Seismic attribute

A seismic attribute is any quantity derived from seismic data using measured time, amplitude, frequency, phase, attenuation or any combination of these. It intends to output a subset of the data that quantifies rock and fluid properties and/or allows the recognition of geological patterns and features. Almost all seismic attributes are post-stack but there are a few pre-stack ones. They can be measured along a single seismic trace or throughout various seismic traces.

The first attributes developed were related to the 1D complex seismic trace and included: envelope amplitude, instantaneous phase, instantaneous frequency, and apparent polarity. Acoustic impedance obtained from seismic inversion can also be considered an attribute and was among the first developed (Sheriff, 2002).

Other attributes commonly used include: coherence, azimuth, dip, instantaneous amplitude, response amplitude, response phase, instantaneous bandwidth, AVO, and spectral decomposition.

Amplitude attributes
Amplitude attributes use the seismic signal amplitude as the basis for their computation.

Mean amplitude
A post-stack attribute that computes the arithmetic mean of the amplitudes of a trace within a specified window. This can be used to observe the trace bias which could indicate the presence of bright spots.

Average Energy
A post-stack attribute that computes the sum of the squared amplitudes divided by the number of samples within the specified window used. This provides a measure of reflectivity and allows one to map direct hydrocarbon indicators within a zone of interest.

RMS (root mean square) amplitude
A post-stack attribute that computes the square root of the sum of squared amplitudes divided by the number of samples within the specified window used. With this, one can measure reflectivity in order to map direct hydrocarbon indicators in a zone of interest. However, RMS is sensitive to noise as it squares every value within the window.

Maximum magnitude
A post-stack attribute that computes the maximum value of the absolute value of the amplitudes within a window. This can be used to map the strongest direct hydrocarbon indicator within a zone of interest.

AVO attributes
AVO (amplitude versus offset) attributes are pre-stack attributes that have as the basis for their computation, the variation in amplitude of a seismic reflection with varying offset. These attributes include: AVO intercept, AVO gradient, intercept multiplied by gradient, far minus near, fluid factor, etc (Castagna and Backus, 1993).

Coherence
A post-stack attribute that measures the continuity between seismic traces in a specified window along a picked horizon. It can be used to map the lateral extent of a formation. It can also be used to see faults, channels or other discontinuous features.

Although it should be used along a specified horizon, some software packages (i.e. Petrel) compute this attribute along arbitrary time-slices.

Dip
A post-stack attribute that computes, for each trace, the best fit plane (3D) or line (2D) between its immediate neighbor traces on a horizon and outputs the magnitude of dip (gradient) of said plane or line measured in degrees. This can be used to create a pseudo paleogeologic map on a horizon slice.

Azimuth
A post-stack attribute that computes, for each trace, the best fit plane (3D) between its immediate neighbor traces on a horizon and outputs the direction of maximum slope (dip direction) measured in degrees, clockwise from north. This is not to be confused with the geological concept of azimuth, which is equivalent to geological strike and is measured 90° counterclockwise from the dip direction.

Curvature
A group of post-stack attributes that are computed from the curvature of a specified horizon. These attributes include: magnitude or direction of maximum curvature, magnitude or direction of minimum curvature, magnitude of curvature along the horizon's azimuth (dip) direction, magnitude of curvature along the horizon's strike direction, magnitude of curvature of a contour line along a horizon.

Frequency attributes
These attributes involve separating and classifying seismic events within each trace based on their frequency content. The application of these attributes is commonly called spectral decomposition. The starting point of spectral decomposition is to decompose each 1D trace from the time domain into its corresponding 2D representation in the time-frequency domain by means of any method of time-frequency decomposition such as: short-time Fourier transform, continuous wavelet transform, Wigner-Ville distribution, matching pursuit, among many others. Once each trace has been transformed into the time-frequency domain, a bandpass filter can be applied to view the amplitudes of seismic data at any frequency or range of frequencies.

Technically, each individual frequency or band of frequencies could be considered an attribute. The seismic data is usually filtered at various frequency ranges in order to show certain geological patterns that may not be obvious in the other frequency bands. There is an inverse relationship between the thickness of a rock layer and the corresponding peak frequency of its seismic reflection. That is, thinner rock layers are much more apparent at higher frequencies and thicker rock layers are much more apparent at lower frequencies. This can be used to qualitatively identify thinning or thickening of a rock unit in different directions.

Spectral decomposition has also been widely used as a direct hydrocarbon indicator.