Nearly Constant Amplitudes with Offset

Here I analyze the second amplitude effect observed along the BSR: nearly constant amplitudes with increasing angle. A representative reflectivity gather and the picked BSR amplitude trend are shown in Figure . Since some residual amplitudes are left at the central offsets between $16^{^{\circ}}$ and $20^{^{\circ}}$ (Chapter 2, section 2.4.3), those amplitudes are edited manually to be in better agreement with the general amplitude trend.

 

gather3
Figure 19 Reflectivity gather shown on the left panel and the picked BSR AVO trend on the right panel. The amplitudes are nearly constant with increasing offset.

As described before, I determine the initial velocity models by averaging the interval velocity determined in Chapter 2 and assuming a Poisson's ratio of 0.45, representative of brine-saturated sediments. This initial velocity model is shown in Figure .

 

vint3
Figure 20 Initial velocity models across the BSR for the case of constant amplitudes with angle.

The synthetic AVO curve corresponding to this velocity model is again calculated using the Zoeppritz equations. It is compared with the AVO trend observed in the data in Figure . The comparison shows that this first velocity model fails to reproduce the zero-offset reflection coefficient, while the general AVO trend appears to be in fairly good agreement. This already suggests that this AVO trend might not represent a hydrate-over-gas model.

 

fit3
Figure 21 Synthetic AVO curve obtained from the initial velocities (solid line) compared with the AVO trend observed in the data (*).

In order to fit the data, the P-wave velocity across the BSR must be increased and the S-wave velocity contrast slightly decreased (Figure ). The result is similar to the P-wave velocity contrast described in the previous section: the hydrate velocity increases to approximately 2.07 km/s while the velocity in the underlying sediment decreases to about 1.55 km/s. The slight decrease in S-wave velocity contrast results in a slightly smaller Poisson's ratio of about 0.43 in the sediments below the hydrate, compared to a value of 0.45 in the hydrated sediment itself. This small decrease in Poisson's ratio is consistent with a hydrate-over-brine model, and may be explained by the heterogeneity of the sediments, which can cause not only variations in P-wave velocity but also in Poisson's ratio.

The resulting AVO curve is shown in Figure . The comparison of the modeled AVO curve with the one observed indicates that this model successfully reproduces the zero-offset reflection coefficient and the AVO trend. This good agreement suggests that there exists a small decrease in S-wave velocity across the BSR. This indicates that the model of hydrate over brine-saturated sediment would be required to reproduce the nearly constant amplitudes with increasing offset.

 

vint3-fit
Figure 22 Interval velocities in the case of slightly decreasing Poisson's ratio in the sediments beneath the hydrate. The solid line represents the initial velocities. The arrow indicates if the modeled velocities have to be increased or decreased.

 

fit3-fit
Figure 23 Synthetic AVO curve obtained from the interval velocity (solid line) compared with the amplitudes observed in the data (*).

The analysis of the nearly constant amplitude trend with increasing angle indicates that the BSR amplitudes require a significant negative P-wave velocity contrast and a small negative S-wave velocity contrast. Based on the fairly unchanged Poisson's ratio at the transition, this would suggest a hydrate-over-brine sediment model for these amplitude variations. However, the significant decrease in the velocities underneath the hydrate, consistent with drilling results and similar seismic analyzes in the region, strongly suggests the presence of free gas Holbrook et al. (1996); Katzman et al. (1994); Lee et al. (1994); Matsumoto et al. (1996); Wood et al. (1994). Consequently, two interpretations are possible:

• brine- and gas-saturated zones are distributed irregularly underneath the BSR, and the amplitude trend does in fact correspond to a hydrate-over-brine region.
• thin layering underneath the BSR distorts the amplitude response, causing the amplitude to resemble a hydrate-over-brine trend instead of a hydrate-over-gas trend.

A heterogeneous distribution of gas and brine saturation is physically plausible in marine sediment. The gas can be trapped in thin layers that serve as permeability barriers, yielding a patchy gas distribution Dillon et al. (1993). However, the resolution of localized patches of brine and gas saturation from surface seismic is highly questionable. Therefore, it is more reasonable to assume that the constant amplitudes result from thin layer tuning effects, rather than patchy brine saturation beneath the BSR.