A case study on semiautomatic seismic interpretation of unconformities and faults in the southwestern Barents Sea

Abstract

Recently, there has been a growing interest in automatic and semiautomatic seismic interpretation, and we have developed methods for extraction of 3D unconformities and faults from seismic data as alternatives to conventional and time-consuming manual interpretation. Our methods can be used separately or together, and they are time efficient and based on easily available 2D and 3D image-processing algorithms, such as morphological operations and image region property operations. The method for extraction of unconformities defines seismic sequences, based on their stratigraphic stacking patterns and seismic amplitudes, and extracts the boundaries between these sequences. The fault-extraction method extracts connected components from a coherence-based fault-likelihood cube where interfering objects are addressed prior to the extraction. We have used industry-based data acquired in a complex geological area and implemented our methods with a case study on the Polhem Subplatform, located in the southwestern Barents Sea north of Norway. For this case study, our methods result in the extraction of two unconformities and twenty-five faults. The unconformities are assumed to be the Base Pleistocene, which separates preglacial and postglacial Cenozoic sediments, and the Base Cretaceous, which separates the severely faulted Mesozoic strata from prograding Paleocene deposits. The faults are assumed to be mainly Jurassic normal faults, and they follow the trends of the eastern and southwestern boundaries of the Polhem Subplatform; the north–south-trending Jason Fault complex; and the northwest–southeast-trending Ringvassøy-Loppa Fault complex.

Automatic extraction of dislocated horizons from 3D seismic data using nonlocal trace matching

“…A 3D view of an intersecting inline and crossline, with three seismic horizons extracted from a seismic section constrained by unconformities. The unconformities have been identified with a semiautomatic method (Bugge et al, 2018), and the seismic horizons are correlated with the method proposed in this study. The case example covers part of the Polhem Subplatform in the southwestern Barents Sea, as illustrated in Figure 1, which includes annotated stratigraphic and structural details.…”

“…The color bars in a and b illustrate correlation accuracy. Wu and Hale, , 2015aBugge et al, 2018). Here, we use a semiautomated method for unconformity extraction introduced by Bugge et al (2018) to constrain our target sequence.…”

“…Wu and Hale, , 2015aBugge et al, 2018). Here, we use a semiautomated method for unconformity extraction introduced by Bugge et al (2018) to constrain our target sequence. This method assumes that the stratigraphic stacking pattern and the seismic amplitudes above and below an unconformity are significantly different.…”

Assessing the accuracy of fault interpretation using machine-learning techniques when risking faults for CO2 storage site assessment

“…The ease and accuracy of seismic interpretation is continually increasing, associated with advancing geophysical and rock physics tools (Chopra and Marfurt, 2007;Avseth et al, 2010), as well as the increased use of automated and semiautomated technologies (e.g. Araya-Polo et al, 2017;Bugge et al, 2018). While technology has progressed to allow users to quickly interpret horizons using facilities such as auto-tracking, the ability for machine learned algorithms for automated fault extraction has, until recently, been lacking.…”