Conclusions

We have introduced a method of separating signal and noise and interpolating irregularly sampled data from regional scale teleseismic recordings using application of industry standard filtering algorithms based on the linear radon transform. With this method, we exploit the plane wave nature of much of the teleseismic wavefield useable for creating structural images of the lithosphere. Arrivals not following plane wave moveouts, within a desired dip range, are easily identified and removed from the recorded wavefield. Additionally, the linear radon transform provides a method for interpolation of planar arrivals that reside within the desired dip range.

We have shown the method to be successful in interpolating real data and synthetic data with a real acquisition geometry created with a complicated wavespeed model. In addition, we have shown this method to be effective at isolating dips differing from the laterally coherent scattering and source related energy even in the presence of severe under sampling. We have demonstrated that our method is more effective than f-k filtering because it has superior dip resolution, allows the separation of dips as a function of time, and can handle (and interpolate) irregular wavefield sampling. Besides the algorithm development, the most important aspect of this study has been the recognition of the effect of current standard teleseismic wavefield acquisition geometries on dip filtering as well as standard seismic imaging methods. We hope that the examples shown in this paper will motivate denser deployments in the future to allow proper pre-processing and insure more accurate images than are presently possible.