Local BSR AVO Inversion

After qualitatively examining the BSR amplitudes in several gathers, I quantify the actual AVO responses along the BSR by first picking the maximum (negative) amplitude along the BSR, and then inverting locally for the P- and S-impedance contrasts. In this way, I minimize errors that might have been introduced by the 2-D inversion due to uncorrected residual moveouts. Furthermore, this analysis results in a direct insight into the transition from hydrate-bearing sediment to the sediment underneath and the possible effect that tuning might have on the resulting impedance contrasts.

The analysis in the previous section suggested that between 25 and 38 km lateral distance the BSR might be underlain by a thin layer. Thus, the BSR amplitudes might suffer from tuning effects. Therefore, I examine only the BSR amplitudes between 40 and 52 km lateral distance.

 

bsr
Figure 9 BSR P-(o) and S-impedance (*) contrasts.

The resulting P- and S-impedance contrasts across the BSR can be seen in Figure , with P-impedance contrast plotted as (o), and S-impedance contrast plotted as (*). The local impedance contrasts display well-constrained, negative P-impedance contrasts, as already expected. However, at about 41.8 km and 44 km, the P-impedance contrasts appear ``spiky''. Looking back at Figures , , and , this behavior seems to be associated with the structure underneath the BSR that is interfering with the BSR response. Therefore, possible tuning effects might account for the anomalously large P-impedance contrasts.

The S-impedance contrast across the BSR displays a laterally more scattered response. Since the limited angle coverage of the data causes the inversion for changes in S-impedance to be less well constrained, more variance can be expected. Most of the S-impedance contrast appears to be positive, but there is significant lateral heterogeneity. In areas of structural interference, the S-impedance contrast is strongly negative, i.e. at 42 km and 44 km. As mentioned before, this amplitude anomaly is most likely based on wavelet interference effects.

The actual BSR AVO function that produced these impedance contrasts is shown in Figure . The figure displays the BSR amplitude difference between the near-offset traces and the far-offset traces. A positive difference means that the amplitudes at the BSR are increasingly negative with increasing offset, while a negative difference represents decreasing amplitudes with increasing offset. The zero crossing (constant amplitudes with increasing offset) is represented by the solid line. As expected from the P- and S-impedance contrasts, the AVO behavior shows decreasing amplitudes with offset at 42 km and 44 km, as well as at about 46.5 km and after 50 km. These regions probably represent a superposition of structure and BSR amplitudes. The BSR AVO behavior shows mostly positive amplitude differences, thus indicating increasing amplitudes with increasing offset, or regions where the difference becomes fairly small, suggesting nearly constant AVO in these regions. Excluding the very apparent tuning effects at 42 km, 44km and 46 km, the lateral BSR amplitudes thus display two different trends:

 

offset
Figure 10 Near - Far offsets of the BSR amplitudes. Positive amplitude difference indicates increasingly negative amplitudes with increasing offset. Negative amplitude difference represents decreasing amplitude with increasing offset.

• increasingly negative amplitudes with increasing offset
• nearly constant amplitudes with constant offset

The increasingly negative amplitudes with offset represent the positive S-impedance contrasts, while the constant amplitudes with offset yield small positive or negative contrasts, depending on the density structure. Figure shows that these amplitude variations probably predict the existence of both brine- and gas-saturated sediment underneath the BSR. However, thin layering might be distorting the BSR amplitudes. Increasingly negative amplitudes with increasing offset have been attributed to the BSR appearance in the region of the Blake Outer Ridge Katzman et al. (1994); Lee et al. (1994). Furthermore, Ostrander 1984 has shown that a thin gas layer can cause amplitudes to be diminished, thus causing either decreasing or constant amplitudes with offset. Therefore, the regions displaying constant amplitudes with offset and thus appearing to indicate the presence of hydrate-bearing sediment overlying brine-saturated sediment might also be caused by tuning effects. AVO modeling of the two different AVO responses observed in the data should help us to distinguish these effects.