Monte Carlo Sampling of Solutions to Velocity Tomography Problem: A Quest for Imaging Accuracy

The main drawback of the classical approach to seismic tomography is a lack of the robust method to estimate the accuracy of tomographic image. This difficulty can be overcome if the Bayesian inverse theory is applied to solve the tomography inverse problem. The method relies on the construction of the posterior probability over the space of all possible velocity images which gives the probability that a given model is the true one. Thus, it provides the mechanism to find not only estimators of the true velocity distribution, like the Maximum Likelihood or average models but also, to estimate the accuracy of the tomography imaging. However, as the tomography problems are highly dimensional, inspecting posterior probability requires Monte Carlo sampling. This paper presents the application of this technique to analysis of the tomography imaging accuracy.