Mark Lindsay
Research Subjects
Mark Lindsay has a total of 152 co-authors. Here are the top co-authors by number of shared publications:
- Mark Jessell
Centre for Exploration Targeting / Mineral Exploration Cooperative Research Centre / DARE Centre, School of Earth Sciences, The University of Western Australia, Perth, Australia
44 shared publications - Jeremie Giraud
School of Earth Sciences, The University of Western Australia, Crawley, 6009, WA, Australia
30 shared publications - Vitaliy Ogarko
Centre for Exploration Targeting (School of Earth Sciences), The University of Western Australia, Crawley, 6009 WA, Australia
27 shared publications - Guillaume Pirot
School of Earth Sciences, The University of Western Australia, Perth, Australia
12 shared publications - Sandra A. Occhipinti
The Centre for Exploration Targeting, School of Earth Sciences, The University of Western Australia, Crawley, Western Australia, 6009, Australia
12 shared publications - Alan Robert Alexander Aitken
The University of Western Australia (Perth, Australia)
10 shared publications
Latest Publications
- Open Access
- 4 November 2026
- Published by Elsevier in SSRN Electronic Journal
3D Pseudo-Trans-Dimensional Geometrical Inversion and Posterior Ensemble Analysis for Undercover Imaging in the Boulia Region (Queensland, Australia)
Because direct observation of undercover geological structures is often impossible, mineral exploration relies on geophysical inversion to derive subsurface models of the structures of interest. However, the inverse problem is non-unique and imaging the large ensemble of geological models that can explain the observed geophysical data remains a challenge. To alleviate this, three-dimensional (3D) pseudo-trans-dimensional geometrical Markov chain Monte Carlo inversion offers a framework that allows exploring a broad range of scenarios by perturbing geological parameters. The algorithm employs a birth–death process that can add or remove rock units from the ensemble, together with random perturbations of geometry and density, sampling an ensemble of models consistent with the data, prior information, and constraints. The first field application using this approach is presented, using gravity data from a prospective exploration area in the Boulia region (Queensland, Australia). Overall, 3 million 3D models of varying complexity are generated to explore the model space. To synthesise this ensemble, an interpretation workflow is presented that combines topological analysis, feature extraction via a 3D convolutional encoder, and dimensionality reduction. Clustering of the resulting features identifies model families, enabling a compact visualisation of the sample. Results indicate the persistence of one or two ‘birthed’ units absent from the a priori geological model: relatively shallow (locally ~1km depth), N–S-elongated bodies of higher density intruding into horizontal layers. These features suggest previously unrecognised intrusive units and demonstrate the potential of the proposed approach to identify exploration targets concealed beneath cover.
- 4 November 2026
- Published by Elsevier in SSRN Electronic Journal
Open Access3D Pseudo-Trans-Dimensional Geometrical Inversion and Posterior Ensemble Analysis for Undercover Imaging in the Boulia Region (Queensland, Australia)
Because direct observation of undercover geological structures is often impossible, mineral exploration relies on geophysical inversion to derive subsurface models of the structures of interest. However, the inverse problem is non-unique and imaging the large ensemble of geological models that can explain the observed geophysical data remains a challenge. To alleviate this, three-dimensional (3D) pseudo-trans-dimensional geometrical Markov chain Monte Carlo inversion offers a framework that allows exploring a broad range of scenarios by perturbing geological parameters. The algorithm employs a birth–death process that can add or remove rock units from the ensemble, together with random perturbations of geometry and density, sampling an ensemble of models consistent with the data, prior information, and constraints. The first field application using this approach is presented, using gravity data from a prospective exploration area in the Boulia region (Queensland, Australia). Overall, 3 million 3D models of varying complexity are generated to explore the model space. To synthesise this ensemble, an interpretation workflow is presented that combines topological analysis, feature extraction via a 3D convolutional encoder, and dimensionality reduction. Clustering of the resulting features identifies model families, enabling a compact visualisation of the sample. Results indicate the persistence of one or two ‘birthed’ units absent from the a priori geological model: relatively shallow (locally ~1km depth), N–S-elongated bodies of higher density intruding into horizontal layers. These features suggest previously unrecognised intrusive units and demonstrate the potential of the proposed approach to identify exploration targets concealed beneath cover.
- Open Access
- 17 September 2026
- Published by Pensoft Publishers in Subterranean Biology
- Vol. 58, 57-91
From order to species: Modelling Amphipoda distributions in the Hamersley Range
Groundwater amphipods in the Pilbara region of Western Australia are under pressure from the impacts of mining operations, particularly from groundwater extraction. To help inform conservation planning for this subterranean group, we model their distribution using a suite of environmental variables and occurrence data, and a Random Forest (RF) classification method. Given the taxonomic uncertainty and cryptic diversity in groundwater amphipods, we compared model results across nested taxonomic datasets to assess the potential benefits and limitations of higher-level taxonomic modelling. In this analysis, we compared all amphipods (Order: Amphipoda), the Paramelitidae family, which comprises 92.2% of filtered amphipod records used in the model, and the species B08 Paramelitidae SOLOMON 1, which accounts for 40.36% of filtered Paramelitidae records. We investigated which environmental factors most strongly influenced patterns in fauna distribution to better understand habitat requirements. Species probability maps offer a geospatial view of species distribution based on model predictions. A range of metrics were used to evaluate models in terms of predictive performance. Comparing the three models, the Amphipoda model achieved higher accuracy (80.4%) than the Paramelitidae model (68.3%), and the B08 Paramelitidae SOLOMON 1 model (69.8%), but lower sensitivity (77.8% vs 84.9% vs 100%), implying that the latter may be more appropriate when detecting species presence is the priority over species absence. Variable importance analysis revealed factors such as the gradient of the landscape, distance to the nearest dyke, and seasonal precipitation played important roles in shaping the distribution of each taxonomic group. The results demonstrate the ability of RF to reliably predict the presence of taxa alongside a set of informative metrics. Our results suggest that order-level modelling may retain some broad-scale ecological patterns observed at the species level under data-limited scenarios. Such information aids the development of effective conservation strategies. The analytical framework developed in this study is adaptable for use in other underground systems, facilitating further research on subterranean environments and adding a vital perspective to conservation decision-making.
- 17 September 2026
- Published by Pensoft Publishers in Subterranean Biology
- Vol. 58, 57-91
Open AccessFrom order to species: Modelling Amphipoda distributions in the Hamersley Range
Groundwater amphipods in the Pilbara region of Western Australia are under pressure from the impacts of mining operations, particularly from groundwater extraction. To help inform conservation planning for this subterranean group, we model their distribution using a suite of environmental variables and occurrence data, and a Random Forest (RF) classification method. Given the taxonomic uncertainty and cryptic diversity in groundwater amphipods, we compared model results across nested taxonomic datasets to assess the potential benefits and limitations of higher-level taxonomic modelling. In this analysis, we compared all amphipods (Order: Amphipoda), the Paramelitidae family, which comprises 92.2% of filtered amphipod records used in the model, and the species B08 Paramelitidae SOLOMON 1, which accounts for 40.36% of filtered Paramelitidae records. We investigated which environmental factors most strongly influenced patterns in fauna distribution to better understand habitat requirements. Species probability maps offer a geospatial view of species distribution based on model predictions. A range of metrics were used to evaluate models in terms of predictive performance. Comparing the three models, the Amphipoda model achieved higher accuracy (80.4%) than the Paramelitidae model (68.3%), and the B08 Paramelitidae SOLOMON 1 model (69.8%), but lower sensitivity (77.8% vs 84.9% vs 100%), implying that the latter may be more appropriate when detecting species presence is the priority over species absence. Variable importance analysis revealed factors such as the gradient of the landscape, distance to the nearest dyke, and seasonal precipitation played important roles in shaping the distribution of each taxonomic group. The results demonstrate the ability of RF to reliably predict the presence of taxa alongside a set of informative metrics. Our results suggest that order-level modelling may retain some broad-scale ecological patterns observed at the species level under data-limited scenarios. Such information aids the development of effective conservation strategies. The analytical framework developed in this study is adaptable for use in other underground systems, facilitating further research on subterranean environments and adding a vital perspective to conservation decision-making.
- Open Access
- 9 April 2026
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 245 (3)
Spatial uncertainty constraints reduce overfitting for potential field geophysical inversion
SUMMARY: Understanding the internal structure of the Earth is achieved using geophysical data and inversion is a powerful mathematical technique used by resource explorers to do so. Inherent ambiguity means that an infinite number of petrophysical models exist that can explain the geophysical data, so constraints such as geological models and petrophysical data have been employed to reduce the solution space. The constraints, like the data, are subject to noise and error, resulting in uncertainty propagating to the final model because inversion is designed to use the algorithm and constraints to find the single ‘best’ solution. Current practice assumes the best solution is found by optimizing for the lowest misfit between the data and model; however, if the data is uncertain, the model fit to that data is likewise uncertain and potentially misrepresentative. Optimizing misfit also means that inversion is subject to overfitting. Overfitting occurs when a model achieves the lowest misfit values by inadvertently fitting to data noise. Overfitting inversion occurs when the model has too many free parameters with no constraints, resulting in near-surface anomalies that can be mistakenly identified as legitimate targets for exploration rather than model artefacts. This contribution describes the use of spatial uncertainty calculated from geophysical data, providing free parameter constraints to reduce overfitting for geophysical inversion. The spatial uncertainty estimate is taken from a geostatistical model calculated using Integrated Nested Laplacian Approximation. A region in the East Kimberley, northern Western Australia, is subject to gravity inversion using Tomofast-x, an open-source inversion platform. Inversion is conducted using different configurations. Inversion is run without spatial uncertainty constraints, as is current practice, and then with spatial uncertainty constraints to test their effect on the resulting petrophysical model. The geostatistical model offers different percentiles from the geophysical model representing the extrema of estimated gravimetry values in the 10th and 90th percentiles. Inversions are run using these ‘extrema’ alongside the current practice of using the 50th percentile (or ‘mean’) gravity models as the observed field. Examination of inversion using and not using spatial uncertainty constraints shows that overfitting can be reduced. Using the extrema percentiles as the observed field has lesser benefits to reduce overfitting.
- 9 April 2026
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 245 (3)
Open AccessSpatial uncertainty constraints reduce overfitting for potential field geophysical inversion
SUMMARY: Understanding the internal structure of the Earth is achieved using geophysical data and inversion is a powerful mathematical technique used by resource explorers to do so. Inherent ambiguity means that an infinite number of petrophysical models exist that can explain the geophysical data, so constraints such as geological models and petrophysical data have been employed to reduce the solution space. The constraints, like the data, are subject to noise and error, resulting in uncertainty propagating to the final model because inversion is designed to use the algorithm and constraints to find the single ‘best’ solution. Current practice assumes the best solution is found by optimizing for the lowest misfit between the data and model; however, if the data is uncertain, the model fit to that data is likewise uncertain and potentially misrepresentative. Optimizing misfit also means that inversion is subject to overfitting. Overfitting occurs when a model achieves the lowest misfit values by inadvertently fitting to data noise. Overfitting inversion occurs when the model has too many free parameters with no constraints, resulting in near-surface anomalies that can be mistakenly identified as legitimate targets for exploration rather than model artefacts. This contribution describes the use of spatial uncertainty calculated from geophysical data, providing free parameter constraints to reduce overfitting for geophysical inversion. The spatial uncertainty estimate is taken from a geostatistical model calculated using Integrated Nested Laplacian Approximation. A region in the East Kimberley, northern Western Australia, is subject to gravity inversion using Tomofast-x, an open-source inversion platform. Inversion is conducted using different configurations. Inversion is run without spatial uncertainty constraints, as is current practice, and then with spatial uncertainty constraints to test their effect on the resulting petrophysical model. The geostatistical model offers different percentiles from the geophysical model representing the extrema of estimated gravimetry values in the 10th and 90th percentiles. Inversions are run using these ‘extrema’ alongside the current practice of using the 50th percentile (or ‘mean’) gravity models as the observed field. Examination of inversion using and not using spatial uncertainty constraints shows that overfitting can be reduced. Using the extrema percentiles as the observed field has lesser benefits to reduce overfitting.
- Open Access
- 13 March 2026
- Published by Copernicus GmbH
Fostering the ethical use of Artificial Intelligence in the Geosciences
A set of practical and actionable recommendations for the ethical application of Artificial Intelligence (AI) in the geosciences is presented by the Task Group on AI Ethics in Geosciences under the IUGS Commission on Geoethics. While geoscientists have long used statistical and machine learning methods, the rapid adoption of frontier and generative AI introduces amplified risks alongside opportunities for scientific discovery and productivity. AI holds immense potential to support the UN Sustainable Development Goals (SDGs), for example by predicting natural resource locations, enhancing understanding of deep geological time, and assisting with natural hazard prediction such as earthquakes and landslides.However, the rapid development and deployment of AI, combined with high-profile ethical concerns, demands clear, actionable guidance. Current high-level ethical frameworks, such as those from UNESCO, lack the practical detail needed for implementation in the geosciences. This report addresses that gap by providing guidance for academic, industry, governmental, and non-governmental geoscientists, society leaders, and policymakers.The methodology adopted a robust approach informed by Critical Realism—encouraging attention to hidden structures and power dynamics—and Virtue Ethics—focusing on the character of a “good and wise geoscientist.” The investigation combined a literature review, longitudinal analysis of deployed AI systems, and counterfactual future thinking, all triangulated against the UNESCO Recommendations on AI Ethics and the International Science Council (ISC) AI Analytical Framework.Eight key themes were identified to address current and anticipated ethical challenges in the geosciences. These recommendations aim to foster a responsible, just, and sustainable integration of AI that serves the public good and upholds scientific integrity:Use AI Responsibly: Treat AI as a tool to support, not replace, geoscientist judgment, avoiding fully autonomous decisions that impact people or ecosystems.Promote Transparency and Explainability: Ensure research is open, traceable, and reproducible, with clear disclosure of data sources, limitations, and uncertainties, particularly for “black box” models.Consider Bias and Fairness: Use diverse, representative datasets and actively address biases that could affect marginalized or Indigenous communities.Obtain Informed Consent and Protect Personal Data: Explicit consent is required for AI training data, and a privacy-by-design approach should be applied, especially for sensitive information.Practice Participatory Design and Community Engagement: Engage meaningfully with communities affected by AI outputs, following the principle “Nothing about us without us.”Advocate for Environmental Protection: Weigh the environmental costs of AI (energy, water, e-waste) against its scientific benefits, promoting sustainable practices such as energy-efficient algorithms.Integrity in Science, Publishing, and Education: Disclose AI use in research, verify AI-generated assertions, and ensure AI does not undermine critical thinking or scientific honesty.Consider Geopolitics: International institutions should remain neutral, avoiding endorsement of cloud platforms that centralize data and risk eroding data sovereignty or reinforcing inequities (“algorithmic colonization”).The report (https://www.geoethics.org/_files/ugd/5195a5_5dcf66f87cca492c958319c3f4cdeffb.pdf) proposes a high-level roadmap for continuous improvement, including practical ethical impact and risk assessments. These recommendations serve as a call to action to safeguard geosciences and ensure responsible stewardship of the Earth.
- 13 March 2026
- Published by Copernicus GmbH
Open AccessFostering the ethical use of Artificial Intelligence in the Geosciences
A set of practical and actionable recommendations for the ethical application of Artificial Intelligence (AI) in the geosciences is presented by the Task Group on AI Ethics in Geosciences under the IUGS Commission on Geoethics. While geoscientists have long used statistical and machine learning methods, the rapid adoption of frontier and generative AI introduces amplified risks alongside opportunities for scientific discovery and productivity. AI holds immense potential to support the UN Sustainable Development Goals (SDGs), for example by predicting natural resource locations, enhancing understanding of deep geological time, and assisting with natural hazard prediction such as earthquakes and landslides.However, the rapid development and deployment of AI, combined with high-profile ethical concerns, demands clear, actionable guidance. Current high-level ethical frameworks, such as those from UNESCO, lack the practical detail needed for implementation in the geosciences. This report addresses that gap by providing guidance for academic, industry, governmental, and non-governmental geoscientists, society leaders, and policymakers.The methodology adopted a robust approach informed by Critical Realism—encouraging attention to hidden structures and power dynamics—and Virtue Ethics—focusing on the character of a “good and wise geoscientist.” The investigation combined a literature review, longitudinal analysis of deployed AI systems, and counterfactual future thinking, all triangulated against the UNESCO Recommendations on AI Ethics and the International Science Council (ISC) AI Analytical Framework.Eight key themes were identified to address current and anticipated ethical challenges in the geosciences. These recommendations aim to foster a responsible, just, and sustainable integration of AI that serves the public good and upholds scientific integrity:Use AI Responsibly: Treat AI as a tool to support, not replace, geoscientist judgment, avoiding fully autonomous decisions that impact people or ecosystems.Promote Transparency and Explainability: Ensure research is open, traceable, and reproducible, with clear disclosure of data sources, limitations, and uncertainties, particularly for “black box” models.Consider Bias and Fairness: Use diverse, representative datasets and actively address biases that could affect marginalized or Indigenous communities.Obtain Informed Consent and Protect Personal Data: Explicit consent is required for AI training data, and a privacy-by-design approach should be applied, especially for sensitive information.Practice Participatory Design and Community Engagement: Engage meaningfully with communities affected by AI outputs, following the principle “Nothing about us without us.”Advocate for Environmental Protection: Weigh the environmental costs of AI (energy, water, e-waste) against its scientific benefits, promoting sustainable practices such as energy-efficient algorithms.Integrity in Science, Publishing, and Education: Disclose AI use in research, verify AI-generated assertions, and ensure AI does not undermine critical thinking or scientific honesty.Consider Geopolitics: International institutions should remain neutral, avoiding endorsement of cloud platforms that centralize data and risk eroding data sovereignty or reinforcing inequities (“algorithmic colonization”).The report (https://www.geoethics.org/_files/ugd/5195a5_5dcf66f87cca492c958319c3f4cdeffb.pdf) proposes a high-level roadmap for continuous improvement, including practical ethical impact and risk assessments. These recommendations serve as a call to action to safeguard geosciences and ensure responsible stewardship of the Earth.
- Open Access
- 6 August 2025
- Published by Elsevier in Next Research
- Vol. 2 (3)
Machine learning to assess troglofauna occurrences in the northern part of Western Australia
- 6 August 2025
- Published by Elsevier in Next Research
- Vol. 2 (3)
Open AccessMachine learning to assess troglofauna occurrences in the northern part of Western Australia
- Open Access
- 13 June 2025
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Machine Learning and Computation
- Vol. 2 (2)
TopoRSNet: A Unique Approach to Maintaining Topological Features for Digital Rock Image Rescaling With Minimal Quality Degradation
The composition, microstructure, and physical behavior of rock at the core‐to‐pore scale are commonly visualized and characterized using 3D X‐ray microcomputed tomography (micro‐CT). However, a key geophysical problem in micro‐CT rock characterization is how to efficiently handle and process large numbers of high‐resolution 3D rock images, and how to preserve multiscale features for characterization at coarse resolution. Typically, a single high‐resolution 3D micro‐CT image can be extremely large with more than voxels, and in the case of dynamic scans, hundreds of such large images will be generated, posing challenges in storage, downstream analysis, modeling, and their application to deep learning. Existing image rescaling methods face challenges such as information loss during down‐sampling and limited feature recovery during up‐sampling. Herein, we present TopoRSNet, an image rescaling network designed to adjust micro‐CT images to an optimal size and resolution while preserving global and local representative features. Feature preservation is ensured by introducing three feature‐based loss functions: adversarial loss, a new feature consistency loss, and a novel persistent homology loss (PH loss). Combined with a pixel consistency loss, we assure the preservation of both pixels and features during rescaling. The efficacy of TopoRSNet is validated on two common geological rocks, and the results are compared to other rescaling methods. This method enables scalable analysis of multiscale rock features, allowing for broader integration of 3D imaging in geoscientific modeling, simulation, and machine learning workflows.
- 13 June 2025
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Machine Learning and Computation
- Vol. 2 (2)
Open AccessTopoRSNet: A Unique Approach to Maintaining Topological Features for Digital Rock Image Rescaling With Minimal Quality Degradation
The composition, microstructure, and physical behavior of rock at the core‐to‐pore scale are commonly visualized and characterized using 3D X‐ray microcomputed tomography (micro‐CT). However, a key geophysical problem in micro‐CT rock characterization is how to efficiently handle and process large numbers of high‐resolution 3D rock images, and how to preserve multiscale features for characterization at coarse resolution. Typically, a single high‐resolution 3D micro‐CT image can be extremely large with more than voxels, and in the case of dynamic scans, hundreds of such large images will be generated, posing challenges in storage, downstream analysis, modeling, and their application to deep learning. Existing image rescaling methods face challenges such as information loss during down‐sampling and limited feature recovery during up‐sampling. Herein, we present TopoRSNet, an image rescaling network designed to adjust micro‐CT images to an optimal size and resolution while preserving global and local representative features. Feature preservation is ensured by introducing three feature‐based loss functions: adversarial loss, a new feature consistency loss, and a novel persistent homology loss (PH loss). Combined with a pixel consistency loss, we assure the preservation of both pixels and features during rescaling. The efficacy of TopoRSNet is validated on two common geological rocks, and the results are compared to other rescaling methods. This method enables scalable analysis of multiscale rock features, allowing for broader integration of 3D imaging in geoscientific modeling, simulation, and machine learning workflows.
- Open Access
- 6 June 2025
- Published by Geoinformatics International in International Journal of Geoinformatics
- Vol. 21 (5) , 37-61
Constructing a Structured Corpus from Geoscience Literature: A Case Study using Western Australia Iron and Lithium Deposits
International Journal of GEOINFORMATICS
- 6 June 2025
- Published by Geoinformatics International in International Journal of Geoinformatics
- Vol. 21 (5) , 37-61
Open AccessConstructing a Structured Corpus from Geoscience Literature: A Case Study using Western Australia Iron and Lithium Deposits
International Journal of GEOINFORMATICS
- Open Access
- 3 June 2025
- Published by Copernicus GmbH
- Vol. 2025, 1-30
Comment on egusphere-2024-3754
Geophysical inversion is an important tool for characterising the structure of the Earth. The utility of geophysical inversion has led to widespread adoption by resource explorers, and used to adapt gravity, magnetic, seismic and electrical datasets into petrophysical models that can be used for targeting. However, inherent ambiguity means that an infinite number of petrophysical models exist that can explain the geophysical data, so constraints such as geological models and petrophysical data have been employed to reduce the solution space. The constraints, like the data, are subject to noise and error resulting in uncertainty propagating to the final model. This is because inversion is designed to use the algorithm and constraints to find the ‘best’ solution by optimising the lowest misfit between the data and model. If the data is uncertain, the model fit to that data is likewise uncertain, and misrepresentative. Optimising misfit also means that inversion is subject to overfitting. Overfitting is when the lowest misfit values are attained by fitting the model to data noise. Overfitting inversion can create anomalies in the near-surface that can be mistakenly identified as legitimate targets for exploration rather than possible model artefacts. This contribution describes the use of spatial error constraints calculated from geophysical data to reduce overfitting for geophysical inversion. The spatial error estimate is derived from a geostatistical model calculated using Integrated Nested Laplacian Approximation (INLA). A region in the East Kimberley, northern Western Australia, is subject to gravity inversion using Tomofast-x, an open-source inversion platform. Inversion using different percentiles from the geophysical model explores whether the extrema of gravimetry values should be considered to explore the model space. Examination of inversion using and not using spatial error constraints shows that overfitting reduction can be achieved while using different percentiles as the observed field has lesser benefits.
- 3 June 2025
- Published by Copernicus GmbH
- Vol. 2025, 1-30
Open AccessComment on egusphere-2024-3754
Geophysical inversion is an important tool for characterising the structure of the Earth. The utility of geophysical inversion has led to widespread adoption by resource explorers, and used to adapt gravity, magnetic, seismic and electrical datasets into petrophysical models that can be used for targeting. However, inherent ambiguity means that an infinite number of petrophysical models exist that can explain the geophysical data, so constraints such as geological models and petrophysical data have been employed to reduce the solution space. The constraints, like the data, are subject to noise and error resulting in uncertainty propagating to the final model. This is because inversion is designed to use the algorithm and constraints to find the ‘best’ solution by optimising the lowest misfit between the data and model. If the data is uncertain, the model fit to that data is likewise uncertain, and misrepresentative. Optimising misfit also means that inversion is subject to overfitting. Overfitting is when the lowest misfit values are attained by fitting the model to data noise. Overfitting inversion can create anomalies in the near-surface that can be mistakenly identified as legitimate targets for exploration rather than possible model artefacts. This contribution describes the use of spatial error constraints calculated from geophysical data to reduce overfitting for geophysical inversion. The spatial error estimate is derived from a geostatistical model calculated using Integrated Nested Laplacian Approximation (INLA). A region in the East Kimberley, northern Western Australia, is subject to gravity inversion using Tomofast-x, an open-source inversion platform. Inversion using different percentiles from the geophysical model explores whether the extrema of gravimetry values should be considered to explore the model space. Examination of inversion using and not using spatial error constraints shows that overfitting reduction can be achieved while using different percentiles as the observed field has lesser benefits.
- Open Access
- 12 April 2025
- Published by Elsevier in Geosystems and Geoenvironment
- Vol. 4 (3)
Lineament extraction and structural mapping using Landsat-9 OLI and Sentinel-1 SAR data in the Proterozoic North Singhbhum Mobile Belt, Eastern India
- 12 April 2025
- Published by Elsevier in Geosystems and Geoenvironment
- Vol. 4 (3)
Open AccessLineament extraction and structural mapping using Landsat-9 OLI and Sentinel-1 SAR data in the Proterozoic North Singhbhum Mobile Belt, Eastern India