Bayesian framework to wavelet estimation and linearized acoustic inversion
L. P. Figueiredo; M. Santos; M. Roisenberg; W. Figueiredo
Seismic inversion is an important tool widely used in Geophysical problems to infer the subsurface properties through seismic wave measurements. In particular, it can improve exploration and management success in the petroleum industry, since it estimates the elastic properties from the seismic data, which has a great correlation with many petrophysical properties. The major challenge of seismic inversion method is to integrate all different kinds of data in order to obtain an accurate and high-resolution set of subsurface parameters, also characterizing the uncertainties of the results of the inversion.
In this work, we show how a seismic inversion method based on a Bayesian framework can be applied on post-stack seismic data to estimate the wavelet, the seismic noise level and the subsurface acoustic impedance. We propose a different linearized forward model and discuss in detail how some stochastic quantities are defined in a geophysical interpretation. The forward model and the Gaussian assumption for the likelihood distributions enable to obtain the conditional distributions. The method is divided in two sequential steps: the wavelet and noise level estimation, in which the posterior distribution is obtained via a Monte Carlo method (Gibbs sampling algorithm), and the acoustic inversion, which uses the proposal forward model and the results of the first step. In the second step, the posterior distribution for acoustic impedance is analytically obtained. Therefore, the maximum a posteriori impedance can be calculated, yielding a very fast inversion algorithm.
In the wavelet estimation, the method presents good results without any assumption about the wavelet phase. The estimated wavelet, and the mean value of noise level yield good results when used as fixed quantities in the proposed acoustic inversion method. The methodology appears to be a good choice for acoustic inversion, due to the possibility of integrating prior knowledge, as the low frequency model, seismic correlation function and wavelet characteristics, in the inversion results conditioned to seismic traces. The similarity between the MAP solution and the Constrained Sparse Spike Inversion indicates the feasibility and reliability of the proposed method. The covariance matrices calculations of the multivariate Gaussian distributions are important procedures to incorporate results of the prior knowledge into the inversion, which directly influences its quality.