Rob Hegge

CFP Operator Estimation And Inversion Demonstrated On a Field Data Set - Part I: Operator Updating

Summary For a reliable image of the subsurface correct focusing operators need be used. Traditionally these are calculated from the velocity-depth model. In the CFP approach they are obtained directly from the seismic data by (iteratively) updating the initial (hyperbolic) focusing operator using the principle of equal travel time. The updating can be done for a random sequence of focal points lying on different boundaries. There is no layer stripping involved. The updating procedure is already shown to work successfully on several complex synthetic data sets ([3], [7], [4] and [6]). In this paper the updating results on a real North Sea data set containing a saltdome are presented. The results show that the operator approach is valid even for focal points below the salt.

Estimation of primaries by sparse inversion using dual-sensor data

Summary Most wave-equation based multiple removal methods, such as the surface-related multiple elimination (SRME) method, are based on prediction and subtraction of the multiples. Although this approach has proven successfully for many cases, it still shows some drawbacks such as handling missing near-offsets and distortion of primaries during the subtraction process. The recently introduced estimation of primaries by sparse inversion (EPSI) method estimates the primaries as unknowns in a multi-dimensional inversion process, where primaries are parameterized by spikes and a corresponding wavelet. Furthermore, it can reconstruct missing near offset data simultaneously. In this paper the EPSI method is extended to handle time-variant wavelet variations in the parameterization, which is necessary to obtain good results on field data. Because of the strong physical constraint between primaries and multiples, EPSI requires the upgoing wavefield at the surface. Therefore, it is very well suited for field data from a dual-sensor measurement, which provides a reliable upgoing wavefield.

Using focusing operators to estimate velocity models

Focusing operators are derived in the common focal point (CFP) formulation of prestack migration. Each operator corresponds essentially to the time-reversed Greens function in the actual medium, when the source is placed at the location to be imaged (the focal point). By applying these operators twice to the data, the downand upward propagation effects are removed and thus a migrated image is obtained. However by using the traveltime information in these operators we can also estimate a layered macro velocity model. The estimation is not limited to homogeneous layers and all layers can be estimated simultaneously. Several aspects, including uncertainty, with regard to this process are considered and illustrated with a number of examples.

3D SRME: Acquisition & processing solutions

This paper describes the issues and possible solutions involved in the application of 3D data-driven multiple removal in a marine production environment. The optimization of marine data acquisition for 3D Surface related multiple elimination (SRME) is discussed. Recommendations for acquisition are given to create a dense grid of sail lines and streamers needed to predict the full 3D characteristics of the multiples. A processing sequence is discussed to predict and remove 3D multiples for each sail line using only the recorded data around the output streamer. The processing strategy does not rely on any a priori information or model of the subsurface to remove 3D multiples. The application of 3D SRME to a field data set from the Norwegian Sea leads to results that could not be obtained using its 2D equivalent.

Application of 3-D SRME And Multiple Diffraction Removal In the Makassar Straits, Indonesia.

It is recognized that 3-D related noise problems and multiple attenuation in particular can be benefit from currently non-standard acquisition techniques, like, for instance, multi-azimuth (Widmaier et al. 2002) and streamer overlap (van Borselen et al. 2005). However processing solutions are still required for historical and new standard marine acquisition datasets. Especially for the application of 3-D Surface Related Multiple Elimination (SRME), the issues of limited crossline sampling and aperture have to be addressed. A variety of methods for dealing with these issues have been proposed in the literature, which can be broadly categorized into a) a full wavefield interpolation or reconstruction, followed by a large number of 3-D inline convolutions to create a dense grid of multiple contributions (MCs) and finally summation in the crossline direction of those MCs (see, for instance, Baumstein and Hadidi, 2004), or b) calculation of 3-D inline convolutions to create MCs wherever possible using the existing data followed by sparse inversion in the crossline direction (van Dedem and Verschuur, 2002; Hokstad and Sollie, 2003). The hybrid approach described in van Borselen et al. (2005) consists of a more limited sailline regularisation and reconstruction in the crossline direction and sparse inversion of the subsequently increased number of crossline MCs. Major advantages are the reduced number of 3-D inline convolutions that need to be calculated compared to full reconstruction approaches, while with more MCs the sparse inversion becomes easier to parameterize and provides better results. This 3-D prediction approach is used in the current case study.