Ernst Schetselaar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ernst Schetselaar has a total of 102 co-authors. Here are the top co-authors by number of shared publications:
- Gilles Bellefleur
Geological Survey of Canada, 601 Booth St., Ottawa, ON, Canada
13 shared publications - Don White
Natural Resources Canada
10 shared publications - Eric A. de Kemp
Geological Survey of Canada, Three-dimensional Earth Imaging and Modelling Lab, 601 Booth Street, Ottawa, Canada, K0E 1E9
10 shared publications - Michael Hillier
Geological Survey of Canada
5 shared publications - Christopher J. M. Lawley
Natural Resources Canada
4 shared publications J A Craven
Natural Resources Canada
4 shared publications
Latest Publications
- 1 January 2026
- Published by Natural Resources Canada/CMSS/Information Management
Multicomponent geospatial data for predictive mapping of Canadian critical minerals
The NRCan Open S&T Repository (OSTR) provides free and open access to science publication and research funded by the Canadian government.
- 1 January 2026
- Published by Natural Resources Canada/CMSS/Information Management
Multicomponent geospatial data for predictive mapping of Canadian critical minerals
The NRCan Open S&T Repository (OSTR) provides free and open access to science publication and research funded by the Canadian government.
- 1 January 2026
- Published by Natural Resources Canada/CMSS/Information Management
Données géospatiales à composantes multiples pour la cartographie prédictive des minéraux critiques canadiens
The NRCan Open S&T Repository (OSTR) provides free and open access to science publication and research funded by the Canadian government.
- 1 January 2026
- Published by Natural Resources Canada/CMSS/Information Management
Données géospatiales à composantes multiples pour la cartographie prédictive des minéraux critiques canadiens
The NRCan Open S&T Repository (OSTR) provides free and open access to science publication and research funded by the Canadian government.
- Open Access
- 1 April 2024
- Published by Canadian Science Publishing in Canadian Journal of Earth Sciences
- Vol. 61 (4) , 431-445
Exploring the Moho beneath the Northern Canadian Cordillera, with seismically constrained gravity inversion
The geometry and depth of the Moho beneath the Northern Canadian Cordillera and adjacent North American craton are modeled through the application of both cokriging and 3-D inversion of gravity data, integrated with sparse seismic depth estimates. Models require a regional density correction, with lower densities beneath the Cordillera than the craton. The lower densities are primarily attributed to a reduction in upper mantle density, ascribed to thermal expansion under regionally higher temperatures. The eastern margin of this low-density zone is broadly aligned with the rapid westward shallowing of the lithospheric–asthenospheric boundary. From the Cordillera to the North American craton, the Moho is broadly flat at a depth of ∼32 km. A zone of deeper Moho (up to ∼38 km) beneath the Mistry Creek embayment has a modeled mantle density that is of a colder cratonic signature, akin to the Mackenzie craton, and is interpreted to represent the preserved remnants of an old rift basin that is a local focus of the diffuse seismicity. Southeast of the Fort Norman structure, seismicity is broadly focused along the eastern edge of the low-density zone. Major structures such as the Denali and Tintina faults, with 100’s km of right-lateral displacement, separate zones of higher and lower upper mantle density, supporting the interpretation of their continuation into the upper mantle. Within the North American craton, upper mantle density steadily increases toward the Great Bear magmatic zone but increases more rapidly beneath the Slave craton in tangent with a deepening of the Moho.
- 1 April 2024
- Published by Canadian Science Publishing in Canadian Journal of Earth Sciences
- Vol. 61 (4) , 431-445
Open AccessExploring the Moho beneath the Northern Canadian Cordillera, with seismically constrained gravity inversion
The geometry and depth of the Moho beneath the Northern Canadian Cordillera and adjacent North American craton are modeled through the application of both cokriging and 3-D inversion of gravity data, integrated with sparse seismic depth estimates. Models require a regional density correction, with lower densities beneath the Cordillera than the craton. The lower densities are primarily attributed to a reduction in upper mantle density, ascribed to thermal expansion under regionally higher temperatures. The eastern margin of this low-density zone is broadly aligned with the rapid westward shallowing of the lithospheric–asthenospheric boundary. From the Cordillera to the North American craton, the Moho is broadly flat at a depth of ∼32 km. A zone of deeper Moho (up to ∼38 km) beneath the Mistry Creek embayment has a modeled mantle density that is of a colder cratonic signature, akin to the Mackenzie craton, and is interpreted to represent the preserved remnants of an old rift basin that is a local focus of the diffuse seismicity. Southeast of the Fort Norman structure, seismicity is broadly focused along the eastern edge of the low-density zone. Major structures such as the Denali and Tintina faults, with 100’s km of right-lateral displacement, separate zones of higher and lower upper mantle density, supporting the interpretation of their continuation into the upper mantle. Within the North American craton, upper mantle density steadily increases toward the Great Bear magmatic zone but increases more rapidly beneath the Slave craton in tangent with a deepening of the Moho.
- 10 October 2023
- Published by Society of Exploration Geophysicists in Geophysics
- Vol. 89 (1) , WA279-WA294
A deep learning benchmark for first break detection from hardrock seismic reflection data
Deep learning techniques are used to tackle a variety of tasks related to seismic data processing and interpretation. Although many works have shown the benefits of deep learning, assessing the generalization capabilities of proposed methods for data acquired in different conditions and geologic environments remains challenging. This is especially true for applications in hardrock environments. The primary factors that impede the adoption of machine learning in geosciences include the lack of publicly available and labeled data sets and the use of inadequate evaluation methodologies. Because machine learning models are prone to overfit and underperform when the data used to train them are site specific, the applicability of these models on new survey data that could be considered “out-of-distribution” is rarely addressed. This is unfortunate, as evaluating predictive models in out-of-distribution settings can provide a good insight into their usefulness in real-world use cases. To tackle these issues, we develop a simple benchmarking methodology for first break picking to evaluate the transferability of deep learning models that are trained across different environments and acquisition conditions. For this, we consider a reflection seismic survey data set acquired at five distinct hardrock mining sites combined with annotations for first break picking. We train and evaluate a baseline deep learning solution based on a U-Net for future comparisons and discuss potential improvements to this approach.
- 10 October 2023
- Published by Society of Exploration Geophysicists in Geophysics
- Vol. 89 (1) , WA279-WA294
A deep learning benchmark for first break detection from hardrock seismic reflection data
Deep learning techniques are used to tackle a variety of tasks related to seismic data processing and interpretation. Although many works have shown the benefits of deep learning, assessing the generalization capabilities of proposed methods for data acquired in different conditions and geologic environments remains challenging. This is especially true for applications in hardrock environments. The primary factors that impede the adoption of machine learning in geosciences include the lack of publicly available and labeled data sets and the use of inadequate evaluation methodologies. Because machine learning models are prone to overfit and underperform when the data used to train them are site specific, the applicability of these models on new survey data that could be considered “out-of-distribution” is rarely addressed. This is unfortunate, as evaluating predictive models in out-of-distribution settings can provide a good insight into their usefulness in real-world use cases. To tackle these issues, we develop a simple benchmarking methodology for first break picking to evaluate the transferability of deep learning models that are trained across different environments and acquisition conditions. For this, we consider a reflection seismic survey data set acquired at five distinct hardrock mining sites combined with annotations for first break picking. We train and evaluate a baseline deep learning solution based on a U-Net for future comparisons and discuss potential improvements to this approach.
- 1 January 2023
- Published by Natural Resources Canada/CMSS/Information Management
Initiating transformative geoscience practice at the Geological Survey of Canada: Canada in 3D
Application of 3D technologies to the wide range of Geosciences knowledge domains is well underway. These have been operationalized in workflows of the hydrocarbon sector for a half-century, and now in mining for over two decades. In Geosciences, algorithms, structured workflows and data integration strategies can support compelling Earth models, however challenges remain to meet the standards of geological plausibility required for most geoscientific studies. There is also missing links in the institutional information infrastructure supporting operational multi-scale 3D data and model development. Canada in 3D (C3D) is a vision and road map for transforming the Geological Survey of Canada's (GSC) work practice by leveraging emerging 3D technologies. Primarily the transformation from 2D geological mapping, to a well-structured 3D modelling practice that is both data-driven and knowledge-driven. It is tempting to imagine that advanced 3D computational methods, coupled with Artificial Intelligence and Big Data tools will automate the bulk of this process. To effectively apply these methods there is a need, however, for data to be in a well-organized, classified, georeferenced (3D) format embedded with key information, such as spatial-temporal relations, and earth process knowledge. Another key challenge for C3D is the relative infancy of 3D geoscience technologies for geological inference and 3D modelling using sparse and heterogeneous regional geoscience information, while preserving the insights and expertise of geoscientists maintaining scientific integrity of digital products. In most geological surveys, there remains considerable educational and operational challenges to achieve this balance of digital automation and expert knowledge. Emerging from the last two decades of research are more efficient workflows, transitioning from cumbersome, explicit (manual) to reproducible implicit semi-automated methods. They are characterized by integrated and iterative, forward and reverse geophysical modelling, coupled with stratigraphic and structural approaches. The full impact of research and development with these 3D tools, geophysical-geological integration and simulation approaches is perhaps unpredictable, but the expectation is that they will produce predictive, instructive models of Canada's geology that will be used to educate, prioritize and influence sustainable policy for stewarding our natural resources. On the horizon are 3D geological modelling methods spanning the gulf between local and frontier or green-fields, as well as deep crustal characterization. These are key components of mineral systems understanding, integrated and coupled hydrological modelling and energy transition applications, e.g. carbon sequestration, in-situ hydrogen mining, and geothermal exploration. Presented are some case study examples at a range of scales from our efforts in C3D.
- 1 January 2023
- Published by Natural Resources Canada/CMSS/Information Management
Initiating transformative geoscience practice at the Geological Survey of Canada: Canada in 3D
Application of 3D technologies to the wide range of Geosciences knowledge domains is well underway. These have been operationalized in workflows of the hydrocarbon sector for a half-century, and now in mining for over two decades. In Geosciences, algorithms, structured workflows and data integration strategies can support compelling Earth models, however challenges remain to meet the standards of geological plausibility required for most geoscientific studies. There is also missing links in the institutional information infrastructure supporting operational multi-scale 3D data and model development. Canada in 3D (C3D) is a vision and road map for transforming the Geological Survey of Canada's (GSC) work practice by leveraging emerging 3D technologies. Primarily the transformation from 2D geological mapping, to a well-structured 3D modelling practice that is both data-driven and knowledge-driven. It is tempting to imagine that advanced 3D computational methods, coupled with Artificial Intelligence and Big Data tools will automate the bulk of this process. To effectively apply these methods there is a need, however, for data to be in a well-organized, classified, georeferenced (3D) format embedded with key information, such as spatial-temporal relations, and earth process knowledge. Another key challenge for C3D is the relative infancy of 3D geoscience technologies for geological inference and 3D modelling using sparse and heterogeneous regional geoscience information, while preserving the insights and expertise of geoscientists maintaining scientific integrity of digital products. In most geological surveys, there remains considerable educational and operational challenges to achieve this balance of digital automation and expert knowledge. Emerging from the last two decades of research are more efficient workflows, transitioning from cumbersome, explicit (manual) to reproducible implicit semi-automated methods. They are characterized by integrated and iterative, forward and reverse geophysical modelling, coupled with stratigraphic and structural approaches. The full impact of research and development with these 3D tools, geophysical-geological integration and simulation approaches is perhaps unpredictable, but the expectation is that they will produce predictive, instructive models of Canada's geology that will be used to educate, prioritize and influence sustainable policy for stewarding our natural resources. On the horizon are 3D geological modelling methods spanning the gulf between local and frontier or green-fields, as well as deep crustal characterization. These are key components of mineral systems understanding, integrated and coupled hydrological modelling and energy transition applications, e.g. carbon sequestration, in-situ hydrogen mining, and geothermal exploration. Presented are some case study examples at a range of scales from our efforts in C3D.
- 1 January 2022
- Published by Natural Resources Canada/CMSS/Information Management
Initiating transformative geoscience practice at the Geological Survey of Canada: Canada in 3D
Application of 3D technologies to the wide range of Geosciences knowledge domains is well underway. These have been operationalized in workflows of the hydrocarbon sector for a half-century, and now in mining for over two decades. In Geosciences, algorithms, structured workflows and data integration strategies can support compelling Earth models, however challenges remain to meet the standards of geological plausibility required for most geoscientific studies. There is also missing links in the institutional information infrastructure supporting operational multi-scale 3D data and model development. Canada in 3D (C3D) is a vision and road map for transforming the Geological Survey of Canada's (GSC) work practice by leveraging emerging 3D technologies. Primarily the transformation from 2D geological mapping, to a well-structured 3D modelling practice that is both data-driven and knowledge-driven. It is tempting to imagine that advanced 3D computational methods, coupled with Artificial Intelligence and Big Data tools will automate the bulk of this process. To effectively apply these methods there is a need, however, for data to be in a well-organized, classified, georeferenced (3D) format embedded with key information, such as spatial-temporal relations, and earth process knowledge. Another key challenge for C3D is the relative infancy of 3D geoscience technologies for geological inference and 3D modelling using sparse and heterogeneous regional geoscience information, while preserving the insights and expertise of geoscientists maintaining scientific integrity of digital products. In most geological surveys, there remains considerable educational and operational challenges to achieve this balance of digital automation and expert knowledge. Emerging from the last two decades of research are more efficient workflows, transitioning from cumbersome, explicit (manual) to reproducible implicit semi-automated methods. They are characterized by integrated and iterative, forward and reverse geophysical modelling, coupled with stratigraphic and structural approaches. The full impact of research and development with these 3D tools, geophysical-geological integration and simulation approaches is perhaps unpredictable, but the expectation is that they will produce predictive, instructive models of Canada's geology that will be used to educate, prioritize and influence sustainable policy for stewarding our natural resources. On the horizon are 3D geological modelling methods spanning the gulf between local and frontier or green-fields, as well as deep crustal characterization. These are key components of mineral systems understanding, integrated and coupled hydrological modelling and energy transition applications, e.g. carbon sequestration, in-situ hydrogen mining, and geothermal exploration. Presented are some case study examples at a range of scales from our efforts in C3D.
- 1 January 2022
- Published by Natural Resources Canada/CMSS/Information Management
Initiating transformative geoscience practice at the Geological Survey of Canada: Canada in 3D
Application of 3D technologies to the wide range of Geosciences knowledge domains is well underway. These have been operationalized in workflows of the hydrocarbon sector for a half-century, and now in mining for over two decades. In Geosciences, algorithms, structured workflows and data integration strategies can support compelling Earth models, however challenges remain to meet the standards of geological plausibility required for most geoscientific studies. There is also missing links in the institutional information infrastructure supporting operational multi-scale 3D data and model development. Canada in 3D (C3D) is a vision and road map for transforming the Geological Survey of Canada's (GSC) work practice by leveraging emerging 3D technologies. Primarily the transformation from 2D geological mapping, to a well-structured 3D modelling practice that is both data-driven and knowledge-driven. It is tempting to imagine that advanced 3D computational methods, coupled with Artificial Intelligence and Big Data tools will automate the bulk of this process. To effectively apply these methods there is a need, however, for data to be in a well-organized, classified, georeferenced (3D) format embedded with key information, such as spatial-temporal relations, and earth process knowledge. Another key challenge for C3D is the relative infancy of 3D geoscience technologies for geological inference and 3D modelling using sparse and heterogeneous regional geoscience information, while preserving the insights and expertise of geoscientists maintaining scientific integrity of digital products. In most geological surveys, there remains considerable educational and operational challenges to achieve this balance of digital automation and expert knowledge. Emerging from the last two decades of research are more efficient workflows, transitioning from cumbersome, explicit (manual) to reproducible implicit semi-automated methods. They are characterized by integrated and iterative, forward and reverse geophysical modelling, coupled with stratigraphic and structural approaches. The full impact of research and development with these 3D tools, geophysical-geological integration and simulation approaches is perhaps unpredictable, but the expectation is that they will produce predictive, instructive models of Canada's geology that will be used to educate, prioritize and influence sustainable policy for stewarding our natural resources. On the horizon are 3D geological modelling methods spanning the gulf between local and frontier or green-fields, as well as deep crustal characterization. These are key components of mineral systems understanding, integrated and coupled hydrological modelling and energy transition applications, e.g. carbon sequestration, in-situ hydrogen mining, and geothermal exploration. Presented are some case study examples at a range of scales from our efforts in C3D.
- 1 January 2022
- Published by Natural Resources Canada/CMSS/Information Management
3D regional scale modelling of the Flin Flon-Glennie Complex: preparatory data analysis and preliminary results, Manitoba and Saskatchewan
GEOSCAN supports the mission of Natural Resources Canada through the timely access to information.
- 1 January 2022
- Published by Natural Resources Canada/CMSS/Information Management
3D regional scale modelling of the Flin Flon-Glennie Complex: preparatory data analysis and preliminary results, Manitoba and Saskatchewan
GEOSCAN supports the mission of Natural Resources Canada through the timely access to information.
- 1 January 2022
- Published by Natural Resources Canada/CMSS/Information Management
Three-dimensional magnetotelluric modelling of the Lalor volcanogenic massive-sulfide deposit, Manitoba
Unconstrained magnetotelluric inversion commonly produces insufficient inherent resolution to image ore-system fluid pathways that were structurally thinned during post-emplacement tectonic activity. To improve the resolution in these complex environments, we synthesized the 3-D magnetotelluric (MT) response for geologically realistic models using a finite-element-based forward-modelling tool with unstructured meshes and applied it to the Lalor volcanogenic massive-sulfide deposit in the Snow Lake mining camp, Manitoba. This new tool is based on mapping interpolated or simulated resistivity values from wireline logs onto unstructured tetrahedral meshes to reflect, with the help of 3-D models obtained from lithostratigraphic and lithofacies drillhole logs, the complexity of the host-rock geological structure. The resulting stochastic model provides a more realistic representation of the heterogeneous spatial distribution of the electric resistivity values around the massive, stringer, and disseminated sulfide ore zones. Both models were combined into one seamless tetrahedral mesh of the resistivity field. To capture the complex resistivity distribution in the geophysical forward model, a finite-element code was developed. Comparative analyses of the forward models with MT data acquired at the Earth's surface show a reasonable agreement that explains the regional variations associated with the host rock geological structure and detects the local anomalies associated with the MT response of the ore zones.
- 1 January 2022
- Published by Natural Resources Canada/CMSS/Information Management
Three-dimensional magnetotelluric modelling of the Lalor volcanogenic massive-sulfide deposit, Manitoba
Unconstrained magnetotelluric inversion commonly produces insufficient inherent resolution to image ore-system fluid pathways that were structurally thinned during post-emplacement tectonic activity. To improve the resolution in these complex environments, we synthesized the 3-D magnetotelluric (MT) response for geologically realistic models using a finite-element-based forward-modelling tool with unstructured meshes and applied it to the Lalor volcanogenic massive-sulfide deposit in the Snow Lake mining camp, Manitoba. This new tool is based on mapping interpolated or simulated resistivity values from wireline logs onto unstructured tetrahedral meshes to reflect, with the help of 3-D models obtained from lithostratigraphic and lithofacies drillhole logs, the complexity of the host-rock geological structure. The resulting stochastic model provides a more realistic representation of the heterogeneous spatial distribution of the electric resistivity values around the massive, stringer, and disseminated sulfide ore zones. Both models were combined into one seamless tetrahedral mesh of the resistivity field. To capture the complex resistivity distribution in the geophysical forward model, a finite-element code was developed. Comparative analyses of the forward models with MT data acquired at the Earth's surface show a reasonable agreement that explains the regional variations associated with the host rock geological structure and detects the local anomalies associated with the MT response of the ore zones.