A Madagascar package for deblending in multiple flavours

Daniel O. Trad

Data acquisition is the most expensive part of seismic methods and the need to decrease its costs leads to cost-effective compromises. In particular, 3-dimensional data surveys always have deficient sampling in at least 2 of the 4 spatial dimensions. Offset and azimuth dimensions are always undersampled and migration algorithms produce suboptimal angle and offset gathers, which affects AVO and AVAz studies. As a consequence, geophysicists make extensive efforts in the mitigation of these sampling problems. These efforts usually involve two directions: data interpolation and simultaneous acquisition. Interpolation is intended to create new seismic traces from the acquired samples by using sparse transformations. Simultaneous acquisition, also known as blending, attempts to mitigate the sampling problem by acquiring more data in less time by recording several shots simultaneously. Its main difficulty is the processing of the resulting seismic data, which requires either shot separation very early in the signal processing chain or the development of new processing/inversion capabilities that can work directly with blended data.

There are several approaches for deblending seismic data: denoising, inversion and physical transforms. From a general point of view, the three are closely related but they differ in how and which transformations are used. In this paper, I will discuss the relationships between these approaches from an algorithmic point of view.