Next Article in Journal
On the Application of 238U, 235U, and 232Th Decay Series Disequilibria to Understanding the Processes and Timescales of Oceanic Basalt Petrogenesis
Previous Article in Journal
Integrating MCDA and Rain-on-Grid Modeling for Flood Hazard Mapping in Bahrah City, Saudi Arabia
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Geosciences
Volume 16
Issue 1
10.3390/geosciences16010033
geosciences-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

3D Multi-Attribute Ant Tracking for Fault and Fracture Delineation—A Case Study from the Anadarko Basin

by
Sreejesh V. Sreedhar
,
Camelia C. Knapp
* and
James H. Knapp
Boone Pickens School of Geology, Oklahoma State University, 105 NRC, Stillwater, OK 74078, USA
*
Author to whom correspondence should be addressed.
Geosciences 2026, 16(1), 33; https://doi.org/10.3390/geosciences16010033
Submission received: 22 October 2025 / Revised: 25 December 2025 / Accepted: 28 December 2025 / Published: 6 January 2026
(This article belongs to the Section Geophysics)
Browse Figures
Versions Notes

Abstract

Faults and fractures play a critical role in subsurface systems; they may act as hydrocarbon traps, compartmentalize reservoirs, or serve as conduits for fluid migration across stratigraphic intervals. Consequently, fault delineation from seismic data plays a key role in reservoir characterization. This study presents a workflow for generating ant-tracking attribute volumes using multiple structural attributes to enhance fault/fracture delineation. Our results were thereafter validated with formation microimager (FMI) data. The workflow involves a sequential process comprising seismic data conditioning, structural attribute computation, and ant-tracking volume generation. Variance, curvature, and amplitude contrast attributes were calculated on conditioned 3D seismic data and subsequently used as input for the ant-tracking process. Parameter optimization was conducted through an iterative process of varying individual parameters and qualitatively assessing the results against key seismic features in both vertical sections and time slices. The ant-tracking volumes generated from individual attribute volumes were integrated to produce a composite volume, which served as input for automatic fault extraction. The resultant fault patch orientations were consistent with the formation microimager (FMI) log orientations. The integration of multiple structural attributes within the ant-tracking workflow significantly enhanced fault and fracture delineation by leveraging the complementary strengths of each attribute.

1. Introduction

Faults and fractures exert critical and diverse controls on subsurface systems. They may serve as hydrocarbon traps, compartmentalize reservoirs [1], serve as conduits for fluid migration into different stratigraphic intervals, or act as potential leakage pathways for subsurface fluids [2,3]. In the context of geologic carbon storage, faults and fractures are equally important, as they can either increase the permeability in tight reservoirs, serve as conduits facilitating CO2 migration into multiple target formations, or pose a risk by providing potential leakage pathways [3,4,5,6]. Consequently, fault and fracture delineation on seismic data plays a key role in reservoir characterization and effective implementation of carbon capture and storage (CCS) strategies.
Seismic attributes are essential tools for enhancing the stratigraphic and structural interpretation of seismic data, and are particularly effective for mapping faults and fractures, including subtle features of sub-seismic resolution. Numerous studies have demonstrated the effectiveness of structural attributes in enhancing fault and fracture interpretation from seismic dataset [7,8,9].
Coherence is one of the earliest attributes developed for discontinuity detection, measuring the similarity of seismic waveforms between neighboring traces. Abrupt changes in coherence values indicate structural or stratigraphic discontinuities [10,11]. Variance is a related attribute that quantifies the variability of waveforms within the analysis window, with higher variance corresponding to lower coherence [12,13].
Curvature, a 3D measure of how much a surface deviates from being planar, helps remove regional dip effects and enhances small-scale features that may be related to primary depositional structures or minor faults [14]. Curvature attributes are widely used for fault and fracture prediction [15,16,17,18]. In a case study from the North Slope of Alaska, curvature and aberrancy attributes were effectively applied to delineate sub-seismic faults [19]. Similarly, attributes such as maximum curvature, variance, t* attenuation, and iso-frequency have proven effective in basement fracture characterization [20]. Amplitude contrast is another seismic attribute used to highlight faults and fractures. The attribute uses a Sobel-based edge detection operator, similar to the classical Sobel filter used in image processing, adapted for seismic data to enhance structural discontinuities [21,22]. The method provides a robust measure of local amplitude gradients and is particularly effective for delineating faults, fractures, and other features like edges of channels that generate sharp amplitude contrasts within seismic volumes [21].
The structural attribute volumes serve as an input for the ant-tracking attribute. Ant tracking applies a swarm intelligence algorithm that traces faults and fractures through seismic attribute volumes, identifying continuous trends within the data [23]. Ant tracking mimics the foraging behavior of ants, deploying virtual agents that navigate a seismic volume. These agents are guided by local attribute values, particularly those indicating structural discontinuities. As they traverse the volume, the agents preferentially follow paths of coherent signal variation, which are commonly associated with faults, while disregarding random noise or incoherent features [23]. This process effectively highlights and connects discontinuities commonly associated with faults. Automatic fault surface extraction is made possible by the application of ant tracking on fault-enhancing attribute volumes.
Several workflows for ant tracking have been proposed in the literature, each varying in terms of input attributes, parameter selection, and postprocessing approaches. One of the studies employs summation of amplitude contrast attributes calculated at 10° azimuthal increments, and the resultant attribute volume is then used as input to ant tracking [22]. In another study, seismic data were divided into low, medium, and high frequency bands, which were then used as input to generate ant-tracking volumes from the variance attributes [24]. Another workflow employs several attributes, including curvature, dip deviation, and cosine of phase, to generate ant-tracking volumes, comparing the results to select dip deviation as the attribute that produced the most reliable fault predictions [25]. A case study from the Taranaki Basin in New Zealand utilized graphic equalizer to adjust the frequency component of the post-stack seismic dataset and generated chaos attribute as an input to ant tracking [26].
While single-attribute ant tracking can highlight faults, each seismic attribute emphasizes different geological features and is sensitive to noise [14,27]. Variance highlights discontinuities [12], curvature emphasizes subtle flexures [15], and amplitude contrast responds to abrupt changes in amplitude [21]; however, no single attribute captures the complete fault network. Conventional seismic reflection data-based fault detection relies heavily on single-attribute discontinuity measures such as variance, semblance, and curvature, as well as standard ant-tracking workflows [8,9,10,28]. Existing multi-attribute integration has been shown to improve fault delineation [29,30], but they provide limited guidance on systematic workflows for parameter optimization, and generally do not employ multiple passes of ant tracking or attribute integration for enhanced detection of faults. This study addresses these gaps by developing a geologically calibrated, weighted multi-attribute ant-tracking workflow that systematically evaluates attribute contributions, optimizes their combination, and demonstrates clear improvements over single-attribute volumes in terms of structural coherence and fault continuity.
This study presents the use of multiple structural seismic attributes for generating the ant-tracking attribute volume for fault and fracture delineation, using a 3D seismic dataset from the Anadarko Basin in Oklahoma, USA. Integrating multiple attributes within an ant-tracking workflow leverages their complementary strengths, improving fault continuity, reducing noise, and increasing confidence in structural interpretation. Although ant tracking has been widely applied for fault detection, its effectiveness depends strongly on attribute selection, parameter tuning, and geological calibration [23,26,31]. In this study, we develop and optimize a multi-attribute workflow that integrates variance, curvature, and amplitude contrast through iterative parameter testing and a weighted attribute combination strategy. This approach, coupled with multiple stage ant-tracking process, and validation against FMI fracture orientations provide a more geologically consistent representation of the fault and fracture system.

2. Study Area

The study area lies within the Anadarko Basin, one of the most prolific hydrocarbon provinces in the cratonic interior of North America (Figure 1). The Anadarko Basin is one of the deepest and most geologically complex sedimentary basins in North America, with a thick stratigraphic succession ranging from to Cambrian to Permian in age [32,33]. The basin comprises a combination of marine and continental deposits, shaped by prolonged subsidence, episodic tectonic activity, and eustatic sea-level fluctuations. The basin extends across western Oklahoma, the Texas Panhandle, and southwestern Kansas. It is bounded by the Amarillo–Wichita Uplift to the southwest, the Nemaha Ridge to the east, and the Anadarko shelf to the west and north [32,33].
The formation of the Anadarko Basin is associated with the evolution of the Southern Oklahoma Aulacogen (SOA), interpreted as a failed arm of an ancient rift triple junction that developed during the Late Precambrian to Middle Cambrian [33,36]. The initial phase of rifting and igneous activity was followed by a prolonged period of thermal subsidence which resulted in the development of the southern Oklahoma trough [33]. This structural depression served as a major depocenter for lower Paleozoic marine sediments, including thick carbonate sequences.
The Anadarko Basin, in its present form, developed during the Pennsylvanian orogeny which induced significant crustal shortening and uplift, most notably the Wichita–Amarillo Uplift to the south and the Nemaha Uplift to the east [32]. The tectonic loading from these uplifts created an asymmetric foreland basin that experienced rapid subsidence and accommodated exceptionally thick accumulations of Pennsylvanian clastic sediments derived from the erosion of the surrounding uplifts [32,36]. The basin axis reaches depths exceeding 12,000 m along its southern margin, making it one of the deepest sedimentary basins in North America [36]. Syntectonic faulting and folding resulted in a series of structural traps and stratigraphic pinch-outs [33].
The Late Permian saw a cessation of tectonic activity and the region stabilized with sedimentation dominated by red beds and evaporites. The Permian red beds and evaporites act as regional seals for the basin [32,36]. The Late Permian represents the final phase of sediment accumulation in the basin. The Mesozoic era is characterized by regional uplift, potentially the distant influence of the Laramide orogeny, and subsequent erosional dominance stripping away the Permian and older rocks. The basin transitioned into a relatively stable cratonic region during this period [32,36].
The Anadarko Basin is crisscrossed by a complex network of faults, reflecting its poly phase tectonic history (Figure 1). The Wichita–Amarillo fault system is the most prominent fault system, defining the southern margin of the deep basin, and it comprises south-dipping, high-angle reverse faults and thrust faults formed during Pennsylvanian transpressional tectonics [33]. The Wichita Uplift faults primarily exhibit a west–northwest to east–southeast (WNW–ESE) trend, consistent with the overall orientation of the Wichita–Amarillo Uplift [33]. The regional compression also created gentle anticlinal and synclinal structures throughout the basin [33].
Nemaha Uplift faults are characterized by north–northeast-trending (NNE-SSW), steeply dipping normal and reverse faults formed during the Pennsylvanian compression, and may likely include reactivated Proterozoic basement fabrics [32]. Intra-basinal normal faults, associated with basement weaknesses, also created structural highs and lows. Localized strike-slip faults and transfer faults are also present, though compressional tectonics dominate the structural framework of the Anadarko Basin [32,37]. Transfer faults typically accommodate differential displacement between normal fault blocks.
The Devonian–Mississippian stratigraphy of the basin has been considered one of the most successful resource plays in the Midcontinent, referred to as the Sooner Trend of the Anadarko Basin in Canadian and Kingfisher Counties (STACK) and the South-Central Oklahoma Oil Province (SCOOP) [38]. These low permeability unconventional reservoirs include the Woodford Shale, the Mississippian Group, and the Springer Group [38]. A generalized stratigraphic column for the Anadarko Basin is shown in Figure 2 [39]. The Mississippian Meramec interval is primarily composed of siliceous limestones, quartz-rich calcareous siltstones, argillaceous siltstones, and organic-rich mudstones [40]. The observed lateral facies distribution and vertical facies successions suggest deposition on a distally steepened ramp [40,41].

3. Data and Methodology

The dataset used for the present study is a 3D seismic volume from the Anadarko Basin, Oklahoma, processed using advanced techniques such as orthorhombic pre-stack time migration. The 3D seismic survey area covers approximately 324 square km (125 square miles). It was acquired with a sample rate of 2 ms, bin spacing of 25 m (82.5 ft), and 6 s record length. The 3D seismic volume is of good quality throughout the survey area; it exhibits strong, coherent reflectors with limited noise. It covers part of the STACK and SCOOP play of the Anadarko Basin. A Formation Microimager (FMI) log also exists in a well-drilled (Well-D) in the study area. The Formation Microimager (FMI) is a wireline tool that produces high-resolution electrical images of the borehole wall by measuring microresistivity, which is used to identify fractures, faults, borehole breakouts, and other stratigraphic or structural features [42]. A previous study analyzed the FMI log from Well-D in the study area for fracture and stress evaluation [43]. The FMI log was acquired in water-based mud, which provided high-quality borehole images [43]. The interpretation of the FMI log from Well-D was used to validate the orientations of faults extracted from 3D seismic data [43]. The comparison was performed over the Meramec Formation interval, corresponding to the depth range of the FMI log. Fracture orientations derived from the FMI log were compared with the orientation of fault patches extracted from the seismic volume. This approach allowed a direct, formation-specific comparison between well-scale discontinuities and seismic-scale discontinuities, providing a basis for assessing the consistency between FMI-derived fracture orientations and 3D seismic fault geometry.
The footprint of the 3D seismic survey used in this study, together with the location of the well containing the FMI log data and a representative seismic profile showing the formation tops, is illustrated in Figure 3. All seismic profiles and attribute images presented in this study were derived from this 3D survey area. The workflow applied to generate the multi-attribute ant-tracking volume and subsequent analyses is summarized in Figure 4. All analyses conducted in this study were performed using the Petrel E&P software platform, version 2023.3 [44]. The dataset is proprietary; hence, time/depth information and inline/crossline details are not clearly stated in figures presented.
The workflow for generating the ant-tracking volumes is depicted in Figure 4 and consists of the following primary stages: (1) seismic data-conditioning to enhance structural continuity and suppress noise, (2) attribute generation to highlight discontinuities, (3) iterative ant tracking, involving parameter optimization and single-attribute ant-tracking volume generation, (4) multi-attribute ant-tracking volume generation, and (5) automatic fault extraction.

3.1. Seismic Data Conditioning

Seismic data conditioning is important for generating effective fault attribute volumes. The goal of the seismic data conditioning process is to reduce noise and enhance structural features. Conditioning of the post-stack seismic data was performed before calculating the attribute volumes. The post-stack conditioning step involved the use of a spatial filter for enhancing signal-to-noise ratio in the seismic data, while retaining the structure (Figure 4). A structural smoothing filter and trace gradient were applied to the post-stack seismic data to enhance the structural features like faults. The structural smoothing filter applies local averaging using a Gaussian-weighted averaging filter, effectively reducing random noise while preserving geological structures such as faults and stratigraphic features [28,45]. The seismic reflectors appear more coherent, and faults appear more pronounced after applying the structural smoothing.
Trace gradient calculates the gradient along a seismic trace, typically over a window of five samples, which gives the rate of change in amplitude between adjacent time samples, effectively capturing how the amplitude varies at each point [44,45]. The trace gradient highlights zones of rapid amplitude variation, with higher values at points of maximum rate of change in amplitude, often corresponding to geological features like layer boundaries, faults, or other discontinuities within the subsurface [44,45]. As an edge-enhancement attribute, trace gradient improves the visibility of these features, which are often subtle or poorly expressed in unconditioned seismic data.

3.2. Structural Attributes

Multiple structural attributes were calculated on the post-stack seismic data in order to enhance the imaging of faults and fractures. Attributes selected to delineate faults in the 3D seismic area were variance, curvature, and amplitude contrast attributes (Figure 4).
The variance seismic attribute is widely used for mapping faults. The variance attribute highlights discontinuities in seismic data such as faults, fractures, stratigraphic terminations, and major structural lineaments [12,13,45]. The variance attribute quantifies the degree of similarity/dissimilarity between adjacent seismic traces within a localized window to estimate lateral heterogeneity [13,44]. High variance values indicate zones with faults/fractures. Dip-guided variance attribute enhances structural features.
Curvature is a geometric attribute that quantifies how sharply a curve bends at a specific point, defined as the rate of change in direction along the curve [15,17,46,47]. In seismic interpretation, curvature attributes are valuable for detecting subtle variations in structural trends and tectonic features, including lineaments [48]. The most extreme curvature highlights the hanging wall and footwall near fault positions and helps to delineating faults/fractures [47,48,49]. The most positive and most negative curvature values are returned as a combined value for the most extreme curvature attribute [44,48,49].
Amplitude contrast attribute is another structural attribute which highlights the whole structure in addition to edge enhancement [21,22]. Amplitude contrast attribute uses a Sobel filter to delineate areas with amplitude discontinuities which may be related to stratigraphic terminations or structural lineaments [21]. The amplitude contrast attribute can delineate faults, fractures, and salt structures [21]. Structural attributes mentioned above are generated on the conditioned seismic volume with optimized parameters. These attribute volumes serve as input for ant tracking.

3.3. Ant Tracking

The ant-tracking algorithm is a widely used method for extracting fault surfaces from seismic data [24,26,29,30,50]. The ant-tracking algorithm uses the principles of swarm intelligence (ant colony system) to delineate faults by picking trends in the seismic data [23,31,51]. Similarly to real ants navigating using pheromone trails, the discontinuities are tracked by multiple “ants” and noise is not tracked or tracked by single “ants”, which means it is removed during the ant-tracking process [31,44].

3.3.1. Parameter Optimization

The structural attributes generated from the conditioned seismic data were used as an input for ant-tracking volume generation. The ant-tracking process was carried out in two stages. In the first stage, the parameters were systematically tested and optimized through multiple iterations to enhance fault continuity (Figure 4). Initial ant boundary, ant step size, legal steps allowed, and illegal steps allowed are the parameters changed iteratively, keeping ant-tracking deviation and stop criteria constant. The final parameter set was selected based on visual inspection and alignment with key seismic features. Ant-tracking volumes were generated for each attribute using custom parameter sets.
The ant-tracking parameters and their definitions are listed below as follows [44]:
  • Initial ant boundary: The number of voxels as the territorial radius around each ant.
  • Ant-tracking deviation: the maximum deviation from a local maximum while tracking.
  • Ant step size: The number of voxels an ant advances for each step.
  • Illegal steps allowed: How far an ant can continue without finding an edge value. These are temporary movements into low-discontinuity areas, which allow the ant to continue in order to reconnect the trend if the discontinuity reappears.
  • Legal steps allowed: Controls how connected a detected edge must be to distinguish edge from noise. These are preferred movements that follow high-discontinuity trends indicated by the seismic attributes.
  • Stop criteria: The percentage of illegal steps allowed throughout a single agent’s life.
Ant-tracking parameter tests included as Supplementary Materials (Tables S1–S3) present the ant-tracking parameter tests conducted for the variance, curvature, and amplitude contrast attribute volumes, respectively.

3.3.2. Ant-Tracking Volumes from Attributes—Single-Attribute Ant-Tracking Volumes

The optimal parameter set obtained from the above iterative parameter testing was used to generate the first-pass (Stage 1) ant-tracking output (Figure 4). This output volume was subsequently used as input for the second-pass (Stage 2) ant tracking, to further enhance fault continuity and resolution, generating single-attribute ant-tracking volumes (Figure 4). The final parameter sets applied in both stage 1 and stage 2 are summarized in Table 1.

3.3.3. Composite Ant-Tracking Volume—Multi-Attribute Ant-Tracking Volume

Composite volumes were generated from the three single-attribute ant-tracking volumes (Figure 4). An average and a weighted composite ant-tracking volume were computed. For evaluation, the volumes were visualized in vertical sections and time slices; they were also co-rendered with the seismic amplitude volume to compare fault continuity, edge sharpness, and noise suppression. The weighted composite volume was selected over the average composite, as it exhibited superior fault continuity, sharper edges, and greater geological consistency upon visual analysis. A higher weight (0.6) was assigned to the curvature-derived ant-tracking volume, as curvature was found to emphasize fault trends with greater continuity and definition. In contrast, lower weights (0.2 each) were assigned to the ant-tracking volumes derived from amplitude contrast and variance attribute volumes, which contributed useful structural information but were more sensitive to noise and had less continuity than curvature-based ant-tracking volume. Another pass of the ant-tracking process was carried out on the composite volume, with the parameters for stage 2 (with *) described in Table 1, for generating the multi-attribute ant-tracking volume (Figure 4).

3.4. Automatic Fault Extraction

The multi-attribute ant-tracking volume was visualized in 3D and a formation-constrained version of the same volume was generated to focus only on the stratigraphic level of interest (Figure 4). The multi-attribute ant-tracking volume was clipped between interpreted formation boundaries (top and bottom surfaces of the formation) to generate the formation-constrained volume. The resultant volume was then utilized for automatic fault extraction that produced a dense distribution of fault patches. A fault patch is a continuous surface representing part of a fault. Multiple fault patches can be combined to form a complete fault surface. Fault patches were extracted from the volume using parameters optimized for high-confidence detection [44]. Table 2 shows parameters for fault extraction.
The extracted patches were filtered by surface area, dip, and patch confidence to remove very small, geologically unrealistic fault patches and to retain higher-confidence patches. The strike directions for fault patches were analyzed using a rose diagram. The orientations (azimuths) of the filtered fault patches were extracted. These were plotted on a rose diagram, providing a quantitative view of the dominant fault trends within the formation. This allowed comparison of fault orientations from seismic data with borehole fault orientations.

4. Results

Seismic data conditioning effectively enhanced structural features and signal-to-noise ratio. Variance, curvature, and amplitude contrast attributes were used as input for single-attribute ant-tracking volume, followed by a weighted composite to generate a multi-attribute ant-tracking volume. Rose diagrams were generated from the extracted seismic fault patch orientations and compared with FMI-derived fracture orientations to evaluate structural consistency.

4.1. Single-Attribute Ant-Tracking Volume

Ant tracking was applied separately to three seismic attribute volumes, namely variance, curvature, and amplitude contrast, to enhance fault and fracture detection (Figure 5). The time slices shown in the figure are at the same TWT (two-way travel time) through the single-attribute ant-tracking volume. A strike-slip fault is highlighted with red arrows in Figure 5. The results demonstrate that each attribute emphasizes different features of the seismic data: variance highlights major discontinuities, curvature responds to subtle flexures, and amplitude contrast delineates abrupt amplitude terminations. The single-attribute ant-tracking volumes generated from variance, curvature, and amplitude-contrast attributes reveal a broadly consistent structural framework; however, the expression and continuity of individual fault segments vary among the attributes (Figure 5). Among the three, the curvature-derived ant-tracking volume provides the most continuous and laterally coherent fault trends. In contrast, the variance- and amplitude contrast-based volumes tend to produce shorter, more fragmented segments and exhibit reduced connectivity. These differences are evident along the highlighted strike-slip fault and across the broader 3D seismic area. The enhanced performance of curvature reflects its sensitivity to subtle reflector bending and flexures, making it more effective for capturing fault geometries in this dataset than the other two attributes.

4.2. Multi-Attribute Ant-Tracking Volume

To combine the complementary strengths of the attributes analyzed in this study, a weighted composite multi-attribute ant-tracking volume was generated (Figure 6). Vertical sections through this volume (inline and crossline co-rendered with amplitude) reveal enhanced fault continuity and improved signal-to-noise ratio compared to single-attribute volumes (Figure 5). A time slice through the volume illustrates the lateral distribution of the fault/fracture network. The multi-attribute ant-tracking volume enhanced fault continuity and connectivity, as depicted in both vertical and horizontal views (Figure 6). The integrated volume captures both major through-going faults and secondary, shorter-offset fractures, providing a more geologically coherent representation of the structural framework. Red arrows in Figure 6 indicate selected fault trends that remain consistent and continuous across multiple viewing directions, confirming the robustness of the integrated response. This multi-attribute approach thus yields a superior fault representation than single-attribute results.
A comparison of the single-attribute (curvature-based) and multi-attribute ant-tracking volumes highlights the benefits of integrating multiple structural attributes (Figure 7). Although curvature produces the most coherent fault definition among the single-attribute ant-tracking volumes (Figure 5), several fault and fracture segments remain discontinuous or only partially expressed. In contrast, the weighted multi-attribute volume enhances overall connectivity, reduces attribute-specific gaps, and reveals additional subtle discontinuities. Highlighted areas in the northern and southern portions of the time slice show marked improvements in fault continuity. This improvement reflects the complementary contributions of variance and amplitude contrast, which reinforce curvature-derived trends and collectively produce a more complete representation of the fault/fracture system. When the variance and amplitude-contrast volumes were integrated with curvature to generate the weighted multi-attribute ant-tracking cube, a noticeable improvement in both continuity and structural definition was observed (Figure 7). As a result, the weighted multi-attribute cube provides a more robust and geologically complete representation of the structural fabric than any individual attribute alone.
While the 3D visualization in Figure 6 demonstrates the continuity of the multi-attribute ant-tracking interpretation, selected inline profiles in Figure 8 highlight structural discontinuities in greater detail. Seismic inline sections (Profiles 1 and 2) show the amplitude volume co-rendered with the multi-attribute ant-tracking volume, with red arrows indicating a prominent strike-slip fault zone (Figure 8). The time slice in Figure 8c shows the spatial relationship of these profiles, and blue arrows on the vertical sections mark the level of the time slice.
Integration of multiple attributes in the ant-tracking volume significantly improves fault detection by capturing features emphasized differently in individual attributes, resulting in enhanced continuity, connectivity, and confidence in the interpreted fault/fracture framework. Compared to single-attribute ant-tracking volumes, the multi-attribute volume integrates complementary attribute responses, producing a more coherent and geologically consistent fault/fracture framework with reduced noise and improved continuity (Figure 7).

4.3. Automatic Fault Extraction

The automatic fault extraction produced a dense distribution of planes of discontinuity, known as fault patches (Figure 9). These fault patches were subsequently filtered using surface area, patch confidence, and dip thresholds to remove low-confidence, small-scale, and sub-horizontal artifacts. The filtered dataset is interpreted to predominantly represent fault-related discontinuities, although it is acknowledged that some non-structural seismic artifacts, like edges of the 3D seismic survey or stratigraphic features like channel margins, may persist.

4.4. Comparison of Seismic-Scale Faults and Borehole-Scale Fractures

A previous study interpreted the FMI (Formation Microimager) log available in Well-D to identify faults and fractures [43]. The interpretation provided information on fracture orientation and density, which was compared to seismic-scale fault orientations extracted from the Meramec Formation in this study (Figure 9).
Fracture orientation data from FMI interpretation in Well-D show dominant fracture strike-oriented ENE–WSW, with secondary orientations trending NNW–SSE (Figure 10a). The final ant-tracking attribute from the present study was extracted on the Meramec surface which shows that the attribute matches the structural features on the surface (Figure 10b). A rose diagram generated from the extracted fault patches within the Meramec Formation reveals dominant fault strikes that align closely with the ENE–WSW trend observed in the FMI data, along with a secondary direction consistent with the NNW–SSE set (Figure 10c). Additionally, seismic data show fractures in subordinate fault orientations with the trending WNW-ESE. The close correspondence between the FMI-derived fracture orientations and the seismic-scale fault orientations supports the reliability of the multi-attribute ant-tracking interpretation.
The single-attribute ant-tracking volumes showed complementary fault and fracture patterns (Figure 5), with the multi-attribute volume providing improved continuity and reduced noise (Figure 6 and Figure 8). Fault orientations extracted from the Meramec surface match the dominant ENE–WSW and secondary NNW–SSE trends observed in the FMI log, with seismic data also indicating subordinate WNW–ESE orientations (Figure 10). These results confirm the consistency of seismic- and borehole-scale structural trends, providing a robust basis for interpretation discussed in the following section.
The multi-attribute ant-tracking approach provides a robust and comprehensive characterization of fault and fracture networks, surpassing the capabilities of individual attribute analyses and thereby improving the reliability of structural interpretations.

5. Discussion

This study presented a calibrated multi-attribute ant-tracking workflow that optimized parameters on individual attributes and integrated them using a validated weighting scheme to enhance subtle fault delineation. The single-attribute ant-tracking volumes (Figure 5) indicated that while the overall fault and fracture framework remained consistent, the clarity and continuity of individual fault segments vary depending on the features enhanced by each attribute. Variance emphasized major discontinuities, curvature highlighted subtle flexures, and amplitude contrast responded to abrupt amplitude terminations. In essence, all three attributes illuminated structural discontinuities like faults and fractures, yet each does so by responding to and emphasizing a different characteristic of the discontinuity.
Integration of these complementary responses in the multi-attribute ant-tracking volume (Figure 6 and Figure 7) resulted in a more continuous and geologically consistent fault and fracture network model. Co-rendering the ant-tracking volume with vertical amplitude sections (inline and crossline) together with lateral expression in the time slice confirmed that the multi-attribute approach improved fault visibility, reduced noise, and enhanced confidence in the structural interpretation.
Comparison of FMI-derived fracture orientations with seismic-scale fault orientations from the Meramec Formation revealed a strong alignment between borehole and seismic observations. The dominant ENE–WSW trend identified in the FMI log was mirrored in the fault patch orientations extracted from the seismic volume, and the secondary WNW–ESE trend was also consistent across both datasets. This finding is consistent with previously reported fracture orientation analysis from the same study area [29]. This correlation suggested that the seismic data effectively delineated the fault and fracture framework within the formation. The agreement validated the multi-attribute ant-tracking workflow for enhanced fault and fracture delineation and supported its applicability for generating a discrete fracture network for reservoir characterization.
The novelty of this work lies not in the ant-tracking algorithm itself, but in the design, optimization, and geological calibration of the multi-attribute workflow. The study demonstrated that a targeted combination of ant tracking from variance, curvature, and amplitude contrast attributes, optimized through iterative parameter testing and integrated using a weighted scheme, provided superior fault delineation compared to equal-weight or single-attribute approaches. The workflow incorporated multiple ant-tracking passes strategy that enhanced both major fault corridors and subordinate fracture trends. Importantly, the final multi-attribute volume was validated against FMI-derived fracture orientations, establishing a direct correspondence between seismic-scale discontinuity patterns and borehole-scale structural data. This integrated and geologically calibrated approach constitutes the primary contribution of this work, particularly within the structurally complex setting of the Anadarko Basin.

Limitations of the Workflow

While the developed multi-attribute ant-tracking workflow provided improved delineation of fault and fracture networks, several limitations should be acknowledged.
  • Dependence on seismic data and data conditioning:
  • The quality of all subsequent attributes and ant-tracking results is inherently tied to the seismic data quality and initial conditioning of the seismic data. Choices such as the parameters of structural smoothing can influence fault sharpness and apparent continuity. Although careful testing was performed, variations in the preprocessing steps may produce subtly different outcomes in highly noisy or structurally complex regions.
  • Subjectivity in parameter optimization:
  • Parameter optimization relied on iterative testing followed by visual inspection of vertical sections and time slices specific to the 3D seismic dataset used. This approach is common in attribute-based fault interpretation but nonetheless introduces a degree of interpreter subjectivity. Alternative parameter combinations may result in slightly different fault expressions.
  • Scale discrepancy between seismic faults and FMI fractures:
  • The comparison between seismic-scale fault patches (tens to hundreds of meters) and borehole-scale fracture orientations (centimeters to meters) assumes that the dominant deformation fabric is persistent across scales. The correlation reflects orientation similarity rather than a direct one-to-one structural correspondence.
  • Non-structural seismic artifacts:
  • Although the filtering and patch-extraction steps were designed to preferentially retain fault-related discontinuities, the resulting fault patches may still contain a limited number of non-structural responses. These include 3D seismic survey-edge effects, any residual processing artifacts, and stratigraphic edges such as channel margins or lithologic boundaries that can generate discontinuity-like responses in attribute volumes.
  • Algorithm and software specificity:
  • The ant-tracking implementation used in this study is a proprietary algorithm within the SLB Petrel platform. The findings are specific to this implementation. Results obtained using other commercial or open-source algorithms may differ.

6. Conclusions

This study presents a calibrated multi-attribute ant-tracking workflow for enhanced fault and fracture delineation using 3D seismic data. By integrating variance, curvature, and amplitude contrast attributes through an optimized and weighted ant-tracking strategy, the workflow significantly improves fault continuity, reduces noise, and yields a more geologically consistent structural model than single-attribute approaches. The resulting multi-attribute ant-tracking volume shows strong agreement with FMI-derived fracture orientations, demonstrating consistency between seismic-scale fault patterns and borehole-scale structural observations. This validation highlights the effectiveness of the proposed workflow for structural interpretation and the discrete fracture network construction for reservoir characterization in structurally complex basins. Overall, the study illustrates that careful attribute selection, iterative parameter optimization, and geological calibration are critical for maximizing the effectiveness of ant-tracking techniques. The proposed workflow provides a robust and practical framework for improved fault and fracture characterization from seismic data, particularly in settings where subtle and complex structural features are present.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/geosciences16010033/s1; Table S1: Ant-tracking parameter tests for variance attribute volume; Table S2: Ant-tracking parameter tests for curvature attribute volume; Table S3: Ant-tracking parameter tests for amplitude contrast attribute volume.

Author Contributions

Conceptualization, S.V.S., C.C.K. and J.H.K.; methodology, S.V.S., C.C.K. and J.H.K.; software, S.V.S.; validation, S.V.S.; formal analysis, C.C.K. and J.H.K.; investigation, S.V.S.; resources, C.C.K. and J.H.K.; data curation, S.V.S.; writing—original draft preparation, S.V.S.; writing—review and editing, S.V.S., C.C.K. and J.H.K.; visualization, S.V.S.; supervision, C.C.K. and J.H.K.; project administration, C.C.K. and J.H.K.; funding acquisition, C.C.K. and J.H.K. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the US Department of Energy and Southeast Regional Carbon Utilization and Storage Acceleration Partnership (SECARB-USA, DE-FE0031830) led by the Southern States Energy Board (DE-FE0031557).

Data Availability Statement

The datasets presented in this article are confidential.

Acknowledgments

The authors express their gratitude to Devon Energy for providing the seismic and well log data used in this study. They also acknowledge SLB for granting access to the Petrel E&P Software Platform, which was instrumental in carrying out this research. The authors acknowledge the support for this project provided by the US Department of Energy and Southeast Regional Carbon Utilization and Storage Acceleration Partnership (SECARB-USA, DE-FE0031830).

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
CCSCarbon capture and storage
FMIFormation microimager
TWTTwo-way travel time
SOASouthern Oklahoma aulacogen
STACKSooner trend of the Anadarko Basin in Canadian and Kingfisher counties
SCOOPSouth-central Oklahoma oil province

References

  1. Jolley, S.J.; Fisher, Q.J.; Ainsworth, R.B. Reservoir Compartmentalization: An Introduction. Geol. Soc. Lond. Spec. Publ. 2010, 347, 1–8. [Google Scholar] [CrossRef]
  2. Lawal, M.A.; Bialik, O.M.; Lazar, M.; Waldmann, N.D.; Foubert, A.; Makovsky, Y. Modes of Gas Migration and Seepage on the Salt-Rooted Palmahim Disturbance, Southeastern Mediterranean. Mar. Pet. Geol. 2023, 153, 106256. [Google Scholar] [CrossRef]
  3. Gilmore, K.A.; Sahu, C.K.; Benham, G.P.; Neufeld, J.A.; Bickle, M.J. Leakage Dynamics of Fault Zones: Experimental and Analytical Study with Application to CO2 Storage. J. Fluid Mech. 2022, 931, A31. [Google Scholar] [CrossRef]
  4. Rutqvist, J. The Geomechanics of CO2 Storage in Deep Sedimentary Formations. Geotech. Geol. Eng. 2012, 30, 525–551. [Google Scholar] [CrossRef]
  5. Pawar, R.J.; Bromhal, G.S.; Carey, J.W.; Foxall, W.; Korre, A.; Ringrose, P.S.; Tucker, O.; Watson, M.N.; White, J.A. Recent Advances in Risk Assessment and Risk Management of Geologic CO2 Storage. Int. J. Greenh. Gas Control. 2015, 40, 292–311. [Google Scholar] [CrossRef]
  6. Burnside, N.M.; Shipton, Z.K.; Dockrill, B.; Ellam, R.M. Man-Made versus Natural CO2 Leakage: A 400 k.y. History of an Analogue for Engineered Geological Storage of CO2. Geology 2013, 41, 471–474. [Google Scholar] [CrossRef]
  7. Tingdahl, K.M.; Steen, Ø.; Ligtenberg, J.H. Semi-automatic Detection of Faults in 3-D Seismic Signals. In Proceedings of the SEG Technical Program Expanded Abstracts, San Antonio, TX, USA, 9–14 September 2001. [Google Scholar]
  8. Chopra, S.; Marfurt, K.J. Seismic Attributes for Fault/Fracture Characterization. In Proceedings of the SEG Technical Program Expanded Abstracts, San Antonio, TX, USA, 23–27 September 2007. [Google Scholar]
  9. Verma, S.; Chopra, S.; Ha, T.; Li, F. A Review of Some Amplitude-Based Seismic Geometric Attributes and Their Applications. Interpretation 2022, 10, B1–B12. [Google Scholar] [CrossRef]
  10. Marfurt, K.J.; Kirlin, R.L.; Farmer, S.L.; Bahorich, M.S. 3-D Seismic Attributes Using a Semblance-based Coherency Algorithm. Geophysics 1998, 63, 1122–1479. [Google Scholar] [CrossRef]
  11. Bahorich, M.; Farmer, S. 3-D Seismic Discontinuity for Faults and Stratigraphic Features: The Coherence Cube. Lead. Edge 1995, 14, 1021–1098. [Google Scholar] [CrossRef]
  12. Van Bemmel, P.P.; Pepper, R.E. Seismic Signal Processing Method and Apparatus for Generating a Cube of Variance Values 2000. Available online: https://patents.google.com/patent/US6151555A/en (accessed on 3 December 2025).
  13. Ochoma, U. Seismic Attributes for Enhanced Structural and Stratigraphic Interpretation in Onshore Fuba Field, Niger Delta, Nigeria. MEJAST 2023, 6, 105–117. [Google Scholar] [CrossRef]
  14. Chopra, S.; Marfurt, K.J. Seismic Attributes—A Historical Perspective. Geophysics 2005, 70, 3SO–28SO. [Google Scholar] [CrossRef]
  15. Roberts, A. Curvature Attributes and Their Application to 3D Interpreted Horizons. First Break. 2001, 19, 85–100. [Google Scholar] [CrossRef]
  16. Hart, B.S.; Pearson, R.; Rawling, G.C. 3-D Seismic Horizon-Based Approaches to Fracture-Swarm Sweet Spot Definition in Tight-Gas Reservoirs. Lead. Edge 2002, 21, 28–35. [Google Scholar] [CrossRef]
  17. Sigismondi, M.E.; Soldo, J.C. Curvature Attributes and Seismic Interpretation: Case Studies from Argentina Basins. Lead. Edge 2003, 22, 1122–1126. [Google Scholar] [CrossRef]
  18. Samuel, S.A.; Knapp, C.C.; Knapp, J.H. Seismic Facies Classification of Salt Structures and Sediments in the Northern Gulf of Mexico Using Self-Organizing Maps. Geosciences 2025, 15, 183. [Google Scholar] [CrossRef]
  19. Verma, S.; Bhattacharya, S. Delineation of Complex Fault Network North Slope, Alaska Using Seismic Attributes. In Proceedings of the SEG Technical Program Expanded Abstracts 2019, San Antonio, TX, USA, 15–20 September 2019. [Google Scholar]
  20. Hameed, S.S.; Hermana, M. Fractured Basement Reservoir Characterization Using Multi Seismic Attributes: A Case Study of Malaysian Basin Field. J. Earth Sci. Technol. 2020, 2, 83–89. [Google Scholar]
  21. Aqrawi, A.A.; Bø, T.H. Amplitude Contrast Seismic Attribute. U.S. Patent No. US20120257477A1, 17 May 2016. [Google Scholar]
  22. Kim, M.; Yu, J.; Kang, N.-K.; Kim, B.-Y. Improved Workflow for Fault Detection and Extraction Using Seismic Attributes and Orientation Clustering. Appl. Sci. 2021, 11, 8734. [Google Scholar] [CrossRef]
  23. Pedersen, S.I.; Randen, T.; Sonneland, L.; Steen, Ø. Automatic Fault Extraction Using Artificial Ants. In SEG Technical Program Expanded Abstracts 2002; Society of Exploration Geophysicists: Salt Lake City, UT, USA, 2002; pp. 512–515. [Google Scholar]
  24. Xie, Q.; Zhao, C.; Rui, Z.; Guan, S.; Zheng, W.; Fan, H. An Improved Ant-Tracking Workflow Based on Divided-Frequency Data for Fracture Detection. J. Geophys. Eng. 2022, 19, 1149–1162. [Google Scholar] [CrossRef]
  25. Tran, H.; Bedle, H.; Lubo-Robles, D.; Vera-Arroyo, A. It Is Not Our Fault... yet: A Multiattribute and Swarm Intelligence Analysis for Upper Basement Region for Fault Identification in Decatur, Illinois. In Proceedings of the SEG/AAPG International Meeting for Applied Geoscience & Energy, Houston, TX, USA, 26–29 August 2024. [Google Scholar]
  26. Zhang, T.; Lin, Y.; Liu, K.H.; Alhakeem, A.; Gao, S.S. Fault Visualization Enhancement Using Ant Tracking Technique and Its Application in the Taranaki Basin, New Zealand. In Proceedings of the SEG International Exposition and 87th Annual Meeting, Houston, TX, USA, 24–29 September 2017. [Google Scholar]
  27. Marfurt, K.J.; Alves, T.M. Pitfalls and Limitations in Seismic Attribute Interpretation of Tectonic Features. Interpretation 2015, 3, SB5–SB15. [Google Scholar] [CrossRef]
  28. Ngeri, A.; Amakiri, A. Ant-Tracker Attributes: An Effective Approach to Enhancing Fault Identification and Interpretation. IOSR J. VLSI Signal Process. 2015, 5, 2319–4200. [Google Scholar]
  29. Basir, H.M.; Javaherian, A.; Yaraki, M.T. Multi-Attribute Ant-Tracking and Neural Network for Fault Detection: A Case Study of an Iranian Oilfield. J. Geophys. Eng. 2013, 10, 015009. [Google Scholar] [CrossRef]
  30. Acuña-Uribe, M.; Pico-Forero, M.C.; Goyes-Peñafiel, P.; Mateus, D. Enhanced Ant Tracking: Using a Multispectral Seismic Attribute Workflow to Improve 3D Fault Detection. Lead. Edge 2021, 40, 502–512. [Google Scholar] [CrossRef]
  31. Pedersen, S.I.; Skov, T.; Randen, T.; Sønneland, L. Automatic Fault Extraction Using Artificial Ants. Mathematical Methods and Modelling in Hydrocarbon Exploration and Production. Math. Ind. 2005, 7, 107–116. [Google Scholar]
  32. Johnson, K.S. Geologic Evolution of the Anadarko Basin. Okla. Geol. Surv. 1989, 90, 3–12. [Google Scholar]
  33. Perry, W.J. Tectonic Evolution of the Anadarko Basin Region, Oklahoma; U.S. Geological Survey Bulletin 1866-A; US Geological Survey: Reston, VA, USA, 1989; pp. A1–A19. [Google Scholar] [CrossRef]
  34. Northcutt, R.A.; Campbell, J.A. Geologic Provinces of Oklahoma. AAPG Bull. 1995, 79. Available online: https://www.osti.gov/biblio/159937 (accessed on 27 December 2025).
  35. Johnson, K.S.; Luza, K.V. Earth Sciences and Mineral Resources of Oklahoma; Oklahoma Geological Survey; Educational Publication 9: Norman, OK, USA, 2008. [Google Scholar]
  36. Ball, M.M.; Henry, M.E.; Frezon, S.E. Petroleum Geology of the Anadarko Basin Region, Province (115), Kansas, Oklahoma, and Texas; Open-File Report; US Geological Survey: Reston, VA, USA, 1991. [Google Scholar] [CrossRef]
  37. Turko, M.; Mitra, S. Macroscopic Structural Styles in the Southeastern Anadarko Basin, Southern Oklahoma. Mar. Pet. Geol. 2021, 125, 104863. [Google Scholar] [CrossRef]
  38. Symcox, C.; Philp, R.P. Heterogeneity of STACK/SCOOP Production in the Anadarko Basin, Oklahoma–Geochemistry of Produced Oils. In Proceedings of the SPE/AAPG/SEG Unconventional Resources Technology Conference, Denver, CO, USA, 22 July 2019; pp. 1249–1262. [Google Scholar]
  39. Fritz, R.D.; Mitchell, J.R. The Anadarko “Super” Basin: 10 Key Characteristics to Understand Its Productivity. Bulletin 2021, 105, 1199–1231. [Google Scholar] [CrossRef]
  40. Miller, J.C.; Pranter, M.J.; Cullen, A.B. Regional Stratigraphy and Organic Richness of the Mississippian Meramec and Associated Strata, Anadarko Basin, Central Oklahoma. Shale Shaker 2019, 70, 50–79. [Google Scholar]
  41. Childress, M.; Grammer, G.M. Characteristics of Debris Flows and Outrunner Blocks—Evidence for Mississippian Deposition on a Distally Steepened Ramp. In Mississippian Reservoirs of the Midcontinent; Grammer, G.M., Gregg, J.M., Puckette, J., Jaiswal, P., Mazzullo, S.J., Pranter, M.J., Goldstein, R.H., Eds.; The American Association of Petroleum Geologists: Tulsa, OK, USA, 2019; Volume 122, ISBN 978-0-89181-399-6. [Google Scholar]
  42. Gaillot, P.; Brewer, T.; Pezard, P.; Yeh, E.-C. Borehole Imaging Tools—Principles and Applications. Sci. Drill. 2007, 5, 1–4. [Google Scholar] [CrossRef]
  43. Bedell, J.; Ismail, A.; Grammer, M.; Puckette, J.; Mewafy, F.; Boykin, B. Multiscale Fractures, Stress and Reservoir Quality Characterization: The Mississippian Meramec and Osage Intervals, STACK Play, Central Oklahoma. J. Appl. Geophys. 2023, 219, 105244. [Google Scholar] [CrossRef]
  44. Petrel, version 2023.3. Software for Analyzing Subsurface Data. Slb: Houston, TX, USA, 2023.
  45. Sarhan, M.; Safa, M. Application of Seismic Attributes for Detecting Different Geologic Features within Kafr El Sheikh Formation, Temsah Concession, Nile Delta Basin. Sci. J. Damietta Fac. Sci. 2017, 7, 26–34. [Google Scholar] [CrossRef]
  46. Chopra, S.; Marfurt, K. Curvature Attribute Applications to 3D Surface Seismic Data. Lead. Edge 2007, 26, 404–414. [Google Scholar] [CrossRef]
  47. Chopra, S.; Marfurt, K. Seismic Curvature Attributes for Mapping Faults/Fractures, and Other Stratigraphic Features. CSEG Rec. Mag. 2007, 32, 37–41. [Google Scholar]
  48. Al-Dossary, S.; Marfurt, K.J. 3D Volumetric Multispectral Estimates of Reflector Curvature and Rotation. Geophysics 2006, 71, P41–P51. [Google Scholar] [CrossRef]
  49. Chopra, S. Interpreting Fractures through 3D Seismic Discontinuity Attributes and Their Visualization. CSEG Rec. 2009, 34, 5–14. [Google Scholar]
  50. Albesher, Z.; Kellogg, J.; Hafiza, I.; Saeid, E. Multi-Attribute Analysis Using Coherency and Ant-Tracking Techniques for Fault and Fracture Detection in La Florida Anticline, Llanos Foothills, Colombia. Geosciences 2020, 10, 154. [Google Scholar] [CrossRef]
  51. Cox, T. Ant Tracking Seismic Volumes for Automated Fault Interpretation. In Proceedings of the CSPG CSEG Convention, Calgary, AB, Canada, 14–17 May 2007. [Google Scholar]
Geosciences 16 00033 g001
Figure 1. Location of the study area in the Anadarko Basin, a major sedimentary basin in the cratonic interior of North America (modified after [34,35]).
Figure 1. Location of the study area in the Anadarko Basin, a major sedimentary basin in the cratonic interior of North America (modified after [34,35]).
Geosciences 16 00033 g001
Geosciences 16 00033 g002
Figure 2. Generalized stratigraphy of the Anadarko Basin (modified after [39]). Black dots represent petroleum pay zones; wavy lines indicate unconformities, and vertical bars represent their duration. The cross symbols denote Precambrian basement rocks composed of crystalline igneous and metamorphic units.
Figure 2. Generalized stratigraphy of the Anadarko Basin (modified after [39]). Black dots represent petroleum pay zones; wavy lines indicate unconformities, and vertical bars represent their duration. The cross symbols denote Precambrian basement rocks composed of crystalline igneous and metamorphic units.
Geosciences 16 00033 g002
Geosciences 16 00033 g003
Figure 3. (a) Footprint of 3D seismic survey showing the location of Well-D and seismic section A–B. (b) Seismic time section along profile A–B intersecting Well-D, showing well tops corresponding to the Meramec, Osage, Woodford, and Hunton formations and corresponding horizons shown in dotted lines. (TWT (ms)—two-way travel time (milliseconds)).
Figure 3. (a) Footprint of 3D seismic survey showing the location of Well-D and seismic section A–B. (b) Seismic time section along profile A–B intersecting Well-D, showing well tops corresponding to the Meramec, Osage, Woodford, and Hunton formations and corresponding horizons shown in dotted lines. (TWT (ms)—two-way travel time (milliseconds)).
Geosciences 16 00033 g003
Geosciences 16 00033 g004
Figure 4. Workflow for generating multi-attribute ant-tracking volume and automatic fault extraction.
Figure 4. Workflow for generating multi-attribute ant-tracking volume and automatic fault extraction.
Geosciences 16 00033 g004
Geosciences 16 00033 g005
Figure 5. Time slices from ant-tracking volumes derived from (a) variance, (b) curvature, and (c) amplitude contrast attributes at same TWT (two-way trevel time), showing consistent fault/fracture framework with attribute-dependent variations in fault expression. Positive values show the presence of discontinuity. A strike-slip fault trend is highlighted with red arrows, showing variations in expression by each attribute.
Figure 5. Time slices from ant-tracking volumes derived from (a) variance, (b) curvature, and (c) amplitude contrast attributes at same TWT (two-way trevel time), showing consistent fault/fracture framework with attribute-dependent variations in fault expression. Positive values show the presence of discontinuity. A strike-slip fault trend is highlighted with red arrows, showing variations in expression by each attribute.
Geosciences 16 00033 g005
Geosciences 16 00033 g006
Figure 6. A 3D visualization of the muti-attribute ant-tracking volume. (a) Inline across volume co-rendered with seismic amplitude volume. (b) Crossline through volume. (c) Time slice through volume. Red arrows highlight selected fault trends to illustrate their continuity and coherence as they extend from the time slice into the vertical sections.
Figure 6. A 3D visualization of the muti-attribute ant-tracking volume. (a) Inline across volume co-rendered with seismic amplitude volume. (b) Crossline through volume. (c) Time slice through volume. Red arrows highlight selected fault trends to illustrate their continuity and coherence as they extend from the time slice into the vertical sections.
Geosciences 16 00033 g006
Geosciences 16 00033 g007
Figure 7. Comparison of single-attribute and multi-attribute ant-tracking results. (a) Time slice through the curvature-based ant-tracking volume. (b) Time slice at the same two-way travel time (TWT) through the weighted multi-attribute ant-tracking volume. Highlighted circles in the northern and southern parts of the time slice show that the multi-attribute volume exhibits enhanced fault/fracture continuity and reveals additional discontinuities compared to the single-attribute result.
Figure 7. Comparison of single-attribute and multi-attribute ant-tracking results. (a) Time slice through the curvature-based ant-tracking volume. (b) Time slice at the same two-way travel time (TWT) through the weighted multi-attribute ant-tracking volume. Highlighted circles in the northern and southern parts of the time slice show that the multi-attribute volume exhibits enhanced fault/fracture continuity and reveals additional discontinuities compared to the single-attribute result.
Geosciences 16 00033 g007
Geosciences 16 00033 g008
Figure 8. (a) Inline Profile-1 (N–S) showing the amplitude volume co-rendered with the multi-attribute ant-tracking volume. (b) Inline Profile-2 (N–S) through the same ant-tracking volume. (c) Time slice showing the spatial location of Profiles 1 and 2. The blue arrow on the vertical sections marks the level of the time slice. Red arrows across all panels highlight a prominent strike-slip fault zone. (TWT (ms)—two-way travel time (milliseconds)).
Figure 8. (a) Inline Profile-1 (N–S) showing the amplitude volume co-rendered with the multi-attribute ant-tracking volume. (b) Inline Profile-2 (N–S) through the same ant-tracking volume. (c) Time slice showing the spatial location of Profiles 1 and 2. The blue arrow on the vertical sections marks the level of the time slice. Red arrows across all panels highlight a prominent strike-slip fault zone. (TWT (ms)—two-way travel time (milliseconds)).
Geosciences 16 00033 g008
Geosciences 16 00033 g009
Figure 9. Three-dimensional visualization of (a) multi-attribute ant-tracking cube cropped to the Mermec formation. (b) Fault patches extracted through automatic fault extraction.
Figure 9. Three-dimensional visualization of (a) multi-attribute ant-tracking cube cropped to the Mermec formation. (b) Fault patches extracted through automatic fault extraction.
Geosciences 16 00033 g009
Geosciences 16 00033 g010
Figure 10. (a) FMI log interpretation from Well-D within the Meramec Formation, showing dominant ENE–WSW fracture orientation (modified after [43]). (b) Multi-attribute ant tracking extracted along the Meramec Formation time structure map—warm colors show high-relief areas. (c) Rose diagram of strike directions for fault patches extracted from the same ant-tracking volume in the Meramec Formation.
Figure 10. (a) FMI log interpretation from Well-D within the Meramec Formation, showing dominant ENE–WSW fracture orientation (modified after [43]). (b) Multi-attribute ant tracking extracted along the Meramec Formation time structure map—warm colors show high-relief areas. (c) Rose diagram of strike directions for fault patches extracted from the same ant-tracking volume in the Meramec Formation.
Geosciences 16 00033 g010
Table 1. Ant-tracking parameters used for different input seismic attribute volumes: variance, curvature, and amplitude contrast.
Table 1. Ant-tracking parameters used for different input seismic attribute volumes: variance, curvature, and amplitude contrast.
Ant-Tracking ParametersVariance AttributeCurvature AttributeAmplitude Contrast Attribute
Stage-1Stage-2 Stage-1Stage-2 * Stage-1Stage-2
Initial ant boundary353747
Ant-track deviation222222
Ant step size333333
Illegal steps allowed223131
Legal steps allowed223313
Stop criteria107105105
* Same parameters used on the composite volume as a final pass of ant tracking.
Table 2. Fault extraction parameters [44].
Table 2. Fault extraction parameters [44].
ParametersValueRemarks
Extraction sampling distance20Minimum distance between extraction seed points in voxels
Extraction sampling thresholdTop 10%Minimum signal level to create extraction points
Extraction background thresholdTop 30%Minimum signal level to incorporate into fault estimate
Deviation from a plane13Distance, in voxels, a fault may deviate from a plane surface fit to the data
Connectivity constraint2Voxel connectivity on one, two, or three faces to be included in the fault patch
Minimum patch size (points)100Fault patches with less than this value will be excluded
Patch downsampling (Voxels)8Controls the density of points within each fault patch
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Sreedhar, S.V.; Knapp, C.C.; Knapp, J.H. 3D Multi-Attribute Ant Tracking for Fault and Fracture Delineation—A Case Study from the Anadarko Basin. Geosciences 2026, 16, 33. https://doi.org/10.3390/geosciences16010033

AMA Style

Sreedhar SV, Knapp CC, Knapp JH. 3D Multi-Attribute Ant Tracking for Fault and Fracture Delineation—A Case Study from the Anadarko Basin. Geosciences. 2026; 16(1):33. https://doi.org/10.3390/geosciences16010033

Chicago/Turabian Style

Sreedhar, Sreejesh V., Camelia C. Knapp, and James H. Knapp. 2026. "3D Multi-Attribute Ant Tracking for Fault and Fracture Delineation—A Case Study from the Anadarko Basin" Geosciences 16, no. 1: 33. https://doi.org/10.3390/geosciences16010033

APA Style

Sreedhar, S. V., Knapp, C. C., & Knapp, J. H. (2026). 3D Multi-Attribute Ant Tracking for Fault and Fracture Delineation—A Case Study from the Anadarko Basin. Geosciences, 16(1), 33. https://doi.org/10.3390/geosciences16010033

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop