You are on page 1of 2

GEOPHYSICS, VOL. 66, NO. 1 (JANUARY-FEBRUARY 2001); P. 47–48, 1 FIG.

Understanding seismic attributes

Alistair R. Brown∗

With the success of 3-D surveys has come the popularity of positional precision is typically around a quarter of a millisec-
seismic attributes. Attributes are valuable for gaining insight ond and the precision of the attribute value somewhat compa-
from the data particularly when displayed spatially over in- rable. Windowed attributes, on the other hand, use the sample
terpreted horizons. However, all the many attributes available values at, for example, 2 or 4 ms intervals. They may add up or
are not independent pieces of information but, in fact, simply average all the sample values to give a gross attribute, they may
represent different ways of presenting a limited amount of ba- select one unique attribute value, or a distribution or trend in
sic information. The key to success lies in selecting the most the attribute values over the window may be calculated.
applicable attribute for the problem. Furthermore, statistical Time-derived attributes help to discern structural detail.
analysis using attributes must be based on understanding, not Amplitude-derived and frequency-derived attributes address
simply mathematical correlation. problems of stratigraphy and reservoir properties. Amplitude
Figure 1 is a classification of seismic attributes aimed at un- attributes are here the most robust and useful, but frequency
derstanding what is available and what use each may have. attributes may help reveal additional geologic layering. The
First, attributes are related to the fundamental information hybrid attributes are particularly interesting because they con-
in seismic data: time, amplitude, frequency, and attenuation. tain elements of amplitude and frequency, and are thus useful
Most of the attributes we use are poststack, that is from the measurements of seismic character.
stacked and migrated data volume loaded on our workstations. Petrophysical studies and seismic modeling, of course, are
The prestack attributes are principally derived from amplitude other sources of understanding. Such understanding is neces-
variations with offset (AVO) measurements. sary to build confidence in observed correlations with reser-
Horizon attributes are extracted along a tracked horizon and voir parameters and must be part of all attribute analysis
thus benefit from the precision of machine autotrackers. The projects.


7706 El Santo Lane, Dallas, Texas 75246-4316. E-mail: alistair@airmail.net.
°
c 2001 Society of Exploration Geophysicists. All rights reserved.

47
48
Geophysics in the new millennium

FIG. 1. Seismic attributes derived from or related to the basic seismic information of time, amplitude, frequency, and attenuation.
The window can be a constant time interval, a constant interval hung from one horizon, or the interval between two horizons.

You might also like