I propose to use the comprehensive 3-D Amoco data set to estimate reservoir
mechanical properties and their spatial variation in a true 3-D sense.
This procedure will require sequential data processing steps involving:
(1) *P* and *S* velocity analysis from 3-D and 2-D reflection data,
(2) 3-D prestack migration images of reservoir structure, stratigraphy
and fault geometry, (3) 3-D simultaneous prestack inversion of the
and reflection data for *P* and *S* relative impedance
contrasts within the reservoir zone, and (4) integration of the *P*
and *S* velocity and impedance contrast information to make estimates
of the *P* and *S* impedances *I*_{P} and *I*_{S},
in a 3-D spatial volume encompassing the reservoir block.

The most significant portion of this research effort is directed at the
problem of estimating impedance contrasts from seismic reflection data.
I will approach this problem in the following manner. Using a reliable
estimate of the *P* and *S* migration velocity models resulting from
the velocity analysis step (1), multiple constant offset elastic migrations
will be performed for each of the and data sets.
This procedure will result in four sequences of maps: ,, and , where
*R* represents the angle-dependent elastic reflectivity estimate,
and represents the reflection angle estimate. The *R* and values make a direct mapping to and reflectivity as a function
of reflection angle, which can subsequently be inverted for *P* and *S*
impedance contrasts, using the Zoeppritz plane-wave response, as a function
of subsurface position. The theory for making these reflectivity and angle
estimates will be based on elastic Kirchhoff integrals and ray-theoretic WKBJ
Green's functions (Lumley and Beydoun, 1991; and Lumley 1992, 1993a and 1993b.)

Once the impedance contrasts are estimated, they can be linearly combined
with the *P* and *S* velocity estimates to make broadband impedance estimates.
Spatial variations in *I*_{P} and *I*_{S} are useful in that they have
physical correlations
with: presence and saturation of gas in pore space (*I*_{P} low, *I*_{S} normal
to high), fractured high
permeability zones (both *I*_{P} and *I*_{S} are low), and can be combined
to estimate 3-D variations in Poisson's Ratio, which can be a good
discriminator of lithology (e.g., sandstone vs. shale vs. coal).
I propose to correlate my *I*_{P} and *I*_{S} results with Amoco borehole,
rock physics, and production engineering data. This correlation will
be at the very least a qualitative integration of these multi-disciplinary
data types, and may possibly extended to a quantitative integration using
existing Markov-Bayes methods of spatial geostatistics.

11/17/1997