Abstract
Summary
Compressive sensing (CS) of seismic data is a new style of seismic acquisition whereby the data are recorded on a pseudorandom grid rather than along densely sampled lines in a conventional design. A CS design with a similar station density will generally yield better quality data at a similar cost compared to a conventional design, whereas a CS design with a lower station density will reduce costs while retaining quality. Previous authors (Mosher, 2014) have shown good results from CS surveys using proprietary methods for the design and processing. In this paper we show results obtained using commercially available services based on published algorithms (Lopez, 2016). This is a necessary requirement for adoption of CS by our industry.
This report documents the results of a 108km2 CS acquisition and processing trial. The acquisition and processing were specifically designed to establish whether CS can be used for suppression of backscattered, low velocity, high frequency surface waves. We demonstrate that CS data can be reconstructed by a commercial contractor and that the suppression of backscattered surface waves is improved by using CS receiver gathers reconstructed to a dense shot grid. We also show that CS acquisition is a reliable alternative to conventional acquisition from which high-quality subsurface images can be formed.
Research on Seismic Data Interpolation and Reconstruction