traveling at 5 km/s has a wavelength of 100 m, making the approximate vertical resolution limit 25 m. Layers closer together than this may appear as a single reflector.
• Subseismic faults: Faults with vertical displacement smaller than roughly one-quarter wavelength may not be visibly resolved. The lecture calls these subseismic faults.
• Horizontal resolution: This depends on the first Fresnel zone. Features smaller than this zone may not be imaged as continuous reflectors and can instead act like point sources that generate diffractions.

The Fresnel-zone radius increases as frequency decreases. Because frequency tends to decrease with depth, horizontal resolution also worsens with depth.

A comparison of high- and low-resolution seismic sections from the Wisting field in the southwest Barents Sea illustrates that higher-frequency data can image thinner layers and faults more clearly.

Multiples

Strongly reflective interfaces, such as the water bottom, can generate repeated reflections called multiples.

A multiple may appear at twice the travel time of the primary reflection. Multiples from dipping layers may have twice the dip of the primary. Peg-leg multiples result when a ray bounces between two subsurface layers.

The lecture shows an example where a multiple can be identified because it dips more steeply than the primary reflector.

Speakers and Sources Featured

• Speaker: An unnamed lecturer presenting the material; no other speakers are identifiable in the subtitles.
• Source/example cited: Chapter 11 of “Almendinger’s lab manual” (the name is transcribed this way in the subtitles; the exact spelling is uncertain).
• Data example mentioned: High- and low-resolution seismic sections from the Wisting field, southwest Barents Sea.