Hersh Gilbert
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
Research Subjects
Hersh Gilbert has a total of 134 co-authors. Here are the top co-authors by number of shared publications:
George Zandt
13 shared publications- Craig H. Jones
University of Colorado Boulder
9 shared publications Megan Anderson
6 shared publicationsThomas J. Owens
Department of Geological Sciences, University of South Carolina, 701 Sumter Street, EWS 617, Columbia, South Carolina 29208, USA
6 shared publicationsPatricia Alvarado
5 shared publicationsStephen Marshak
5 shared publications
Latest Publications
- Open Access
- 13 November 2026
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 53 (22)
Hydraulic Uncertainty Complicates Estimating How Much Injection Wells Individually Contribute to Fault Pressure Loading Near Peace River, Alberta
In multiwell saltwater disposal systems, induced earthquakes may reflect overlapping pressure changes, complicating assessment of individual wells' contributions. Near Peace River, Alberta, multiple wells inject into the Leduc Formation, a fringing carbonate reef complex, including wells active for decades approximately from the earthquake cluster center. Yet previous studies linked the 2022 sequence to a well about away. We test whether its contribution to pressure averaged over a basement fault region exceeds that of all other wells combined. To address hydraulic uncertainty, we ran a diffusion ensemble using recorded injection histories from 15 wells and varying reservoir and fault permeabilities. Its contribution exceeded all others combined under some, but not all, sampled hydraulic conditions. Faster pressure transmission allowed decades of distant injection to prevent that well from dominating. Distant wells therefore warrant consideration as potential contributors to fault pressure loading when regional pressure transmission is poorly constrained.
- 13 November 2026
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 53 (22)
Open AccessHydraulic Uncertainty Complicates Estimating How Much Injection Wells Individually Contribute to Fault Pressure Loading Near Peace River, Alberta
In multiwell saltwater disposal systems, induced earthquakes may reflect overlapping pressure changes, complicating assessment of individual wells' contributions. Near Peace River, Alberta, multiple wells inject into the Leduc Formation, a fringing carbonate reef complex, including wells active for decades approximately from the earthquake cluster center. Yet previous studies linked the 2022 sequence to a well about away. We test whether its contribution to pressure averaged over a basement fault region exceeds that of all other wells combined. To address hydraulic uncertainty, we ran a diffusion ensemble using recorded injection histories from 15 wells and varying reservoir and fault permeabilities. Its contribution exceeded all others combined under some, but not all, sampled hydraulic conditions. Faster pressure transmission allowed decades of distant injection to prevent that well from dominating. Distant wells therefore warrant consideration as potential contributors to fault pressure loading when regional pressure transmission is poorly constrained.
- Open Access
- 14 January 2026
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 244 (3)
A Fourier neural operator surrogate model for nonlinear electrical resistivity tomography
SUMMARY: Electrical resistivity tomography (ERT) is used to infer the subsurface resistivity structure. ERT requires solving a nonlinear inverse problem that is often approximated as linear to reduce computational time. However, the approximation requires assumptions that cause limitations for the data analysis. Most of the computational time is due to the forward problem that requires solving the Poisson equation. Recently, similar forward problems have been shown to be replaceable with a surrogate model of lower computational cost. We present a geoelectric surrogate based on Fourier neural operators and demonstrate a successful application in nonlinear inversion. The standard deviation of the surrogate prediction error for unseen samples is $<$5 per cent. Furthermore, the surrogate reduces computational time by over three orders of magnitude per realization, enabling ERT for previously intractable settings. We apply the surrogate in Markov chain Monte Carlo inversion of simulated data. The results resolve sharp resistivity changes with plausible uncertainties.
- 14 January 2026
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 244 (3)
Open AccessA Fourier neural operator surrogate model for nonlinear electrical resistivity tomography
SUMMARY: Electrical resistivity tomography (ERT) is used to infer the subsurface resistivity structure. ERT requires solving a nonlinear inverse problem that is often approximated as linear to reduce computational time. However, the approximation requires assumptions that cause limitations for the data analysis. Most of the computational time is due to the forward problem that requires solving the Poisson equation. Recently, similar forward problems have been shown to be replaceable with a surrogate model of lower computational cost. We present a geoelectric surrogate based on Fourier neural operators and demonstrate a successful application in nonlinear inversion. The standard deviation of the surrogate prediction error for unseen samples is $<$5 per cent. Furthermore, the surrogate reduces computational time by over three orders of magnitude per realization, enabling ERT for previously intractable settings. We apply the surrogate in Markov chain Monte Carlo inversion of simulated data. The results resolve sharp resistivity changes with plausible uncertainties.
- Open Access
- 2 July 2025
- Published by Wiley
A Fourier Neural Operator Surrogate Model for nonlinear Electrical Resistivity Tomography
Electrical resistivity tomography (ERT) is used to infer the subsurface resistivity structure. ERT requires solving a nonlinear inverse problem that is often approximated as linear to reduce computational time. However, the approximation requires assumptions that cause limitations for the data analysis. Most of the computational time is due to the forward problem that requires solving the Poisson equation. Recently, similar forward problems have been shown to be replaceable with a surrogate model of lower computational cost. We present a geoelectric surrogate based on Fourier Neural Operators (FNO) and demonstrate a successful application in nonlinear inversion. The standard deviation of Surrogate prediction errors for unseen samples are $<$5\%. Furthermore, the surrogate reduces computational time by over three orders of magnitude per realization, enabling ERT for previously intractable settings. We apply the surrogate in Markov chain Monte Carlo (MCMC) inversion of simulated data. The results resolve sharp resistivity changes with plausible uncertainties.
- 2 July 2025
- Published by Wiley
Open AccessA Fourier Neural Operator Surrogate Model for nonlinear Electrical Resistivity Tomography
Electrical resistivity tomography (ERT) is used to infer the subsurface resistivity structure. ERT requires solving a nonlinear inverse problem that is often approximated as linear to reduce computational time. However, the approximation requires assumptions that cause limitations for the data analysis. Most of the computational time is due to the forward problem that requires solving the Poisson equation. Recently, similar forward problems have been shown to be replaceable with a surrogate model of lower computational cost. We present a geoelectric surrogate based on Fourier Neural Operators (FNO) and demonstrate a successful application in nonlinear inversion. The standard deviation of Surrogate prediction errors for unseen samples are $<$5\%. Furthermore, the surrogate reduces computational time by over three orders of magnitude per realization, enabling ERT for previously intractable settings. We apply the surrogate in Markov chain Monte Carlo (MCMC) inversion of simulated data. The results resolve sharp resistivity changes with plausible uncertainties.
- Open Access
- 5 May 2025
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 242 (1)
Multidimensional reconstruction of noise correlation functions and its application in improving surface-wave inversion
SUMMARY: The cross-correlation of the ambient noise recordings, also known as noise correlation functions (NCFs), can converge to Green’s functions (GFs) which describe wave propagation between a pair of stations. However, the NCFs are often biased from the true GFs due to the presence of random noise and spurious arrivals arising from non-diffuse wavefields. Additionally, the limited spatial and temporal coverage of recording stations can lead to large data gaps in the retrieved virtual shot gathers, particularly at large interstation distances (far offsets). Both these factors impose great challenges to retrieving high-quality NCFs and conducting reliable subsurface imaging. In this study, we propose a multidimensional (4-D) reconstruction method to compensate for the insufficient station coverage and simultaneously attenuate incoherent noise in the NCFs. We test the feasibility of the proposed method using a dense seismic array deployed in western Canada. Our results demonstrate that the reconstructed virtual common midpoint gather can greatly improve the stability and reliability of the surface-wave dispersion measurements and subsequent shear velocity inversions compared to the conventional approaches. The proposed ambient noise processing framework enables us to construct accurate 3-D velocity model of the subsurface.
- 5 May 2025
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 242 (1)
Open AccessMultidimensional reconstruction of noise correlation functions and its application in improving surface-wave inversion
SUMMARY: The cross-correlation of the ambient noise recordings, also known as noise correlation functions (NCFs), can converge to Green’s functions (GFs) which describe wave propagation between a pair of stations. However, the NCFs are often biased from the true GFs due to the presence of random noise and spurious arrivals arising from non-diffuse wavefields. Additionally, the limited spatial and temporal coverage of recording stations can lead to large data gaps in the retrieved virtual shot gathers, particularly at large interstation distances (far offsets). Both these factors impose great challenges to retrieving high-quality NCFs and conducting reliable subsurface imaging. In this study, we propose a multidimensional (4-D) reconstruction method to compensate for the insufficient station coverage and simultaneously attenuate incoherent noise in the NCFs. We test the feasibility of the proposed method using a dense seismic array deployed in western Canada. Our results demonstrate that the reconstructed virtual common midpoint gather can greatly improve the stability and reliability of the surface-wave dispersion measurements and subsequent shear velocity inversions compared to the conventional approaches. The proposed ambient noise processing framework enables us to construct accurate 3-D velocity model of the subsurface.
- Open Access
- 2 May 2025
- Published by Wiley
A Fourier Neural Operator Surrogate Model for nonlinear Electrical Resistivity Tomograph
Original Article Electrical resistivity tomography (ERT) is used to infer the subsurface resistivity structure. ERT requires solving a nonlinear inverse problem that is often approximated as linear to reduce computational time. However, the approximation requires assumptions that cause limitations for the data analysis. Most of the computational time is due to the forward problem that requires solving the Poisson equation. Recently, similar forward problems have been shown to be replaceable with a surrogate model of lower computational cost. We present a geoelectric surrogate based on Fourier Neural Operators (FNO) and demonstrate a successful application in nonlinear inversion. The standard deviation of Surrogate prediction errors for unseen samples are <5%. Furthermore, the surrogate reduces computational time by over two orders of magnitude, enabling ERT for previously intractable settings. We apply the surrogate in Markov chain Monte Carlo (MCMC) inversion of simulated data. The results resolve sharp resistivity changes with plausible uncertainties.
- 2 May 2025
- Published by Wiley
Open AccessA Fourier Neural Operator Surrogate Model for nonlinear Electrical Resistivity Tomograph
Original Article Electrical resistivity tomography (ERT) is used to infer the subsurface resistivity structure. ERT requires solving a nonlinear inverse problem that is often approximated as linear to reduce computational time. However, the approximation requires assumptions that cause limitations for the data analysis. Most of the computational time is due to the forward problem that requires solving the Poisson equation. Recently, similar forward problems have been shown to be replaceable with a surrogate model of lower computational cost. We present a geoelectric surrogate based on Fourier Neural Operators (FNO) and demonstrate a successful application in nonlinear inversion. The standard deviation of Surrogate prediction errors for unseen samples are <5%. Furthermore, the surrogate reduces computational time by over two orders of magnitude, enabling ERT for previously intractable settings. We apply the surrogate in Markov chain Monte Carlo (MCMC) inversion of simulated data. The results resolve sharp resistivity changes with plausible uncertainties.
- Open Access
- 15 October 2024
- Published by Canadian Science Publishing in Canadian Journal of Earth Sciences
Crustal stress near the Yakutat microplate collision from probabilistic earthquake focal mechanisms
Earthquake source characteristics provide a valuable constraint on crustal stress and regional plate tectonics. Earthquake source studies in Yukon and northern British Columbia have been limited by sparse seismic network coverage. In this work, we leverage recent seismic network improvements to estimate focal mechanisms for small- and moderate-magnitude ( M > 2.0) earthquakes from P-wave first-motion polarity data. We invert these data within a probabilistic framework to rigorously quantify mechanism uncertainties. Subsequently, probabilistic earthquake focal mechanisms are used as input for inversions for the orientation of principal stress axes, and stress shape ratio, for spatial windows throughout the region. We implement a novel, data-driven approach to propagate focal mechanism uncertainty in stress inversion. Our results improve the spatial coverage of existing earthquake focal mechanisms and enable an orogenic-scale study of crustal stress near the Yakutat–North America collision. Overall, the region is characterized by a transpressive stress regime. Maximum horizontal compressive stress orientations exhibit a pattern orthogonal to the Yakutat collision syntaxis. Stress inversion results also reveal an abrupt change in orientation east-to-west, which we interpret as a change in stress regime across the Fairweather–Connector–Totschunda fault system that is likely related to coupling between the subducting Yakutat slab and North America. This work improves our understanding of potential earthquake hazards in southwestern Yukon and the surrounding region.
- 15 October 2024
- Published by Canadian Science Publishing in Canadian Journal of Earth Sciences
Open AccessCrustal stress near the Yakutat microplate collision from probabilistic earthquake focal mechanisms
Earthquake source characteristics provide a valuable constraint on crustal stress and regional plate tectonics. Earthquake source studies in Yukon and northern British Columbia have been limited by sparse seismic network coverage. In this work, we leverage recent seismic network improvements to estimate focal mechanisms for small- and moderate-magnitude ( M > 2.0) earthquakes from P-wave first-motion polarity data. We invert these data within a probabilistic framework to rigorously quantify mechanism uncertainties. Subsequently, probabilistic earthquake focal mechanisms are used as input for inversions for the orientation of principal stress axes, and stress shape ratio, for spatial windows throughout the region. We implement a novel, data-driven approach to propagate focal mechanism uncertainty in stress inversion. Our results improve the spatial coverage of existing earthquake focal mechanisms and enable an orogenic-scale study of crustal stress near the Yakutat–North America collision. Overall, the region is characterized by a transpressive stress regime. Maximum horizontal compressive stress orientations exhibit a pattern orthogonal to the Yakutat collision syntaxis. Stress inversion results also reveal an abrupt change in orientation east-to-west, which we interpret as a change in stress regime across the Fairweather–Connector–Totschunda fault system that is likely related to coupling between the subducting Yakutat slab and North America. This work improves our understanding of potential earthquake hazards in southwestern Yukon and the surrounding region.
- 8 December 2022
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Solid Earth
- Vol. 127 (12)
Monitoring CO2 Injection at the CaMI Field Research Station Using Microseismic Noise Sources
Monitoring subsurface velocity variations due to industrial activities using passive seismic imaging techniques has gained popularity in recent years. In this study, we examine the spatiotemporal variations of persistent, non‐ambient seismic noise during CO2 sequestration at the Containment and Monitoring Institute Field Research Station near Brooks, Alberta, Canada. Based on the temporal migration and power spectral density (PSD) analysis of continuous seismic records from both a dense geophone array and “X”‐shaped geophone lines operated from June to August 2019, we detect two non‐ambient local noise sources correlated with the local industrial activities during the two months: a dominant noise source (1–5 Hz) southeast of the study region and a slightly weaker noise source in the higher frequency range (5–15 Hz) around the injection well. The former noise source overlaps with the operations near a submersible disposal pump. The substantial diurnal variations in noise levels in PSD as a function of time of day, month and location are further evidence of these two noise sources. We propose that the persistent noise source around the injection well originated from the subsurface microtremors caused by the coupled interaction between the injected CO2 fluid and formation rocks. The proposed methods based on passive microseismic noise offer a potentially valuable strategy for long‐term evaluation of the safety of CO2 sequestration, which can be extended to future integrity monitoring of underground energy storage.
- 8 December 2022
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Solid Earth
- Vol. 127 (12)
Monitoring CO2 Injection at the CaMI Field Research Station Using Microseismic Noise Sources
Monitoring subsurface velocity variations due to industrial activities using passive seismic imaging techniques has gained popularity in recent years. In this study, we examine the spatiotemporal variations of persistent, non‐ambient seismic noise during CO2 sequestration at the Containment and Monitoring Institute Field Research Station near Brooks, Alberta, Canada. Based on the temporal migration and power spectral density (PSD) analysis of continuous seismic records from both a dense geophone array and “X”‐shaped geophone lines operated from June to August 2019, we detect two non‐ambient local noise sources correlated with the local industrial activities during the two months: a dominant noise source (1–5 Hz) southeast of the study region and a slightly weaker noise source in the higher frequency range (5–15 Hz) around the injection well. The former noise source overlaps with the operations near a submersible disposal pump. The substantial diurnal variations in noise levels in PSD as a function of time of day, month and location are further evidence of these two noise sources. We propose that the persistent noise source around the injection well originated from the subsurface microtremors caused by the coupled interaction between the injected CO2 fluid and formation rocks. The proposed methods based on passive microseismic noise offer a potentially valuable strategy for long‐term evaluation of the safety of CO2 sequestration, which can be extended to future integrity monitoring of underground energy storage.
- Open Access
- 6 October 2021
- Published by American Geophysical Union (AGU) in Tectonics
- Vol. 40 (10)
Resolving the Cenozoic History of Rock Exhumation Along the Central Rocky Mountain Trench Using Apatite Low‐Temperature Thermochronology
The Rocky Mountain Trench (RMT) extends through the Canadian Cordillera from the Yukon to northwest Montana. The RMT's impressive length and continuous nature have been interpreted to be the result of faulting, and localized erosion, along a continent‐scale zone of crustal weakness located beneath the RMT, possibly associated with the ancient continental margin. Despite the continuous nature of the RMT, its kinematics vary along strike from dextral strike‐slip faults in the north to normal faults in the south. The central RMT is thought to be the transition between these two structural zones. We use low‐temperature thermochronology to compare the cooling histories across the central RMT in eastern British Columbia. We report apatite fission track ages from 23 samples and apatite (U‐Th)/He ages from 25 samples along with thermal‐history models. Our results reveal three phases of rapid cooling that followed the Cretaceous‐Paleocene cordilleran thrusting in the Eocene, the early–mid Miocene, and from the late Miocene to recent. We find that the Malton Gneiss Complex exhumed as a horst structure 20–10 Ma bounded by the North Thompson Albreda Fault and the RMT. Normal faulting along the NTAF continued in the late Miocene as well as west side down normal faulting along the RMT. Our data suggest that extends along the southern RMT continued northward to at least this portion of the central RMT in multiple episodes during the Cenozoic. We suggest transtension occurred in our study area that was driven by orogenic collapse and lithospheric mantle delamination.
- 6 October 2021
- Published by American Geophysical Union (AGU) in Tectonics
- Vol. 40 (10)
Open AccessResolving the Cenozoic History of Rock Exhumation Along the Central Rocky Mountain Trench Using Apatite Low‐Temperature Thermochronology
The Rocky Mountain Trench (RMT) extends through the Canadian Cordillera from the Yukon to northwest Montana. The RMT's impressive length and continuous nature have been interpreted to be the result of faulting, and localized erosion, along a continent‐scale zone of crustal weakness located beneath the RMT, possibly associated with the ancient continental margin. Despite the continuous nature of the RMT, its kinematics vary along strike from dextral strike‐slip faults in the north to normal faults in the south. The central RMT is thought to be the transition between these two structural zones. We use low‐temperature thermochronology to compare the cooling histories across the central RMT in eastern British Columbia. We report apatite fission track ages from 23 samples and apatite (U‐Th)/He ages from 25 samples along with thermal‐history models. Our results reveal three phases of rapid cooling that followed the Cretaceous‐Paleocene cordilleran thrusting in the Eocene, the early–mid Miocene, and from the late Miocene to recent. We find that the Malton Gneiss Complex exhumed as a horst structure 20–10 Ma bounded by the North Thompson Albreda Fault and the RMT. Normal faulting along the NTAF continued in the late Miocene as well as west side down normal faulting along the RMT. Our data suggest that extends along the southern RMT continued northward to at least this portion of the central RMT in multiple episodes during the Cenozoic. We suggest transtension occurred in our study area that was driven by orogenic collapse and lithospheric mantle delamination.
- 21 April 2021
- Published by Seismological Society of America (SSA) in Seismological Research Letters
- Vol. 92 (5) , 3087-3099
Structure and Dynamics of the Southern Rocky Mountain Trench near Valemount, British Columbia, Inferred from Local Seismicity
Stretching nearly the extent of the Canadian Cordillera, the Rocky Mountain trench (RMT) forms one of the longest valleys on Earth. Yet, the level of seismicity, and style of faulting, on the RMT remains poorly known. We assess earthquakes in the southern RMT using a temporary network of seismometers around Valemount, British Columbia, and identify active structures using a probabilistic earthquake catalog spanning from September 2017 to August 2018. Together with results from earlier geological and seismic studies, our new earthquake catalog provides a constraint on the geometry of subsurface faults and their level of activity during a year of recording. The tectonic analysis presented here benefits from the catalog of 47 earthquakes, including robust horizontal and vertical uncertainty quantification. The westward dip of the southern RMT fault is one of the prominent subsurface structures that we observe. The seismicity observed here occurs on smaller surrounding faults away from the RMT and shifts from the east to the west of the trench from north to south of Valemount. The change in distribution of earthquakes follows changes in the style of deformation along the length of the RMT. Focal mechanisms calculated for two earthquakes with particularly clear waveforms reveal northeast–southwest‐oriented thrusting. The seismicity reveals a change in the pattern of deformation from narrowly focused transpression north of Valemount to more broadly distributed activity in an area characterized by normal faulting to the south. Six sets of repeating events detected here produce similar waveforms whose P waves exhibit correlation coefficients that exceed 0.7 and may result from the migration of fluids through the fractured crust.
- 21 April 2021
- Published by Seismological Society of America (SSA) in Seismological Research Letters
- Vol. 92 (5) , 3087-3099
Structure and Dynamics of the Southern Rocky Mountain Trench near Valemount, British Columbia, Inferred from Local Seismicity
Stretching nearly the extent of the Canadian Cordillera, the Rocky Mountain trench (RMT) forms one of the longest valleys on Earth. Yet, the level of seismicity, and style of faulting, on the RMT remains poorly known. We assess earthquakes in the southern RMT using a temporary network of seismometers around Valemount, British Columbia, and identify active structures using a probabilistic earthquake catalog spanning from September 2017 to August 2018. Together with results from earlier geological and seismic studies, our new earthquake catalog provides a constraint on the geometry of subsurface faults and their level of activity during a year of recording. The tectonic analysis presented here benefits from the catalog of 47 earthquakes, including robust horizontal and vertical uncertainty quantification. The westward dip of the southern RMT fault is one of the prominent subsurface structures that we observe. The seismicity observed here occurs on smaller surrounding faults away from the RMT and shifts from the east to the west of the trench from north to south of Valemount. The change in distribution of earthquakes follows changes in the style of deformation along the length of the RMT. Focal mechanisms calculated for two earthquakes with particularly clear waveforms reveal northeast–southwest‐oriented thrusting. The seismicity reveals a change in the pattern of deformation from narrowly focused transpression north of Valemount to more broadly distributed activity in an area characterized by normal faulting to the south. Six sets of repeating events detected here produce similar waveforms whose P waves exhibit correlation coefficients that exceed 0.7 and may result from the migration of fluids through the fractured crust.
- 1 January 2020
- Published by Elsevier in Journal of South American Earth Sciences
- Vol. 97
Lithospheric structure of the Pampean flat slab region from double-difference tomography
- 1 January 2020
- Published by Elsevier in Journal of South American Earth Sciences
- Vol. 97
Lithospheric structure of the Pampean flat slab region from double-difference tomography