Using 2D Ring Arrays to Remove Back-scattered Surface Noise from Land Seismic Data

Abstract

Surface noise often dominates signal in land seismic acquisition by a large amount, perhaps 2x to 20x, which limits the usefulness of land seismic data, especially converted waves and possibly attributes. Much of this noise is also back-scattered. We propose an approach of reducing back-scattered surface noise by designing field acquisition to better sample the noise in X & Y so that it can be more easily removed in processing. The key concept is that deliberate 2D arrays that sample the noise can reduce noise with an effectiveness that scales by an array element count of N or greater while the approach of stacking randomness using large fold reduces noise by √. The difference is large when N is about 9 or greater. The challenges in making this new approach work are designing the 2D arrays, careful inverse filtering of the array data, handling a wide frequency range, and addressing irregular illumination in processing.