Estimating Seismic Uncertainity in Model Space

We present a new method to estimate the uncertainty in models (velocity, density, anisotropy, etc.) used to image seismic data.

We achieve this objective by calculating the geologically significant principal components of both data and the models to estimate where the model does or does not represent the data and how large the error is.

This method is tested on synthetic data generated over Marmousi model. We plot the uncertainties for two variations of Marmousi model and show how these uncertainties can be used in exploration processes.