Seismic inversion is a quantitative analysis technique in reservoir geophysics to
reveal subsurface physical properties from surface-recorded seismic data. But the
most widely used inversion in oil and gas exploration for decades is PP-wave based.
P-to-S converted wave, which has shown great success in the imaging of gas clouds,
has a different response to rocks and pore-fluids from the PP-wave. A joint use of the
PS-wave and PP-wave in the inversion can reduce the ill-posedness of the inverse
problem and in particular enables simultaneous inversion for three independent elastic
parameters.
Conventionally, prestack seismic inversion is based on the incidence
angle-dependent reflection coefficients. In my research, I define the seismic
reflections and impedances along the ray paths of wave propagation, and these ray
paths obey Snell’s law. I adopt the ray-impedance concept, which is a
frequency-dependent parameter and is sensitive to fluid contents. Joined interpretation
of PP- and PS-wave ray impedances can identify reservoirs, and also has potential in
fluid discrimination.
Joint inversion of PP- and PS-waves is performed on the constant ray parameter
(CRP) profiles. For a constant ray parameter, a pair of PP- and PS-wave traces has
exactly the same ray path between the source and the reflection point, which means
the PP- and PS-wave reflection events represent exactly the same reflection point, in
the horizontal direction. Therefore, PP and PS-wave calibration transforms PS-wave
reflection events from PS-wave time to the corresponding PP-wave time, and
reflections events in a pair of PP- and calibrated PS-wave traces with a constant ray
parameter should correspond to each other, sample by sample, both horizontally and
vertically. I also present a procedure which preserves the original wavelets in the
transformed PS-wave trace.
I use a bending ray-tracing method to construct the common image point (CIP)
gathers in the ray-parameter domain. I estimate mixed-phase wavelets for each
constant ray-parameter (CRP) profile through a frequency domain high-order
statistical method, and then invert for the reflectivity series using weighted constraints.
From the reflectivity sections, I estimate PP- and PS-wave ray impedances separately
and also estimate three elastic parameters simultaneously in a joint inversion.
I have applied the entire procedure to a couple of field data sets, to verify the
robustness and effectiveness of the method, and to demonstrate the great potential of joint inversion in ray-parameter domain.