De-noising of marine seismic data

Abstract

Marine seismic acquisition represents one of the most used geophysical exploration techniques employed in the petroleum industry today. However, one major challenge is that the meaningful part of the data will be distorted by a certain amount of noise originating from various sources.

This thesis will look for an optimized de-noising flow of a marine seismic line that was discarded (scrapped line) because the noise threshold values were considered to be too high. This line was then reacquired (reference line) and is going to be processed along with the discarded line and simultaneously serves as a benchmark. The main objective of this work is to see if it is possible to raise the quality of the scrapped data during processing so that it resembles the quality of the reference data.

Since seismic processing techniques have evolved significantly within the last decades, it might thus be acceptable to acquire data in rougher weather conditions. Accordingly, the noise threshold values could be adjusted.

After extensive testing, an optimized de-noising combination was identified. When applied to the scrapped line as well as the reference line, very similar results were obtained. Both visual inspection and calculated RMS values have been taken into account to assure the quality of the final results. These observations support the basic idea of accepting more noise in future marine acquisitions, due to advances in seismic processing (e.g. de-noising).