Increasing Amplitudes with Offset

Here I analyze one of the two amplitude effects observed in the data: increasing amplitudes with increasing offset. Figure shows a reflectivity gather and the corresponding BSR AVO trend picked along the BSR reflection. The reflectivity gather displays a fairly well-resolved BSR wavelet. The AVO trend shows clearly the increasingly negative amplitudes with increasing angle. The offsets were converted into angles by using the reflection angles at the BSR, which were computed by the prestack migration (see Chapter 2). For incidence angles between $17^{^{\circ}}$ and $18^{^{\circ}}$, the amplitudes are anomalously low as a result of uncorrected amplitude effects as described in Chapter2, section 2.4.3 (i.e. hydrophone array attenuation at central offsets). Therefore, these two points are given small weights in the following AVO modeling procedure.

 

gather1
Figure 12 Reflectivity gather shown on the left panel and corresponding BSR AVO trend on the right panel. The amplitudes are increasingly negative with increasing angles.

The initial P- and S-wave velocity used in the modeling are shown in Figure . The P-wave velocity above and below the BSR is calculated directly by averaging the interval velocity obtained from velocity analysis. The S-wave velocity is determined by assuming a Poisson's ratio of 0.45, which is consistent with the brine-saturated, highly unconsolidated sediments typical of this region.

In the first attempt to model the observed AVO amplitudes, these initial velocities are used as input velocities. The resulting AVO curve is obtained using the Zoeppritz equations and is compared with the one observed in the data (Figure ). The comparison of both curves shows that with nearly constant amplitudes with increasing angles, the initial velocity model fails not only to reproduce the zero-offset reflection coefficient, but also to reproduce the general AVO trend. Assuming negligibly small density contributions, the near-offset amplitudes are mainly dependent on the P-wave velocity contrast at the reflector, while the AVO trend is characterized primarily by the S-wave velocity contrast. Thus, the AVO response resulting from the initial velocity model implies the use of both incorrect P- and S-wave velocities at the BSR.

 

vint1
Figure 13 Average interval velocities across the BSR for the case of increasing amplitudes with angle.

 

fit1
Figure 14 AVO curve obtained from the initial P-wave velocity and assumed brine Poisson's ratio (solid line) compared with the one observed in the data (*).

Based on this result, the subsequent modeling attempts to increase the P-wave velocity contrast across the BSR in order to recreate the observed zero-offset reflection amplitudes. The required increase is obtained by increasing the velocity in the hydrate layer and simultaneously decreasing the velocity in the layer underneath (Figure . This yields a thinner hydrate layer over brine sediment. The S-wave velocity is again determined using a Poisson's ratio of 0.45.

 

vint1-vp
Figure 15 Interval velocities across the BSR for a thin hydrate-over-brine model in the case of increasing amplitudes with angle. The solid line represents the initial velocities. The arrows indicate if the modeled velocities have to be increased or decreased.

 

fit1-vp
Figure 16 AVO curve obtained by increasing the P-wave velocity contrast across the BSR (solid line) compared with the one observed in the data (*).

The resulting AVO trend is shown in Figure . The comparison of the modeled AVO response with the observed response indicates that this model can successfully reproduce the zero-offset data. This suggests a P-wave velocity of 2.05 km/s in the hydrate and 1.58 km/s in the underlying sediments might resemble the actual conditions at the BSR. However, the calculated AVO trend is still contrary to the observed one, displaying nearly constant amplitudes with increasing angles. Hence, a change in Poisson's ratio seems to be required at the transition from hydrate-bearing sediments to the sediments underneath.

The observed AVO trend of the data suggests that the hydrate-bearing sediment has a higher Poisson's ratio than the sediment underneath (see Figure ). Since it does not make sense physically to increase the hydrate Poisson's ratio above 0.45 (fluids have Poisson's ratios of about 0.5), I decrease the Poisson's ratio in the layer underneath to simulate the drop in Poisson's ratio. This yields the same P-wave velocities as described before, but an increase in S-wave velocity of about 0.25 km/s across the BSR (Figure ). The resulting velocities represent a Poisson's ratio of 0.45 in the hydrate and approximately 0.28 in the sediments underneath. Instead of keeping the hydrate Poisson's ratio constant, I could have increased the S-wave velocity there as well. This would have yielded a lower Poisson's ratio in the hydrate and would have required an even lower Poisson's ratio than 0.28 in the sediments underneath. Nonetheless, both cases would require a strong increase in S-wave velocity across the BSR. Since the drilling at the Blake Outer Ridge Matsumoto et al. (1996) has shown that the overall shear structure in the sediments is very weak, increasing the model S-wave velocities too much would generate conditions that do not resemble in-situ conditions.

 

vint1-fit
Figure 17 Interval velocities for hydrate-bearing sediments overlying gas-saturated sediments. The initial velocities are represented by the solid lines. The arrows indicate whether the modeled velocities have to be decreased or increased.

 

fit1-fit
Figure 18 AVO curve obtained by the hydrate-over-gas velocity model (solid line). It is compared with the AVO behavior observed in the real data.

The significant drop in P-wave velocity at the BSR as well as the decrease of Poisson's ratio suggests the presence of free gas underneath the hydrate layer. Domenico showed that the presence of gas can cause a decrease in Poisson's ratio down to 0.1 Domenico (1976). However, because the sediments at the Blake Outer Ridge are highly unconsolidated and have shale contents of more than 50% Matsumoto et al. (1996), I would not expect the Poisson's ratio to drop as low as 0.1 in the presence of free gas.

A comparison of the synthetic AVO curve obtained from the model in Figure with the AVO trend observed in the data is shown in Figure . The synthetic curve agrees well with the real data for both near and far offsets. Thus, a strong increase in S-wave velocity and a simultaneous decrease in P-wave velocity at the transition from hydrate-bearing sediments to sediments containing free gas is required to explain the seismic data. The inferred velocity contrasts are, furthermore, in good agreement with the prediction based on the negative P-impedance contrast and positive S-impedance contrast obtained from the impedance inversion.