AVO Attribute Computations2126 ProMAX® Reference

AVO Attribute Computations

AVO Attribute Computations computes the output of AVO

Attribute Stacks and AVO Weighted Stacks into a more
meaningful interpretation dataset.

Theory
The primary outputs from AVO Attribute Stacks and AVO
Weighted Stacks are traces which relate to the P-wave
impedance contrast and S-wave impedance contrast of
reflecting horizons. These primary attributes may not be in
the most suitable form for interpretation of the AVO
anomalies which they represent. The purpose of AVO
Attribute Computations is to allow these primary attributes
to be transformed into a more useful form for interpretation.
The transformations involve mathematical operations
between the two outputs so that those anomalies which are
of interest may stand out more clearly from a background
trend.

Usage
Outputs from both AVO Attribute Stacks and from AVO
Weighted Stacks may be read into this process. Since the
header attribute AVO_ATTR is uniquely defined for these two
processes, AVO Attribute Computations is able to determine
the source of input and handle the differences appropriately.

The statistical attributes from the least squares amplitude

fitting in the AVO Attribute Stacks process, runs statistic
(AVO_ATTR = 8) and correlation coefficient (AVO_ATTR = 7),
are not available in AVO Weighted Stacks. Therefore, any
editing and scaling which may involve these attributes have
no effect if the input data comes from AVO Weighted Stacks.

This process outputs a single stacked trace for each input

ensemble. The only input attributes which are used are those
with header AVO_ATTR = 1, 2, 11, and 12. These are the
Intercept and Gradient traces for AVO Attribute Stacks and
the Vstack and Wstack traces for AVO Weighted Stacks.

The following are the types and method of computation:

• Intercept times Gradient

Other Docs Search Page Known Problems

AVO Attribute Computations2127 ProMAX® Reference

The first attribute, the product of intercept and gradient,

is a good prospecting tool since it highlights events with
strong zero offset reflection coefficients also having
appreciable AVO characteristics. An increase in absolute
amplitude with offset is a peak and a decrease in AVO is
a trough on the seismic section.
For the AVO Weighted Stack input data, this is
computed as:
Output = Vstack × ( Vstack + Wstack )

• Sign(Intercept) times Gradient

The sixth attribute is the product of the sign of the
intercept and the gradient. This retains the value of the
gradient but the polarity varies with the combined
polarities of the gradient and intercept.
For the AVO Weighted Stack input data, this is
computed as:
Output = Sign(Vstack ) × ( Vstack + Wstack )

• Fluid Factor
The Fluid Factor is computed using Castagna’s mudrock
equation as an estimate of the P-wave to S-wave velocity
ratio. Deviation of intercept and gradient from a regional
trend will show up as a Fluid Factor anomaly. The Fluid
Factor equation is:
dVp Vs dVs
dF = ------------- – 1.16 × ------- × ------------
dVpa Vp dVsa

Where for AVO Attribute Stacks:

dVp
------------- = 1.6 × Intercept
dVpa

and
dVs
------------ = Intercept – Gradient
dVsa

And for AVO Weighted Stacks:

dVp
------------- = Vstack
dVpa

Other Docs Search Page Known Problems

AVO Attribute Computations2128 ProMAX® Reference

and
dVs
------------ = Wstack
dVsa

The mudrock trend used to compute the background

shear wave velocity for the Vs/Vp ratio is:

Vs = 0.8621 × Vp – 1.1724

Where the units of velocity are km/s, and the

coefficients of the linear mudrock equation are available
as parameters: Mudrock line slope coefficient and
Mudrock line intercept coefficient.
• Least squares Difference
The difference between the two primary AVO attributes
has been shown to be a hydrocarbon indicator,
particularly when the two attributes are scaled to
approximately equal amplitudes. If the two attributes
have approximately equal amplitudes, then the only
evident differences will be the anomalies which stand
out from the background trend. So that some differences
will be seen, the scaling function is computed over a gate
specified by the parameter: Gate length for least
squares scaling. The second attribute is scaled by this
function and subtracted from the first attribute.
• Least squares Scalers
The least squares scalers are computed as described
above for Least squares Difference. They can be output
as a trace.
• Scaled Difference
This method computes the difference between the two
primary AVO attributes, but allows the specification of a
time-invariant scalar value. This value may be positive or
negative. A negative value will result in the addition of
the two attributes after scaling the second attribute.
• Angle of Crossover
This AVO attribute corresponds to the angle of incidence
at which a polarity reversal occurs. This attribute is

Other Docs Search Page Known Problems

AVO Attribute Computations2129 ProMAX® Reference

calculated as long as the crossover angle is determined

to be less than 30 degrees. If the crossover angle is
greater than 30 degrees, a value of 31 degrees is output.
Absolute amplitude decreases with offset until it reaches
the crossover angle, flips polarity, and subsequently
increases. Events with this response typically appear
weak on a stacked section. Nonetheless, this polarity
reversal effect has been known to indicate rock
properties consistent with pay zones.
• Poissons Reflectivity
Poissons Reflectivity has been described in terms of both
intercept and gradient and weighted stacks In both
cases it has been defined as:
∆Vp ∆Vs
PoissonsReflectivity = ----------- – ----------
Vp Vs
This may be computed as the sum of intercept and
gradient divided by two. It may also be computed as the
difference between Vstack and Wstack. It is computed
appropriately depending on the type of input data.

References
Castagna, J.P, 1993, AVO analysis - tutorial and review: in Castagna, J. P. and
Backus, M. M., Eds., Offset-dependent reflectivity - Theory and Practice of AVO
analysis: Soc. Expl. Geophys.

Koefoed, O., 1955, On the effect of Poisson’s ratios of rock strata on the reflection
coefficients of plane waves: Geophysical Prospecting, 3, 381-387.

Ostrander, W.J., 1984, Plane-wave reflection coefficients for gas sands at non-
normal angles of incidence: Geophysics, 49, 1637-1648.

Shuey, R.T.,1985, A Simplification of the Zoeppritz Equations: Geophysics, 50,

609-614.

Smith, G. C., and Gidlow, P.M., 1987, Weighted stacking for rock property estima-
tion and detection of gas: Geophysical Prospecting, 35, 993-1014.

Walden, A.T., 1991, Making AVO sections more robust. Geophysical Prospecting,
39, 915-942.

Parameters

Type of output

Select the type of output:

Other Docs Search Page Known Problems

AVO Attribute Computations2130 ProMAX® Reference

• Intercept x Gradient outputs the product of the inter-

cept and gradient or the product of Vstack and Vstack
plus Wstack depending on the type of input data.
• Sign(Intercept) x Gradient outputs the product of the
sign of intercept and gradient or the product of the sign
of Vstack and Vstack plus Wstack depending on the type
of input data.
• Fluid Factor outputs amplitudes of the distance from
the water-saturated mudrock line. Anomalies are rela-
tive to this line.
• Lsqr Difference outputs the difference between the two
primary input attributes, after removing the background
trend. The background trend is computed over a speci-
fied time gate.
• Lsqr Scalers outputs the scalers which define the ratio
of amplitudes over a gate as defined for the Lsqr Differ-
ence output.
• Scaled Difference outputs the difference between the
two primary input attributes after multiplying the sec-
ond attribute by a scalar value.
• Angle of Crossover outputs the angle of incidence at
which the gradient (slope) of the AVO effect crosses the
zero-amplitude axis. If greater than 30 degrees, 31 is
output.
• Poissons Reflectivity outputs the sum of the intercept
and gradient or the difference between the Vstack and
Wstack.

Is this a spatially variant velocity calculation?

This appears if Fluid Factor to Type of output. Select Yes if

this a spatially variant velocity calculation. However, the
velocity field input can only have a single function. This flag
allows minor speed-up for spatial invariance of the velocity
field.

Get velocities from the database?

This appears if Fluid Factor to Type of output. Select Yes if

the velocity functions already exist in a velocity parameter
file. Select No to enter the velocity function.

Other Docs Search Page Known Problems

AVO Attribute Computations2131 ProMAX® Reference

Gate Length for least squares scaling

This appears if Lsqr Difference or Lsqr Scalers to Type of

output. Enter the gate length in ms to use for computing the
amplitude ratio of the two primary input attributes.

Scalar value for scaled difference

This appears if Scaled Difference to Type of output. Enter

the scalar value to use for multiplying the second (gradient or
Wstack) of the two primary input attributes before
subtracting from the first attribute (intercept or Vstack). The
attributes may be summed by submitting a negative value for
this parameter.

Select velocity parameter file

This appears if Yes to Get velocities from the database.

Enter a velocity parameter file.

Velocity function

This appears No to Get velocities from the database. Time,

and possibly spatially, variant velocity field indexed by CDP.

Mudrock line slope coefficient

This appears if Fluid Factor to Type of output. Enter the

slope of the mudrock equation. This value controls the fit of
the P-wave to the S-wave reflectivity for the output Fluid
Factor. It represents the rate of change of S-wave velocity for
a change in P-wave velocity.

Vs = 0.8621 × Vp – 1.1724

Mudrock line intercept coefficient

This appears if Fluid Factor to Type of output. Enter the

intercept of the mudrock equation. This value controls the fit
of the P-wave to the S-wave reflectivity for the output Fluid
Factor. It represents the S-wave velocity intercept for a zero
P-wave velocity and has units of meters per second.

Other Docs Search Page Known Problems

AVO Attribute Computations2132 ProMAX® Reference

Vs = 0.8621 × Vp – 1.1724

Minimum correlation coefficient allowed

Enter a correlation coefficient threshold value. The default

value of 0.0 results in no editing based on the correlation
coefficient. Values greater than 0.5 will result in strong
editing. Correlation coefficients with absolute values less
than the submitted threshold value will result in the output
gradient values at those time samples being zeroed.
Note: This parameter only has effect if the input data come
from AVO Attribute Stacks, not from AVO Weighted Stacks.

Maximum runs test statistic allowed

Enter a runs statistic threshold value. The default value of

100 results in no editing based on the runs statistic. Values
greater than 5 will result in little editing. Values less than 3
may result in strong editing. Editing with such low runs
statistics may result in valid data being removed due to
expected variations in the distribution of seismic amplitudes.
Runs statistics with absolute values less than the submitted
threshold value will result in the output gradient values at
those time samples being zeroed.

Note: This parameter only has effect if the input data come
from AVO Attribute Stacks, not from AVO Weighted Stacks.

Correlation coefficient scaling (0 to 1)

Enter a value between 0 and 1.0 to control the weighting of

the intercept and gradient data based on the input
correlation coefficient attribute (AVO_ATTR=7). A value of 0.0
results in no scaling while a value of 1.0 results in complete
scaling. The scaling function for the input scalar S and
correlation coefficient C is:
Scalar = ( 1.0 – S ) + S × C

Other Docs Search Page Known Problems

AVO Attribute Computations2133 ProMAX® Reference

Note: This parameter only has effect if the input data come
from AVO Attribute Stacks, not from AVO Weighted Stacks.

Note: Interval velocities in time are required. The velocity

conversion should be done using Velocity Manipulation. It
is strongly recommended that the interval velocities be
carefully examined before attempting AVO analysis.

Runs test statistic scaling (0 to 1)

Enter a value between 0 and 1.0 to control the weighting of

the intercept and gradient data based on the input runs test
attribute (AVO_ATTR=8). A value of 0.0 results in no scaling
while a value of 1.0 results in complete scaling. The scaling
function for the input scalar S and runs statistic R is:
S
Scalar = ( 1.0 – S ) + -------------------------
( R + 1.0 )

Note: This parameter only has effect if the input data come
from AVO Attribute Stacks, not from AVO Weighted Stacks.

Other Docs Search Page Known Problems