Seismic texture attributes

Seismic Texture Attributes. Subsection 1.2 introduced seismic sequence stratigraphy as a means of explaining the structure of the subsurface, helping determine the depositional environments and possible rock type distributions. Essentially, this process boils down to an analysis of seismic bodies defined by their internal textures and external shape, often referred to as seismic facies analysis. This type of analysis is a must in seismic interpretation to locate potential reservoirs, especially in complex oilfields. 

For further information, extensive overviews of the interplay between seismic facies analysis and seismic texture attributes are given in the chapter [8] of Carrillat and Vallfe. 

A. Carrillat, T. Randen, L. Spnneland, and G. Elvebakk (2002) Automated mapping of carbonate mounds using 3D seismic texture attributes. Proceedings of the Society of Exploration Geophysicists, Annual Meeting. 

D seismic texture attributes presented in this chapter are subdivided into two groups. The first includes kinematic texture attributes that capture the reflector orientation or the reflector continuity information. The second defines dynamic texture attributes that capture features in the seismic signal... 

The traditional approach of seismic attribute computation is to extract attributes along vertical traces, irrespective of any dipping nature of the reflections. This industry standard clearly implies a risk of introducing artefacts, when the stratigraphic pattern is not layer cake and flat. As a more consistent alternative, seismic texture attributes compensate for the dip and azimuth or make the attribute extraction invariant to local dip and azimuth. In addition, they are genuine 3D with no trace bias as opposed to coherency and semblance attributes. Moreover, these attributes are amplitude-invariant. 

A geometrical tensor is used in 3D seismic texture attributes for dip and azimuth estimation. This local dip and azimuth estimation (local orientation estimation) approach is based on three steps (see [35, Section 2]) ... 

The general inversion scheme for reservoir characterisation and delineation from seismic multicomponent data involves the transformation of the converted shear waves (PS data) to PP time domain. This operation is performed in order to have the multi-component data with the same time reference for allowing direct comparison and analysis of both PP and PS data. The transformation of PS data to PP time requires a detailed analysis of the overburden, and interpretation of correlative reflection events on both data sets. The next step in the general inversion scheme involves 3D seismic facies analysis using seismic texture attributes as described earlier. 

A few moves towards texture attribute extraction in seismic data are presented by Sheriff et al. [15]. The probably most common seismic texture... 

Fig. 4. [Reproduced in colour in Plate 5 on page 423.] An example of a seismic cube with a chaotic pattern representing a gas chimney is shown in (a), its dip and Eizimuth attributes in (b) and (c) and a chaos texture attribute highlighting the gas migration path in (d).

T. Randen, E. Monsen, C. Signer, A. Abrahamsen, J.O. Hansen, T. Saeter, J. Schlaf, and L. Spnneland (2000) Three-dimensional texture attributes for seismic data analysis. Expanded Abstr., Int. Mtg., Soc. Exploration Geophys. 

B.P. West, S.R. May, J.E. Eastwood, and C. Rossen (2002) Interactive seismic facies classification using textural attributes and neural networks. The Leading Edge 10, 1042-1049. 

In our approach, the 3D nature of the seismic data is preserved. The analysis is made using a neural network algorithm producing a 3D classification output. The value of this approach to 3D seismic texture mapping has already been demonstrated in the analysis of gas chimneys [25, 34]. Distinct advantages of the approach are that (1) it is independent from and requires no previous horizon interpretation, (2) multiple attributes are simultaneously... 

Dynamic Texture Attributes. Dynamic texture attributes are generated from the original seismic cube. The volume reflection spectrum (VRS) attributes realize spectral analysis of the reflectivity response for each seismic trace [39]. Each trace is characterized in terms of its eigenvalue (spectral attribute) and the associated eigenvector (orthogonal, polynomial), to a > proximate the reflection amplitude along the trace in a least squares sense. For texture mapping a set of discrete spectral VRS coefficients is combined into a composite spectral representation. 

Fig. 3. [Reproduced in colour in Plate 16 on page 433.] Workflow for seismic facies mapping using texture attributes and 3D classification based on neural network algorithm.

Fig. 6. [Reproduced in colonr in Plate 19 on page 434.] Classification of seismic facies based on texture attributes defines a geological/structural model. Using iterative and hierarchical classification capability, seismic facies can be calibrated and assigned to Uthology and fluids using well data or another set of attributes such as amplitude-based, or AVO data.

In the 2D case, a cross-section from an azimuth attribute cube [16] can be used as input data. An attribute is defined as a mathematical operator, or a composition of operators, capturing properties from seismic data. The azimuth attribute has, together with other attributes, such as dip, chaotic texture and continuity, been successfully applied to mapping carbonate reef structures, gas chimneys, channels, fans, faults etc. 

---