Geologic pattern recognition from seismic attributes: Principal component analysis and self-organizing maps

Abstract

Interpretation of seismic reflection data routinely involves powerful multiple-central-processing-unit computers, advanced visualization techniques, and generation of numerous seismic data types and attributes. Even with these technologies at the disposal of interpreters, there are additional techniques to derive even more useful information from our data. Over the last few years, there have been efforts to distill numerous seismic attributes into volumes that are easily evaluated for their geologic significance and improved seismic interpretation. Seismic attributes are any measurable property of seismic data. Commonly used categories of seismic attributes include instantaneous, geometric, amplitude accentuating, amplitude-variation with offset, spectral decomposition, and inversion. Principal component analysis (PCA), a linear quantitative technique, has proven to be an excellent approach for use in understanding which seismic attributes or combination of seismic attributes has interpretive significance. The PCA reduces a large set of seismic attributes to indicate variations in the data, which often relate to geologic features of interest. PCA, as a tool used in an interpretation workflow, can help to determine meaningful seismic attributes. In turn, these attributes are input to self-organizing-map (SOM) training. The SOM, a form of unsupervised neural networks, has proven to take many of these seismic attributes and produce meaningful and easily interpretable results. SOM analysis reveals the natural clustering and patterns in data and has been beneficial in defining stratigraphy, seismic facies, direct hydrocarbon indicator features, and aspects of shale plays, such as fault/fracture trends and sweet spots. With modern visualization capabilities and the application of 2D color maps, SOM routinely identifies meaningful geologic patterns. Recent work using SOM and PCA has revealed geologic features that were not previously identified or easily interpreted from the seismic data. The ultimate goal in this multiattribute analysis is to enable the geoscientist to produce a more accurate interpretation and reduce exploration and development risk.

Seismic expression and geomorphology of igneous bodies: A Taranaki Basin, New Zealand, case study

“…In contrast, voxels that exhibit a very distinct attribute behavior (they lie far from each other in 4D space), project onto different parts of the manifold and appear as different colors. Details can be found in Zhao et al (2016) and Roden et al (2015). Figure 23 shows a vertical slice connecting the four Kora wells illustrating the distribution of the chaotic moderate and the continuous high-amplitude seismic facies.…”

Semiautomatic and Automatic Cooperative Inversion of Seismic and Magnetotelluric Data

“…Examples of well-known geostatistical methods that may have value in cooperative inversion include basic methods such as kriging or co-kriging and clustering techniques such as k-means, Fuzzy logic, principal component analysis (PCA) and neural networking (Bedrosian et al 2007;De Benedetto et al 2012;Di Giuseppe et al 2014;Dubrule 2003;Kieu and Kepic 2015;Klose 2006;Roden et al 2015;Ward et al 2014).…”

“…9. We should reiterate that only large subvolumes are used to create the prior conductivity model after which full 3D MT inversion is applied to recover detailed conductivity distribution and associated fitting error statistics for comparison with other strategies Note that we broadly follow the methodologies expressed in Roden et al (2015), with the main difference being that Roden et al (2015) used self-organising maps (SOM) for clustering, while we use the k-means method (Alférez et al 2015;Nanni et al 2008). The .…”

“…For the two reference drill holes located at the Nevada site, we have used geostatistical methods to divide the rock into classes based on a suite of geochemistry elements. For the geochemical data, PCA and k-means clustering techniques have been used to reduce the data dimension to highlight a limited number of geochemically similar rock types (Alférez et al 2015;Grunsky and Smee 2003;Jolliffe 2002;Nanni et al 2008;Roden et al 2015). The purpose of the analysis was to check the consistency of geochemical groupings with seismic boundaries, prior model conductivity distributions and the final conductivity distribution from semiautomatic and automatic cooperative inversion strategies.…”

Classificação de sismofáceis carbonáticas a partir da técnica Self-Organizing Maps (SOM).

“…Estes processos utilizam das mais diversas técnicas relacionadas à classificação e predição de informações geológicas, a saber: regressão linear multiatributos, análise por principais componentes (PCA), k-means, lógica fuzzy, RNA's supervisionadas e não supervisionadas dentre as principais (RODEN; SMITH; SACREY, 2015,ZHAO et al, 2015,SHAKIBA et al, 2015.…”