Gijs J. O. Vermeer

Reciprocal Offset-vector Tiles In Various Acquisition Geometries

With the recent developments in marine streamer acquisition, all three main types of acquisition geometry – orthogonal, areal, and parallel – can now be used to acquire wideazimuth seismic data. On top of this, marine streamer acquisition may also be used to acquire multi-azimuth data. In this paper these geometry types are compared with each other on basis of offset-vector tile (OVT) gathers. Contrary to popular belief that reciprocal azimuths do not contribute to better illumination, reciprocal OVTs are highly desirable to compensate for coarse sampling. These reciprocal OVTs are not always integrated in the design criteria of wide-azimuth streamer acquisition. Instead, the parameters of this acquisition technique tend to suffer from serious compromises.

Processing orthogonal geometry — what is missing?

The differences in properties between parallel geometry and orthogonal geometry call for different approaches to various prestack processing steps. Indeed, a number of authors have published results of prestack processing of orthogonal geometry data on basis of geometry-oriented approaches, such as 3Dfk-filtering of cross-spreads. Yet, this treatment of orthogonal geometry is far from being standard in the industry.

Alternative strategies for tackling scattered noise

Scattered ground roll is one of the most serious noise problems in the Middle East and similar high-velocity areas. Meunier has shown that the noise of point scatterers can best be discriminated from the desired signal in 3D shot or 3D receiver gathers, whereas part of that noise looks like signal in crossspreads. This paper makes a detailed analysis of the scattered noise problem, both for point and line scatterers, in 3D shot gathers and in cross-spreads. The findings of Meunier are confirmed and also apply to line scatterers. Then alternative strategies are proposed to solve the problem. The most successful method acquiring data with areal geometry (3D shot or 3D receiver gathers) is also the most expensive. As an alternative, orthogonal geometry can be used provided the receiver and/or source lines are laid out as 2D ribbons rather than 1D strings. This can be implemented using a single-point technique and also by using array-based acquisition. The arraybased technique is the least expensive of the three alternatives.

A comparison of two different approaches to 3D seismic survey design

This paper compares conventional 3D seismic survey design with the 3D symmetric sampling approach. Conventional survey design focuses on regular offset distributions and on midpoints in bin centers. The 3D symmetric sampling approach focuses on spatial continuity. The differences are illustrated by the way in which obstacles are tackled: skidding and offsetting as in the conventional approach, or smooth acquisition lines as promoted in 3D symmetric sampling. The latter approach allows successful prestack noise suppression and reduces prestack migration artifacts to a minimum.

Converted Waves: Properties And 3D Survey Design

3D survey design for converted waves should take their specific properties into account. In the first part of this paper properties of PS-waves are investigated in various minimal data sets (3D single-fold basic subsets of acquisition geometry). The apparent velocities in the 3D receiver gather are much larger than in the 3D shot gather. The cross-spread shows significant asymmetry. Illumination and resolution depend strongly on the minimal data set. In the second part the consequences of these properties for 3D survey design are reviewed. It turns out that parallel geometry is a better choice than orthogonal geometry, unless azimuth-dependent effects need be analyzed. Receiver sampling is preferably denser than shot sampling.

Processing with offset‐vector‐slot gathers

Conventional prestack processing suffers from the absence of proper common-offset gathers in the crossed-array geometries. This requires a new approach to prestack processing, which recognizes the particular requirements of those geometries. This paper provides a strategy for prestack processing based on the construction of pseudo-minimal data sets (pMDSs), i.e., data sets, which are as nearly as possible MDSs, yet extend across the whole survey area. The strategy assumes 3-D symmetric sampling of the input data. The most suitable pMDS in orthogonal geometry is a collection of offset-vector slots (OVS gathers). Each OVS contains data with a limited in-line offset range and a limited cross-line offset range.

Symmetric sampling of multiple‐coverage seismic data

The pre-stack seismic data set is in its simplest form a function of three independent variables: shot coordinate x receiver coordinate ‘r’ and time t. A linear tra&formation links the spatial coordinates x ,x to the midpoint coordinate x and the offgetr coordinate x . For each spatial coordinate there is a corresgonding wavenumber: k k ,k and k . Consid&artio! of thg properties of the prestack wave field leads to the definition of the symmetric sampling technique as the preferred way of recording Z-D multiple-coverage seismic data. The effects of asymmetric spatial sampling are discussed and illustrated with an example of asymmetry in the common midpoint gather of a center-spread geometry. Symmetric spatial sampling is compared with the stack-array approach proposed by Anstey. Symmetric sampling will lead to better data quality, especially in case of complex geology or strongly varying groundroll. The combined effect of field arrays and CHP array is to be analysed in the two-dimensional (km,ko) domain, rather than as a function of only one wavenumber. Examples of the total stack response are discussed.