Uncertainty is an inherent problem existing in velocity analysis. It is important for geophysicists to assess the variability of the velocity quantitatively. As an alternative to a common geostatistical method (), Clapp introduced multiple realization method for complex operators. Clapp modified the standard geophysical inversion technique by adding random noise into the model styling goal to achieve multiple realizations. By comparing and contrasting the equal-probable realizations, the variability can be evaluated. Since the subsurface image is obtained based on the new velocity model, the uncertainty of velocity model will cause the uncertainty of amplitude information we can acquired from image (). Using the multiple realization method, Clapp showed how the velocity uncertainty affected the amplitude information.
Amplitudes carry important information about rock properties. Amplitude variation with offset (AVO) is a widely used technique in petroleum industry because AVO anomalies often indicates hydrocarbon existence. A good review of AVO analysis is provided by Castagna . Since AVO is dependent on intrinsic rock parameters such as compressional-wave velocity, shear-wave velocity, density, anisotropy and attenuation, AVO can be used to assess information for rock properties, such as lithology, porosity and pore fluid content. Castagna provide a rock physics framework for AVO analysis.
The relationship between AVO and rock properties make us guess there may exist empirical relationships between AVO uncertainty and rock information. For example, if we get high variance of AVO attributes (which can be evaluated from multiple realizations) at specific subsurface areas, we can conjecture that there may be some change in rock information in the same area, such as impedance, velocity or shale/sand ratio. In this paper, we explored such relationships.
Instead of extracting amplitude variations with offset, we adopted amplitude variation with angle (AVA) analysis because realistic velocities usually break the simple relationship between offset and angle. The dataset we used was from South America. We evaluated the variability of AVA attributes by using a 3-D histogram. A 2-D section was extracted and shale volume along the wells in this section were calculated. We found the well with low shale volume has obvious higher AVA uncertainty than other two wells, which made us conjecture the low shale/sand ratio will cause high AVA uncertainty. The further work need to be done is to use more real data to exam whether our guess is true or there exist other empirical relationships between AVA uncertainty and rock information, such as shale volume, impedance or velocity.