Land seismic data is known to be far noisier than its' marine equivalent - the questions we want to begin to address are why and is there anything that can be done aboutit? Noise issues with land data are often just explained by `poorcoupling',whateverthatmaymean.Webelievethat the main issue is not the connection between the source/receiver and the subsurface, but what happens to theenergyinthetop10-100mduringboththedowngoing andupgoingphases. Modern dense acquisitionclearlyhelps,but does not solve the noise on its own, especially where micro-scattering noise is concerned. However, when this modern acquisition is coupled with new powerful noise removal processingtechnology, noisesparsityallowsustohandle it more effectively. This is especially important for shallow scattering noise, which is difficult to sample unaliased in X & Y. The new algorithms use the same principals as compressive sensing, utilizing sparsity to handle aliased noise, so optimizing acquisition for these algorithms is, effectively, compressive sensing for the noise.

Open-Access Online Publication: March 01, 2023