LSLisa Stright

Lisa Stright

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Lisa Stright has a total of 79 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Integrating field-based and digital datasets to improve geometric accuracy of geologic cross sections constructed from outcrop

    by Thomas G. E. Peploe, Paul R. Nesbit, Rebecca G. Englert, Miquel Poyatos-Moré, Brian W. Romans, Lisa Stright, Stephen M. Hubbard

    Outcrop exposures of sedimentary rocks preserve vital records of depositional processes and offer key insights into Earth's depositional history. Traditional field-based techniques, such as measuring bed orientations, sketches, and field photography, have long been used to document stratigraphic patterns, while modern approaches now leverage high-resolution panoramas and drone-based photogrammetry to quantify and geolocate measurements. Although Digital Outcrop Models (DOMs) derived from UAVs (drones) markedly improve our ability to capture and interpret spatial details, these modern approaches, when used to generate 2-D stratigraphic correlation panels, are still challenged by distortions caused by outcrop perspective and structural deformation. To accurately constrain stratigraphic reconstructions derived from DOMs, calibrated field data is necessary to corroborate these observations with thicknesses, length scales, and lithologies. Here, we present a method to correct stratigraphic interpretations from DOMs for structural overprint and outcrop perspective using field-based paleocurrent and strike-and-dip measurements to produce geometrically accurate surface length, height, and dip in stratigraphic correlation panels. We demonstrate this method on Cerro Cazador, a 15 km-long outcrop transect in Patagonia, Chile, recognized as containing continental shelf and shelf-margin sedimentary deposits. By systematically comparing three data projection strategies: no correction, a single correction, and a novel multi-correction workflow that applies localized adjustments to distinct structural regions, we show that the multi-correction method yields depositional geometries consistent with expected stratigraphic patterns (e.g., flat topset and bottomset deposits, sloping foresets, shelf-edge trajectories). These findings underscore the need for incorporating field-based calibrations into drone-derived models to accurately capture depositional patterns in geologically complex settings and highlight important implications for quantitative sedimentary analysis across diverse outcrop environments.

  • Understanding uncertainty in deepwater channel seismic facies classification applying random forest on outcrop-constrained 3D models and synthetic seismic data

    by Karelia La Marca, Heather Bedle, Lisa Stright, Mario Ballinas, Kurt Marfurt

    To understand the uncertainties in seismic interpretation, especially for deepwater seismic facies, we apply a novel approach using outcrop-based synthetic data as ground truth. By applying the random forest (RF) supervised machine learning method we can better understand the influence of classifier hyperparameters on facies prediction accuracy. Based on previous analysis, we chose six seismic attributes that are able to differentiate five deepwater architectural facies: shale (thin-bedded turbidites), channel axis, off-axis, margin, and mass transport deposit. Hyperparameter testing indicated that optimization of the RF classifier is sensitive to the (1) choice of training attributes, (2) original facies proportions, (3) similarity in the seismic expression of different facies, and (4) seismic data resolution and seismic data signal-to-noise ratio. A simple classifier using common RF hyperparameters developed for fluid saturation predicted the facies with 74% accuracy. Although computationally more expensive, optimizing the RF hyperparameters provided approximately 89% accuracy, demonstrating the importance of hyperparameter tuning. When applying the best model to an unseen portion of the model, the position of the channel complex set was predicted accurately, but the accuracy was only 58%. This highlights the limitations of universal models and the persistent uncertainties faced when using ML to enhance our interpretations. Although we use a correlation matrix approach, we acknowledge that uncertainties can only be assessed and not quantified. Optimization of attribute choice and hyperparameter tuning reduce the epistemic uncertainty. We encourage future research to explore this in more detail.


  • Uncertainty assessment in unsupervised machine-learning methods for deepwater channel seismic facies using outcrop-derived 3D models and synthetic seismic data

    by Karelia La Marca, Heather Bedle, Lisa Stright, Kurt Marfurt

    Unsupervised machine-learning (ML) techniques have been widely applied to analyze seismic reflection data, including the identification of seismic facies and structural features. However, interpreting the resulting clusters often relies on geoscientists’ expertise, necessitating a robustness assessment of these methods. To evaluate their reliability, synthetic data generated from an actual outcrop model are used to demonstrate how two unsupervised methods, self-organizing maps (SOMs) and generative topographic maps (GTMs), cluster deepwater channel-related seismic facies and then measure the associated error. Six seismic attributes, comprising root-mean-square amplitude, instantaneous envelope, peak magnitude, and spectral decomposition frequencies at 20, 40, and 55 Hz, served as input variables. Geobodies are assigned to each cluster formed, and the error in facies clustering is quantified by comparing the actual 3D model with the facies grouped by ML methods on a voxel-by-voxel basis. This allowed for error quantification and the computation of metrics such as F1 score and accuracy through correlation matrices. Key findings reveal that (1) GTM and SOM exhibit similar performance, with a clustering configuration of 81 for GTM slightly outperforming others. (2) Error rates are approximately 2% for the predominant facies (background shale) but significantly higher for individual channel-related facies, suggesting that channel clusters might represent multiple facies. (3) Resolution and imbalanced data distribution impact seismic facies predictability, resulting in nonuniqueness in cluster generation. (4) Using synthetic seismic data proved valuable for experimenting with different unsupervised MLs, highlighting the need for assessing uncertainty in these methods, given their implications for crucial economic decisions reliant on reservoir interpretation, modeling, and volumetric estimations.


  • Open Access

    A good figure is worth a thousand words — Part 2: A quick guide to quality illustrations

    by Lisa Stright, Balazs Nemeth, Robert Merrill, Vsevolod Egorov

  • The influence of inter- and intra-channel architecture on deep-water turbidite reservoir performance

    by Casey D. Meirovitz, Lisa Stright, Stephen M. Hubbard, Brian W. Romans

    Bed-scale heterogeneity in channelized deep-water reservoirs can significantly influence reservoir performance, but reservoir simulation typically requires cell sizes much greater than the scale of intra-channel element architecture. Here, bed- to geobody-scale simulations elucidate the influence of bed-scale architecture and channel element stacking on flow and connectivity, informing full-field reservoir model development and evaluation. Models consist of two channel element segments, each 300 m (985 ft) wide by 14 m (45 ft) thick and 550 m (1805 ft) long, stacked in 12 different stacking arrangements. Bed-scale architecture is captured in six deterministic element fills, highlighting interbedded sandstone and mudstone (thin bed) presence (homogeneous v. heterogeneous elements), position (symmetrical v. asymmetrical), and proportion (low v. high element net-to-gross). Each model is flow simulated to illuminate how element stacking and intra-element heterogeneity impacts reservoir performance. Thin bed presence and position have the greatest impact on reservoir connectivity/performance when elements are laterally offset; impacts are minimal when elements are vertically aligned. Impacts are exacerbated when the thin-bed proportion is increased. Where bed-scale architecture is represented, complex flow behaviours generate a significant variability in production timing and the cumulative volumes produced. Simulations consisting of a homogenous element architecture fail to capture complex flow behaviours, producing comparatively optimistic results.


  • Static connectivity of stacked deep-water channel elements constrained by high-resolution digital outcrop models

    by Allie Jackson, Lisa Stright, Stephen M. Hubbard, Brian W. Romans

    High-resolution digital outcrop models of stacked deep-water channel elements are constructed from the Laguna Figueroa section of the well-exposed Upper Cretaceous Tres Pasos Formation in Chilean Patagonia. The models are based on greater than 1600 m (>5250 ft) of centimeter-scale measured sections, greater than 100 paleoflow measurements, and thousands of differential global positioning system points (10-cm [4-in.] accuracy) from an outcrop belt that is approximately 2.5 km (∼1.5 mi) long and 130 m (425 ft) thick. The models elucidate the effects bed-to-geobody–scale architecture has on static sandstone connectivity among a series of stacked deep-water channel elements and how that connectivity is altered by grid cell size. Static connectivity analyses show that channel element base drapes can strongly influence sandstone connectivity and that smaller channel element widths are more likely to produce disconnected sandstone geobodies. Net-to-gross (NTG) is not directly correlated with connectivity because of the presence of thin channel element base drapes, which do not significantly contribute to NTG. Upscaling the models consistently increases channel element contact (up to 10%) but decreases sandstone connectivity (up to 2%–3%). Channel element stacking patterns strongly impact connectivity. For example, connectivity is reduced in cases of high lateral channel element offsets. Increasing drape coverage markedly decreases connectivity. Evaluating connectivity in a vertical, along-system profile is critical to understanding flow units and reservoir piping. Ultimately, this work constrains uncertainty related to the impact of subseismic-scale stratigraphic architecture on reservoir connectivity by providing concrete knowledge that can be used to guide the model building process.


  • Does fluvial channel-belt clustering predict net sand to gross rock volume? Architectural metrics and point-pattern analysis of a digital outcrop model

    by Alexander R. Koch, Cari L. Johnson, Lisa Stright

    Spatial point-pattern analyses (PPAs) are used to quantify clustering, randomness, and uniformity of the distribution of channel belts in fluvial strata. Point patterns may reflect end-member fluvial architecture, e.g., uniform compensational stacking and avulsion-generated clustering, which may change laterally, especially at greater scales. To investigate spatial and temporal changes in fluvial systems, we performed PPA and architectural analyses on extensive outcrops of the Cretaceous John Henry Member of the Straight Cliffs Formation in southern Utah, USA. Digital outcrop models (DOMs) produced using unmanned aircraft system-based stereophotogrammetry form the basis of detailed interpretations of a 250-m-thick fluvial succession over a total outcrop length of 4.5 km. The outcrops are oriented roughly perpendicular to fluvial transport direction. This transverse cross-sectional exposure of the fluvial system allows a study of the system's variation along depositional strike. We developed a workflow that examines spatial point patterns using the quadrat method, and architectural metrics such as net sand to gross rock volume (NTG), amalgamation index, and channel-belt width and thickness within moving windows. Quadrat cell sizes that are ∼ 50% of the average channel-belt width-to-thickness ratio (16:1 aspect ratio) provide an optimized scale to investigate laterally elongate distributions of fluvial-channel-belt centroids. Large-scale quadrat point patterns were recognized using an array of four quadrat cells, each with 237× greater area than the median channel belt. Large-scale point patterns and NTG correlate negatively, which is a result of using centroid-based PPA on a dataset with disparately sized channel belts. Small-scale quadrat point patterns were recognized using an array of 16 quadrat cells, each with 21× greater area than the median channel belt. Small-scale point patterns and NTG correlate positively, and match previously observed stratigraphic trends in the fluvial John Henry Member, suggesting that these are regional trends. There are deviations from these trends in architectural statistics over small distances (hundreds of meters) which are interpreted to reflect autogenic avulsion processes. Small-scale autogenic processes result in architecture that is difficult to correlate between 1D datasets, for example when characterizing a reservoir using well logs. We show that 1D NTG provides the most accurate prediction for surrounding 2D architecture.


  • The influence of stratigraphic architecture on seismic response: Reflectivity modeling of outcropping deepwater channel units

    by Erin A. L. Pemberton, Lisa Stright, Sean Fletcher, Stephen M. Hubbard

    The size, shape, stacking patterns, and internal architecture of deepwater deposits control reservoir fluid flow connectivity. Predicting deepwater stratigraphic architecture as a function of position along a deepwater slope from seismic-reflection data is critical for successful hydrocarbon exploration and development projects. Stratigraphic architecture from confined and weakly confined segments of a deepwater sediment-routing system is analyzed in outcrop from the Tres Pasos Formation (Upper Cretaceous), southern Chile. Outcrop observations are the basis of two geocellular models: confined channel deposits at Laguna Figueroa and weakly confined channel and scour deposits at Arroyo Picana. Key stratigraphic surfaces and facies relationships observed in outcrop are forward seismic modeled at high to low resolution to bridge the gap in subseismic scale interpretation of deepwater reservoirs and demonstrate challenges associated with identification of varied reservoir architecture. The outcrop-constrained geometry of architectural elements, their stacking arrangement, and the varied internal distribution of facies each impart a strong influence on seismic reflectivity. Key outcomes from the analysis include (1) stratigraphic architecture transitions down-paleoslope from vertically aligned low-aspect-ratio channel elements to a more weakly confined depocenter characterized by a breadth of laterally offset low- and high-aspect-ratio channel and scour elements. Seismic reflections, down to 30 Hz frequencies, record aspects of these stratigraphic changes. (2) Key seismic reflections are often comprised of multiple outcrop-constrained stratigraphic surfaces. Tuning effects result in composite seismic surfaces that are vertically offset from the known position of sedimentary units; this hinders accurate interpretation of stratigraphic surfaces from seismic-reflection data. This is particularly problematic in the weakly confined system in which shifted stratigraphic surfaces, which bound deposits characterized by numerous similar architectural elements, can alter the interpretability of sandstone connectivity within and across zones. Furthermore, misinterpretation of surfaces is problematic when they are flow barriers draped with debris flows, slumps/slides, or thin-bedded turbidites. (3) Tuning effects also impart significant control on volume-based interpretations from seismic data. In particular, calculations of gross rock volume from seismic reflection data that do not consider the tuning or architectural element stacking pattern can overestimate actual volumes by 10%–50%, with implications for reservoir prediction and hydrocarbon volume estimation.