Recent developments - 3-D Seismic Interpretation - page 218

Fig. A3.10 Section through a Poisson's ratio volume showing (arrowed) a zone of low Poisson's
ratio which corresponds to a hydrocarbon-bearing sand mound within a shaly sequence between the
two picked horizons. Colour scale: magenta/red = low Poisson's ratio, green/cyan = high Poisson's
ratio.

(1) Events need to be aligned on the same TWT across the offset range. This is of course the objective
of NMO removal prior to stack. However, the requirements are more stringent than for normal
stack, where a small misalignment causes only a little blurring in the summed trace. This implies
very careful velocity analysis, probably including allowance for non-hyperbolic moveout.
(2) Attention needs to be given to the difference in frequency content between near-offset and far-
offset traces. This is partly a consequence of the longer travel path at far offsets and partly a
result of NMO stretch, the distortion produced in the trace as it is shifted by an amount varying
with TWT to correct for NMO. In both cases the effect is that the frequencies present are lower
at the far offsets. It is a good idea to try to minimise the change in frequency with offset by
filtering and deconvolution. A complementary approach is to invert using an offset-dependent
wavelet.
(3) Amplitude variation with offset needs to be free from artefacts of acquisition or processing. This
is an issue for AVO studies in general, of course. Where good-quality well log data are available,
it is useful to create an offset synthetic and compare the way that amplitude varies with offset
between synthetic and real data; it may then be possible to apply an offset-dependent gain to the
real data in order to make it resemble the synthetic more closely.
As well as choosing a

χ

value in an empirical way to match a desired target log, there are some
theoretical results that offer guidance on what information is available at different angles (Whitcombe
et al. , 2002) . With reasonable assumptions on velocity-density and V p − V s relations, then bulk
modulus is proportional to elastic impedance at χ = 12 ◦ , and shear modulus to elastic impedance

at 51 ◦ ; shear impedance (product of shear velocity and density) can be found from elastic

impedance at χ =− 45 ◦ , and V p / V s from χ = 45 ◦ . From the latter we can also calculate Poisson's
ratio. The significance of this is that bulk modulus is sensitive to fluid content whereas shear modulus
is not. Thus a bulk modulus volume is a good place to look for direct hydrocarbon indicators (DHIs)
caused by reduction in modulus by the presence of oil or gas instead of brine; a shear modulus or shear
impedance volume will show the effects of lithology but not porefill. A real DHI would therefore be
expected to show anomalous low values at χ = 12 ◦ but not at χ =− 45 ◦ . Another useful lithology
indicator is a Poisson's ratio volume; as we have seen in chapter 5 , low values for Poisson's ratio are
often a useful indicator of high sand/shale ratio. An example of a Poisson's ratio volume is shown in
fig. A3.10 .

χ =−