PRISM :: Browsing by Author "Dettmer, Jan"

Browsing by Author "Dettmer, Jan"

Now showing 1 - 20 of 23

Open Access
Advancing Tailings Dam Performance Monitoring with Distributed Acoustic Sensing
(2024-08-30) Ouellet, Susanne; Dettmer, Jan; Hutchinson, Jean; Lato, Matthew; Pidlisecky, Adam
Tailings dams require careful design, planning and monitoring throughout their life cycle to prevent a potential failure. Such failures can occur due to a variety of causes, such as overtopping, structural and foundation conditions, static or seismic liquefaction, and erosional processes. As such, a tailings dam monitoring system should be capable of identifying potential behaviours leading up to different failure modes and with adequate warning time. This thesis investigates an emerging fiber optic sensing technology, distributed acoustic sensing (DAS), with a passive geophysical technique (coda wave interferometry; CWI), to advance monitoring of tailings dam performance. The first study provides a proof-of-concept of CWI using geophones at an active tailings dam, from June to July 2020. Following this study, a fiber optic cable located near the geophones was then employed with DAS and CWI from April to August 2021. CWI was used to infer changes in shear wave velocities up to ~1.9% over depths ranging from ~6 to ~16 m. An inverse correlation between the dv/v and tailings pond levels in the summer months aligned with the earlier geophone study. This study demonstrated how DAS can augment traditional geotechnical monitoring by providing higher spatial resolution of dv/v as a proxy for dam performance. While higher frequency (> 1Hz) DAS measurements can be used with CWI, lower frequency (<1 Hz) DAS measurements can provide complementary information on changes in strain near the cable. A three-day low-frequency DAS dataset encompassing a rainfall event was acquired from the Hollin Hill slow-moving landslide observatory. A conceptual framework was developed to interpret the onset of movement, retrogression, and a flow lobe surge with nanostrain-rate sensitivity, providing new insights into the kinematics of a slow-moving landslide. Collocated slope inclinometer data correlated with the inferred DAS displacements. This study demonstrates how lower frequency DAS can also provide valuable information for tailings dam monitoring applications, using slope inclinometer data (commonly installed for tailings dam monitoring) to evaluate DAS performance. Together, these findings demonstrate how DAS can be used to advance monitoring of tailings dam performance to reduce the risk of future dam failures.
Embargo
Application of machine learning, ambient noise and 2-D seismic interpretation to investigate induced seismicity in western Canada
(2024-04-30) Rojas-Parra, Jesus; Eaton, David; Dettmer, Jan; Karchewski, Brandon Anthony James
In western Canada, the development of unconventional oil and gas resources in low-permeability reservoirs has increased dramatically in the last two decades. Production of these resources typically requires hydraulicfracturing stimulation to increase permeability, as well as disposal of excess brines into permeable rock units. Both of these industrial processes can lead to induced seismicity (human-caused earthquakes). This thesis applies several approaches in two study areas in western Canada where induced earthquakes have occurred, mainly using existing open-source software. Machine learning (ML), a rapidly growing artificial intelligence approach, is combined with a probabilistic non-linear global search algorithm to construct a new seismicity catalog for the northern Montney play in eastern BC using data from both public and private seismograph stations. PhaseNet, a popular ML method that has been pre-trained using historical earthquake catalogs from around the world, is used for detection of P- and S-wave arrivals (phases). A Gaussian mixture model associator, GaMMA, is used to associate phases into events. Hypocentre locations are determined from the phase picks using the NonLinLoc algorithm. Geological context for the seismicity event distribution is supplied by existing published faults as well as new fault interpretations using 2-D seismic profiles. A growing cloud of seismicity, consistent with the Kaiser effect, is observed along a mapped fault near a disposal zone. This phenomenon is modelled using pore-pressure diffusion, numerically simulated here using code that I developed. In the Musreau Lake region ofwestern Alberta, a different type of analysis is applied, in part because a comprehensive seismicity catalog has been published using a similar ML approach to that described above. Ambient-noise interferometry, which uses cross-correlation of background noise between pairs of seismograph stations, was used to investigate temporal changes in subsurface velocity associated with pore-pressure diffusion and induced seismicity. The results using the NoisePy software show a good recovery of relative velocity changes between pair stations. Further, statistical analysis is required to construct robust observations containing all retrieved signals. Finally, areas for possible future work are proposed, including the use of satellite interferometric synthetic aperture radar (InSAR) as a complementary method for seismological studies of induced earthquakes and slow-slip events, as well as fully coupled physics-based simulation and inversion of induced seismicity processes.
Open Access
Application of multiple reflections in seismic imaging
(2023-11-17) Huang, Shang; Trad, Daniel Osvaldo; Dettmer, Jan; Innanen, Kristopher A.H.; Ferguson, Robert James; Karchewski, Brandon Anthony James; Sacchi, Mauricio Dino
Multiple reflections in seismic imaging provide additional information and insights into subsurface structures. However, their practical utilization could be improved due to their complex interaction with the subsurface and interference with primary reflections and noise. This thesis aims to use multiples in seismic imaging in two contexts: classical and machine learning imaging. For the first context, I attempt to extend aperture illumination in a phase shift plus interpolation (PSPI) migration by adding scattering terms in the phase-shift wavefield propagation operator. This method iteratively adds scattering terms for each reflector as the source and receiver wavefields propagate downward into the subsurface, which helps in efficiently extending the illumination of horizontal reflector edges. In the second context, I consider improving images from reverse time migration (RTM). RTM with multiple reflections (RTMM) can improve illumination but suffers from interferences between different orders of multiples. To overcome this limitation, I proposed a method based on a convolutional neural network (CNN) and U-Nets that approximates the inverse of the Hessian similarly to least squares migration, but with less computational cost. The U-NET is trained to learn patterns representing the relation between the reflectivity obtained through RTMM and the true reflectivity. I further developed this by adding a discrete wavelet transform (DWT) input channel which provides an additional constraint that helps to enhance image resolution. Finally, as a key application of the above techniques, I consider the problem of time-lapse seismic monitoring which attempts to detect very weak signal differences produced by changes in reservoirs. This technique is affected by changes in the near-surface noise and insufficient illumination to detect the weak changes. I proposed a novel method that leverages stacked long short-term memory (SD-LSTM) and U-Net neural networks to predict and mitigate noise in monitor data. I test the method in a field dataset, DAS VSP data from the CaMI FRS project. The output provides meaningful information and prediction for CO2 injection migration within a target area. This result aligns closely with the CaMI FRS project CO2 injection plan, providing valuable insights for monitoring the CO2 migration paths for the Basal Belly River Sandstone Formation.
Open Access
Characterization of the southern Rocky Mountain Trench near Valemount, British Columbia, using receiver functions
(2020-05-21) Coffey, Juliann Rose; Gilbert, Hersh; Eaton, David William S.; Dettmer, Jan
The Rocky Mountain Trench (RMT) within the eastern Canadian Cordillera marks a prominent topographic and physiographic boundary within the exotic terrains that accreted onto the North American continent. Previous geophysical data have shown that the RMT coincides with a ∼10 - 15 km decrease in crustal and lithospheric thickness from east to west. Receiver function (RF) analysis is a suitable method for investigating the RMT because it provides estimates of the depths of lithospheric boundaries. A deployment of ten broadband seismometers between July 2017 and July 2019 within the RMT near Valemount, B. C., recorded over 300 teleseismic earthquakes with magnitudes greater than 5.5 at epicentral distances between 30◦ and 100◦ degrees. Due to high levels of cultural noise in the region, removing low signal-to-noise ratio events left between 57 and 17 events per station to use for further analysis. Assuming average crustal velocities, stacks of low frequency RFs show a high-amplitude positive arrival at depths greater than 45 km. When viewed at higher frequencies, these RFs reveal arrivals at depths between 20-30 km. These RFs are analyzed through joint inversions with a surface wave dispersion curve, common conversion point stacking, and H-k stacking. There are three major findings of this study: (1) The Moho near the RMT around Valemount is relatively deep (> 45 km) and the crust is thicker to the south, interpreted to indicate that this location marks the northern most segment of the portion of the southwestern Canadian Cordillera where the crust underwent a greater amount of shortening and thickening compared to the rest of the Cordillera. (2) Strong midcrustal discontinuities are present in the RF stacks, and they are interpreted to mark the depth of the Rocky Mountain basal detachment, corresponding to the interface between the North American basement and accreted terranes. (3) The RF signals vary strongly with back-azimuth, likely resulting from a combination of crustal heterogeneity, anisotropy, and dipping layers.
Open Access
Distributed acoustic sensing: modelling, full waveform inversion, and its use in seismic monitoring
(2022-01) Eaid, Matthew; Innanen, Kristopher; Lawton, Donald; Trad, Daniel; Dettmer, Jan; Ajo-Franklin, Jonathan
Distributed acoustic sensing (DAS) is a rapidly evolving seismic acquisition technology. Employing rugged and small optical fibers, DAS offers an opportunity for access to acquisition geometries often not accessible to more conventional geophone sensors. This is a highly attractive property of DAS fibers, that offers access to the often unrecorded transmission wavefield modes that are crucial to support land full waveform inversion (FWI). The inherently different sampling that DAS fibers provide leads to a requirement for modeling strategies that differ from those used to simulate point sensor data. To maximize the potential of DAS which only samples tangential strain, and is therefore a single component sensor, fibers are often shaped to improve their wavefield sampling. In this thesis I propose a robust method for simulating DAS data from arbitrarily shaped fibers, that couples a geometric model of the fiber to an elastic wavefield propagator to provide the strain sensed along the tangent of a fiber. The simulation methods are used to generate a large synthetic dataset that supports a machine learning study for extracting source mechanism information from DAS data, and to support FWI. The shape of the DAS fiber, both on large-scales such as when it tracks a deviated well, or on small scales when it is wrapped in some characteristic shape affects the sensitivity of the fiber to different wavefield components, which has an important influence on the recorded data. The data recorded by a DAS fiber is a function of the fiber shape, and it is therefore expected that fiber shape will have an important influence on parameter estimates in FWI. To address this question, I first develop a method for incorporating data from an arbitrarily shaped fiber in FWI. Using a 2D isotropic-elastic FWI over synthetic toy models, I then examine the role of fiber gauge length, fiber shape, and their interplay in FWI. It is determined that fiber shape has an important influence on parameter resolution, but that the optimal fiber shape is acquisition geometry and model dependent. The work presented in this thesis lays out a sandbox for appraising fiber geometry prior to field deployment, and allows for the optimization of fiber shape to support FWI. It is also shown that short gauge length fibers (where short is described in relation to the fiber geometry) can push DAS fibers towards multi-component point sensors. FWI results obtained with short gauge length fibers and orthogonal-point-sensing geophones agree favorably. To provide confidence to the synthetically derived conclusions, these insights must also transfer to field data. The methods for the inclusion of DAS data in FWI are used to invert field data from a straight DAS fiber both in isolation and in various combinations with collocated accelerometer data from a field research station focusing on the monitoring of injected carbon dioxide. The field FWI is also 2D and assumes isotropic-elastic wavefield propagation, and incorporates a single-parameter parameterization that leverages prior information from well-log data. Models obtained from inversions of each dataset are correlated in their overall structure, but each provides differing models of the subsurface. Incorporation of both datasets into a single objective function is observed to stabilize the inversion, and leads to a more robust estimate of the elastic subsurface parameter distribution. The models obtained in this study can be used as baseline models of the research station to support further time lapse analysis.
Open Access
Double-Difference Seismic Event Relocation: A Study of the Applications and Limitations of the Relocation Problem
(2024-08-12) Biegel, Katherine M.; Dettmer, Jan; Eaton, David W. S.; Gilbert, Hersh Joseph; Innanen, Kristopher A. H.; Kao, Honn
Earthquake relocation provides refined earthquake catalogs based on additional considerations or data beyond the initial earthquake location estimates. Double-difference relocation, in particular, utilizes the assumption that ray paths in close proximity to one another will pass through a similar earth velocity structure. The method minimizes the residual difference for pairs of these travel time observations to refine event locations. This thesis focuses on the application of three double-difference methods based on differing geometries of data pairing: event-pair relocation, station-pair relocation, and double-pair relocation. In this thesis, I introduce a previously unavailable software, relocDD-py, that implements all three of these relocation methods along with a complete workflow, including data preparation, automated variable selection, and post-relocation uncertainty analysis. This software is developed in Python using many widely implemented Python packages to allow for integration into existing seismic workflows. In addition, I present studies that apply this software and methodology at various scales, including (1) a large-scale tectonic study of complex plate subduction and the resulting seismicity in Alaska and Yukon where we find direct evidence of the Totschunda-Fairweather Connector fault; (2) a regionally observed induced seismicity event with analysis of the entire earthquake sequence from Peace River, Alberta where we find activation of multiple sub-parallel faults; and finally (3) a densely monitored induced seismicity experiment which ruptured a complex network of pre-existing faults near Fox Creek, Alberta where we find limits on the depth of seismicity to the hydraulic fracturing depth. Double-difference relocation can improve existing catalogs even in sparsely monitored areas and can reduce relative location uncertainty to such a degree that detailed three-dimensional interpretation of fault structures is possible. In many cases, event-pair relocation, the most widely applied double-difference method, is sufficient to improve relative uncertainties and maintain absolute location uncertainties. However, in cases with complex velocity models and dense monitoring, the double-pair method can improve relative uncertainties beyond event-pair relocation, which may identify additional seismicity features. The double-pair method is more computationally expensive and, therefore, is not necessary in all cases. All three double-difference relocation methods have an important role to play in earthquake and other seismic event relocation and catalog refinement. My work provides a software tool and a standardized workflow for the implementation of double-difference relocation in seismic studies.
Open Access
Estimating Earth structure from seismograms with unknown earthquakes source characteristics
(University of Calgary, 2018-09-24) Zhang, Haoze; Dettmer, Jan
Open Access
Examination of Shallow Structure in Geothermal Sites of Western Canada Using Microtremor Measurements: Mount Meager and Burwash Landing Case Studies
(2024-02-15) Berumen Borrego, Fernando; Gilbert, Hersh Joseph; Dettmer, Jan; Eaton, David W. S.
Canada has enormous geothermal potential, a sustainable alternative energy in the form of heat. Despite its significant role in moving Canada closer to net-zero CO2 goals, geothermal resource production faces numerous exploration challenges, notably the risks associated with drilling. However, with the emergence of new technologies, e.g., closed-loop systems and declining costs, geothermal energy is increasingly competing with conventional resources like coal. This thesis investigates the use of the Horizontal-to-Vertical Spectral Ratio (HVSR) method, also known as Nakamura’s method or microtremor HVSR (mHVSR), in two geothermal exploration projects. Commonly referred to as HVSR in the geophysical community, this method analyzes microtremors and is effective in determining seismic structures up to 200 meters deep at Mount Meager and 500 meters at Burwash Landing. HVSR complements other geophysical methods that focus on deeper structures and overcomes the limitations of surface geology. At Mount Meager, British Columbia, HVSR has been instrumental in identifying resonance frequencies and understanding the subsurface structure. It has revealed the variability in the shallow subsurface, aiding in the estimation of volcanic unit thicknesses and expanding upon existing geological and seismic data. This has enhanced the understanding of subsurface geology and associated risks. In Burwash Landing, Yukon Territory, where geothermal energy is crucial for local energy needs, the application of a trans-dimensional Bayesian inversion algorithm on HVSR curves has provided detailed insights into seismic structures. This method has offered accurate estimations of bedrock depth and sediment compaction, refining previous interpretations. Notably, it helped correct overestimated sediment layer thicknesses encountered during a drilling operation in November 2022. The trans-dimensional approach avoids the use of predefined models, quantifies uncertainty, and adds objectivity to interpretations. Overall, these studies highlight the importance of HVSR in geothermal exploration, especially in areas with complex topography and geology. HVSR is crucial for understanding shallow structures, assessing associated hazards, and informing studies on deeper geothermal resources. The findings significantly contribute to the understanding of seismic structures in these regions, with broader implications for global sustainability, energy self-sufficiency, and environmental objectives.
Open Access
Inversion Modelling of Copper Transport in Saccharomyces Cerevisiae
(2020-05-12) Wilkins, Aaron Francis; Wieser, Michael E.; Karchewski, Brandon; Dettmer, Jan; Strous, Marc
Copper is an essential nutrient but the uptake into cells is poorly understood. This dissertation summarizes the development of a mathematical system of equations to model the transport of copper in S. cerevisiae. Yeast is a model organism for studying the copper transport in human hepatic cells because the chaperone proteins and structures are well conserved between the species. An experiment is performed to investigate the transport between the growth media and the cells to model the process behind this important pathway. Transport mechanisms for this process are presented, mathematically modelled, and evaluated. Rate limited diffusion did not appear to be adequate in modelling the transport, but a term including a target copper concentration which cells actively maintain was introduced, and with a delayed activation, fit the data much more effectively. With this model, a framework is established for incorporating organelles to eventually model the intracellular copper transport and analyze the copper isotope distributions in the future. This work contributes to a larger initiative to incorporate copper isotope analysis as an innovative medical diagnostic tool in assessing human cellular pathology.
Open Access
Machine learning methods modeling waveform, multi-parameter full waveform inversion, and uncertainty quantification
(2024-01-02) Zhang, Tianze; Innanen, Kristopher; Trad, Daniel; Dettmer, Jan; Karchewski, Brandon; Zhu, Tieyuan
Full waveform inversion (FWI) is a potent technology capable of estimating subsurface parameters using seismic records. However, several challenges hinder its widespread application in large-scale field operations. In this thesis, I introduce strategies that leverage machine learning techniques to address the challenges faced with FWI, making FWI more feasible to implement in complex media and providing a confidence analysis for the inversion results. A primary concern is the necessity for accurate initial models in FWI; without these, FWI risks becoming ensnared in local minima. In this thesis, I propose recurrent neural network (RNN) isotropic elastic FWI. Then, I proposed the elastic implicit full waveform inversion. Instead of directly updating the elastic parameters like in the conventional FWI, I use neural networks to generate elastic models and update the weights in the neural network to decrease data misfits. Furthermore, how to develop a feasible uncertainty quantification method for FWI out- comes remains an open question. The estimation of the prior uncertainty and a method that effectively evaluates the inverse Hessian matrix are critical steps for the uncertainty quantification under the Bayesian inference. I introduce a method that uses the Bayesian neural network (BNN) to provide prior uncertainty for the elastic models and an algorithm that efficiently approximates inverse Hessian. During the inversion process, simplifications in wave propagation physics are often made to enhance computational efficiency. Yet, the extent to which such simplifications (mod- elling errors) impact the inversion results is seldom addressed. I proposed a method for the viscoelastic FWI that can quantify a specific type of modelling error, which can quantify the modelling error caused by the insufficient ability of the relaxation variables to model the constant Q model or a Q model that we desire, i.e., obtained from the field records. I also analyze the effect on the inversion results of such modelling errors. Our community is strongly motivated to find means of accurately computing wavefields with minimal computational expense. I propose the one-connection Fourier neural operator (OCFNO). I test the ability of such a network to “learn” to solve elastic wave equations. Computational speed-ups are significant.
Open Access
Magnitude correlations and criticality in a self-similar model of seismicity
(2019-03-25) Zambrano Moreno, Andres Felipe; Davidsen, Jörn; Hobill, David W.; Jackel, Brian J.; Dettmer, Jan; Simon, Ch M.
We present an analysis of the statistical relation between subsequent magnitudes for a previously proposed self-similar aftershock rates model of seismicity whose main distinguishing feature arises as a consequence of the dependency of the rates on the magnitude difference between trigger-triggered events. By means of a particular statistical measure we studied the level of magnitude correlations among time-ordered subsequent events for various magnitude thresholds under specific types of time conditioning, explained their provenance through the model and found that the type of null model chosen in the analysis plays a pivotal role in the type of observed correlations. With particular time conditioning between subsequent events, we also analyzed and compared a model catalog to data from Southern California (SC), and found that model and real-world catalogues are consistent with each other within our statistical measure. When comparing a culled SC and the corresponding culled model catalogue under a specific type of time conditioning and magnitude thresholds, correlations do not significantly deviate from zero for the model but are present in the subset SC catalogue; these observed correlations in the SC catalogue we attributed to a process where a fraction of events are missed in the seismic recordings. The former was substantiated when we looked at the culled SC catalog for higher magnitude thresholds and saw no significant correlations, although the amount of data for this already shortened catalog is relatively low. By creating synthetic catalogues between 2 − 26 times the culled model catalogue, we estimated the length of a catalogue needed to begin to observe magnitude correlations at 3σ to be ∼ 15 times the current culled model catalogue ( ∼ 150 years worth of data) while the existence of magnitude correlations can be used in earthquake forecasting, only the time variations in the frequency-magnitude distribution might lead to a significant improvement in forecasting. We also studied the criticality of the self-similar model from the perspective of finite size effects through the introduction of an upper magnitude cut-off in the analysis and established that the model lies in the sub-critical regime for the SC parameters.
Open Access
Near surface investigation with DAS for CO2 sequestration and monitoring
(2024-01-11) Qu, Luping; Innanen, Kris; Dettmer, Jan; Martin, Eileen; Liao, Wenyuan; Trad, Daniel
In this thesis, I investigate near-surface seismic properties and several geophysical methods mainly including surface wave dispersion inversion (SWDI) and full waveform inversion (FWI) for CO2 monitoring, focusing on the capabilities of Distributed Acoustic Sensing (DAS) technology. The study is underpinned by data collected from Newell County Facility, Alberta, Canada, employing seismic data acquired from both surface-deployed and vertical seismic profi le (VSP) DAS fiber. DAS data, characterized by broadband and dense spatial sampling, facilitate the extraction of high-resolution near-surface velocity pro les due to their enhanced signal-to-noise ratio and resolution in low-frequency components and multimode dispersion curves. The first segment of the study introduced several types of dispersion curves, explores trans-dimensional (TD) inversion, employing multimode dispersion curves and reversible-jump Markov Chain Monte Carlo (MCMC) sampling to generate probabilistic posterior density (PPD) estimates of model parameters. This approach yields inversion results that align well with known lithological data. This research showcases the potential of horizontal DAS data in high-resolution, near-surface investigations. Additionally, I developed a multi-step multiscale surface wave FWI. Utilizing DAS recorded surface waves, high-resolution S-wave velocity (Vs) and attenuation (quality factor Qs) models of the near-surface are obtained through FWI, offering improved lateral resolution and depth penetration compared to conventional surface-wave analysis. The inclusion of low-frequency components in DAS data effectively mitigates the cycle skipping challenge commonly associated with FWI, leading to high-resolution VS models that capture lateral variations effectively. I also addressed the challenge of noise in seismic data, particularly its impact on acoustic and elastic FWI models. By incorporating the data covariance matrix into the mis t function, this approach mitigates the effects of noise, improving the accuracy of the FWI models. Building on these methods, I applied anisotropic FWI with variable density to DAS recorded walk-away VSP data for characterizing subsurface velocity, anisotropy, and density structures. This technique, essential for time-lapse studies of CO2 injection and storage, proved effective in providing more accurate P-wave velocity, density models, and anisotropy parameters compared to isotropic FWI. In conclusion, this thesis demonstrates the potential of using DAS technology and advanced geophysical methods for near-surface investigation and CO2 monitoring. The integration of DAS data with trans-dimensional and varied FWI approaches, alongside noise mitigation strategies, offers a signi cant step forward in accurate and e cient subsurface characterization, crucial for environmental monitoring and carbon capture and storage initiatives.
Open Access
Near-surface S-wave traveltime corrections and inversion: a raypath-consistent and interferometric approach
(2017) Cova Gamero, Raul Jose; Innanen, Kristopher A. H.; Frederiksen, Andrew; Lawton, Donald Caleb; Krebes, Edward Stephen; Dettmer, Jan
Removing near-surface effects in the processing of 3C data is key to exploiting the information provided by converted waves, particularly for the case of the PS mode where converted energy travels back to the surface as S-waves. The very low velocity of S-waves amplifies the distortions introduced by the near-surface in the PS-traveltimes. This is usually solved by computing the vertical traveltimes in the near-surface layer and removing them from the data in a surface-consistent framework. However, if the velocity change between the near-surface layer and the medium underneath is small, the vertical raypath assumption that supports the surface-consistent approach is no longer valid. This property results in a non-stationary change of the near-surface traveltimes that needs to be addressed to properly remove its effect. I show how the delays introduced by the presence of very low S-wave velocities in the near-surface can introduce raypath-dependent effects which can be larger than what can be considered a residual static. In this study, a raypath-consistent solution for removing near-surface traveltime effects is proposed. This is achieved by transforming the data, organized into receiver gathers, to the tau-p domain and performing crosscorrelation and convolution operations to capture and remove the near-surface delays from the data. The tau-differences captured during the interferometric processing of the near-surface effects are then used in an inversion algorithm to estimate the S-wave velocities in the near-surface. This processing work-flow provides not only a set of corrections but also a velocity model that is based on them. I tested this method on synthetic and field data. In both cases, removing near-surface time delays in a raypath-consistent framework improved coherency and stacking power of shallow and deep events simultaneously. Shallow events benefited most from this processing due to their wider range of reflection angles. This approach can be useful in the processing of wide-angle broadband data, where the kinematics of wave propagation are not consistent with vertical raypath approximations in the near-surface. Additionally, this method provides a near-surface S-wave velocity model that can be used for building migration velocity models or to initialize elastic full waveform inversions.
Open Access
Nonlinear Bayesian estimation of centroid moment tensors using multiple seismic data sets in the Kiskatinaw seismic monitoring and mitigation area
(2022-12-22) Hamidbeygi, Mahdi; Dettmer, Jan; Eaton, David; Gilbert, Hersh
Centroid moment tensor (CMT) parameters of earthquakes are routinely estimated to gain information on structures and regional tectonics. However, for small earthquakes (M<4) it is still challenging to determine CMTs due to the lack of high-quality waveform data. In this study, we propose to improve solutions for small earthquakes by incorporating multiple seismic data types in Bayesian joint inversion: polarities picked on broadband signals, amplitude spectra for intermediate frequency bands (0.2--2.0 Hz), and waveforms at low frequencies (0.05--0.2 Hz). Both measurement and theory errors are accounted for by iterative estimation of non-Toeplitz covariance matrices, allowing to objectively determine weights for the different data types in the joint parameter estimation. Validity and applicability of the method are demonstrated on simulation and field data. Results demonstrate that the combination of data, such as a single high quality waveform, a few amplitude spectra and many waveform polarities are able to resolve CMT parameters to comparable quality as if many high quality waveforms were available. Results of 10 induced events that occurred in northeastern British Columbia between January 2020 and February 2022 indicate predominant strike slip focal mechanisms with low non double-couple components. These events appear to be located at shallow depth with a short duration as expected for induced seismicity. These results are consistent with previous studies. Therefore, we learn that this method reduces the dependency of source inversion on high-quality waveforms and permits to resolve CMTs for earthquakes as small as ML 1.6.
Open Access
Probabilistic Joint Inversion of Gravity and Magnetic Data with 3D Trans-Dimensional Earth and Noise Models
(2022-01) Ghaleh Noei, Emad; Kim, ‪Jeong Woo; Dettmer, Jan; Sideris, Michael George; Rangelova, Elena
Gravity and magnetic data resolve Earth models with variable spatial resolution, and Earth structure exhibits both discontinuous and gradual changes. Therefore, model parametrization complexity should address such variability, for example, by locally adapting to the spatial resolution of the data. The reversible-jump Markov chain Monte Carlo (rjMcMC) algorithm can be employed to explore Bayesian models with variable spatial resolution that is consistent with data information. To address non-uniqueness in joint inversion of potential field data, I employ novel hierarchical models that are based on irregular spatial partitioning and incorporate geological constraints about the subsurface as prior information. Spatial partitioning choices include nested Voronoi cells, linear interpolation and alpha shapes, and Voronoi cells and planes. These parametrizations partition the subsurface in terms of rock types, such as sedimentary rocks, rock salt and basement rocks. Therefore, meaningful prior information can be included in the inversion which reduces non-uniqueness. In addition, nonoverlapping prior distributions are used for density contrast and susceptibility between rock types. Another significant challenge for potential field data is unknown noise characteristics. In particular, poorly estimated noise characteristics can significantly change model spatial resolution and complexity. I consider empirical and hierarchical approaches to noise estimation that include theory and measurement errors. The empirical estimation of full data covariance matrices is based on residuals and an iterative scheme. The hierarchical approach employs a trans-D autoregressive noise model that quantifies the impact of spatial noise correlations on geophysical parameters. Both 1-D and 2-D spatial correlations are considered for 2-D and 3-D inversions, respectively. The method is applied to gravity and magnetic data to study salt and basement structures. This thesis demonstrates that meaningful partitioning of the subsurface into sediment, salt, and basement structures is achieved by these methods without requiring regularization. Multiple simulated- and field-data examples are presented. Simulation results show a clear delineation of salt and basement structures while resolving variable length scales. The field data results are consistent with observations made in the simulations. This work resolves geologically plausible structures with varying length scales and clearly differentiates salt structure and basement topography.
Open Access
Seismicity and tectonic interpretation of the Southern Rocky Mountain Trench near Valemount, British Columbia, Canada
(2020-05-15) Purba, Joshua Chris Shadday; Dettmer, Jan; Gilbert, Hersh; Eaton, David William S.; Trad, Daniel O.
The Rocky Mountain Trench is a large geological feature in the Canadian Cordillera with complex structures. Although efforts to understand the structure of the trench have been conducted through refraction-seismic, reflection-seismic, and geological studies, detailed knowledge of the trench is still sparse but crucial to understanding its evolution. Here, I conduct a local seismic study that involves earthquake detection, arrival-time picking, earthquake location, and a tectonic interpretation of the Rocky Mountain Trench in the area of Valemount, British Columbia. I developed and employed a nonlinear, probabilistic multiple-earthquake location for earthquakes detected here. The location provides both earthquake locations and rigorous estimates of their uncertainties. Based on analysing one year of data, my catalogue includes 47 local earthquakes that I identified and located. The results of the multiple-earthquake location presented here illustrates uncertainty reduction in depth from 18 to 5 km compared to the depths of earthquakes calculated based on single-earthquake locations. This lower depth uncertainty permits better inferences of the tectonic development of the Rocky Mountain Trench. The earthquake locations determined here displays a change in the distribution of seismicity around Valemount. Seismicity extends to the west of the RMT and the south of Valemount. While to the north, the seismicity is primarily confined to the trench and areas to the east. The distribution of seismicity also supports the dome-shaped Malton Gneiss in the subsurface. Seismic velocities are consistent with metamorphic rocks and the presence of significant amount of quartz in crustal rocks.
Open Access
Source Parameters and Tectonic Setting of the 2017 St. Elias Earthquake Sequence near the Southern Terminus of the Eastern Denali Fault, Northwestern Canada
(2019-09-18) Choi, Minhee; Eaton, David W. S.; Enkelmann, Eva; Dettmer, Jan; Gilbert, Hersh J.
On May 1, 2017 two earthquakes of magnitude 6.2 and 6.3, respectively, occurred in proximity to the eastern Denali fault (EDF). These double mainshock events were followed by more than 2,700 aftershocks. Moment-tensor inversion of the mainshock signals shows that the initial event produced reverse slip on a steeply dipping fault with a NW-SE strike direction, while the second produced left-lateral strike-slip on a near-vertical fault with an E-W strike direction. A double-difference relocation method, coupled with clustering analysis, was applied to the aftershock distribution, confirming that seismicity was localized along two previously unmapped fault structures. Stress inversion indicates that the maximum principal stress axis is oriented almost perpendicular to the EDF, suggesting that the fault system is not well oriented for strike-slip in the contemporary stress field. Coulomb stress analysis indicates that the second event was likely triggered by the first one (static stress triggering), with a delay of about two hours. A generalized model is developed to explain the observations, wherein gravitational potential from the >4,000 m high Mount Fairweather, as well as strain partitioning along the plate boundary, produce a stress regime that extends inboard towards the EDF.
Open Access
The Stochastic Characterization of Natural and Hydraulic Fractures in Unconventional Reservoirs
(2023-01-13) McKean, Scott Harold; Dettmer, Jan; Priest, Jeffrey Alan; Eaton, David WS; Wan, Richard G; Davidsen, Joern; Dusseault, Maurice Bernard
An informed understanding of the subsurface is critical for mining, tunnelling, wastewater injection, carbon sequestration, and hydraulic fracturing (HF). Unfortunately, subsurface characterization is full of uncertainty. This is especially true when trying to understand or mitigate induced seismicity (IS), or the triggering of earthquakes by anthropogenic processes. This research focuses on hydraulic fracturing caused IS in unconventional reservoirs. The interaction between HF and IS is complicated by geomechanical variability and the presence of natural fractures. Our research accomplishes three objectives. We study natural fractures through outcrop analogues, discrete fracture network modelling, and induced seismicity. We characterise geomechanical rock properties along with their uncertainty. Finally, we develop a repeatable and scalable workflow to separate HF microseismicity from IS in order to characterise hydraulic and natural fractures. The research focuses on the Duvernay Formation in the Western Canadian Sedimentary Basin. An alpine outcrop equivalent of the Duvernay is characterized to quantify small- and large-scale fractures. This study reveals irreducible small-scale heterogeneity, as well as discernable patterns in large-scale fractures. Statistics and geostatistics are used to investigate elastic moduli and brittleness. The work shows how measurement and modelling uncertainity can propogate from laboratory to basin-scale. It reveals fundamental differences between elastic moduli and brittleness and shows why holistic modelling and uncertainty quantification approaches are essential to understanding and modelling the subsurface. We then introduce methods for the separation of HF microseismicity from IS. Physics-based clustering and Bayesian inference of diffusivity are used for the separation. This permits HF characterization which highlights the large variability of diffusivity and HF dimensions. We show why physics-based constraints are essential for microseismic analysis. The separated IS allows us to infer information about the natural fractures linked with induced seismicity. Application of the methods to the Duvernay shows HFs propogating directly into natural fractures and rotating away from the maximum principal stress direction towards natural fractures. Discrete fracture network modelling and parameter estimation is able to constrain the architecture of multiple fracture sets. We demonstrate that aseismic fracture sets are essential for establishing pressure connectivity and displaying IS.
Open Access
Time-Lapse Seismic Imaging, Full-waveform Inversion, and Uncertainty Quantification
(2023-09-15) Fu, Xin; Innanen, Kristopher A.H.; Innanen, Kristopher A.H.; Dettmer, Jan; Trad, Daniel Osvaldo; James, Brandon Anthony; Liao, Wenyuan; Malcolm, Alison E.
Time-lapse seismic, also known as 4D seismic, is a powerful tool for monitoring subsurface changes over time. By comparing seismic data acquired at different intervals, it enables the detection and characterization of dynamic reservoir processes, aiding in reservoir management, production optimization, and enhanced oil recovery. It has applications in geothermal energy, CO2 storage monitoring, and environmental impact assessment. However, accurate analysis of time-lapse seismic data remains a challenging task. It requires well-repeated time-lapse seismic surveys, including well-repeated acquisition geometry and equipment as well as well-repeated ambient noise. This thesis is to alleviate the non-repeatability issues in time-lapse seismic imaging and full-waveform inversion (FWI), and to realize the uncertain quantification for time-lapse seismic waveform inversion. A time-lapse imaging approach that involves two new frequency-domain matching filters is developed. The first filter requires source wavelet estimates from both baseline and monitoring data, while the second filter is source-independent but more sensitive to data noise. By applying these filters, we successfully reduce source wavelet non-repeatability, and the new approach improves the accuracy of time-lapse imaging. Furthermore, a stepsize-sharing time-lapse FWI strategy is designed to reduce artifacts caused by the variability of convergence in conventional strategies. The strategy demonstrates good adaptivity in different tested realistic scenarios using synthetic data. It is stable for scenarios using biased starting models, while the conventional strategies fail in this regard. Moreover, to realize the uncertain quantification, a Bayesian time-lapse FWI procedure, based on a Markov chain Monte Carlo (MCMC) algorithm, is formulated. The formulation employs several existing strategies, including the use of a double-difference time-lapse FWI, incorporation of time-domain multi-source data, and application of a local-updating target-oriented inversion. It incorporates these within a stochastic framework, involving the computation of model covariance with an adaptive Metropolis algorithm, and a method to estimate data error statistics based on the features of time-lapse difference data is incorporated. A random walk Metropolis-Hastings MCMC is adopted for optimization.
Open Access
Time-lapse VSP monitoring of CO2 sequestration at the CaMI Field Research Station
(2022-04-07) Kolkman-Quinn, Brendan James; Lawton, Donald Caleb; Innanen, Kris; Dettmer, Jan
The Containment and Monitoring Institute Field Research Station (CaMI.FRS) is a carbon sequestration field experiment near Brooks, Alberta. CO2 is injected into a saline aquifer in the 6 m thick, 10% porosity Basal Belly River Sandstone at 300 m depth. This CO2 plume simulates a shallow leak scenario from an industrial scale CO2 storage reservoir. Vertical Seismic Profiles (VSP) were collected between 2017 and 2021 to determine a detection threshold and delineate the plume. These data had high repeatability, with permanent borehole sensors and identical source coordinates. Seasonal variations in surface and near-surface conditions were the main source of dissimilarity between baseline and monitor data. A time-lapse compliant processing workflow was developed for the 10 Hz – 150 Hz shallow data. This workflow produced directly comparable amplitudes between baseline and monitor data, except at higher frequency bands affected by variable near-surface filtering. Spectral differences were eliminated with high-cut filters designed for each source location. Avoiding the use of shaping filters preserved the subtle amplitude response of the CO2 plume in the remaining bandwidth. After 33 t of injection, the CO2 plume produced an observable time-lapse amplitude anomaly on multiple geophone datasets on two monitoring lines. The interpreted CO2 plume matched forward modeling expectations of approximately 50 m lateral extent, but with an asymmetric distribution around the injection well. Equivalent DAS datasets possessed lower signal strength than geophone data, with instrument noise interfering with the CO2 amplitude response. A weak seismic anomaly was observed on only one DAS monitoring line. The high-confidence geophone results establish a leak detection threshold at 33 t of CO2 for this geological setting. The ability to detect a 33 t CO2 plume with VSP field data lends confidence to Measurement, Monitoring, and Verification capabilities at industrial operations with injection rates of kilotonnes or megatonnes per year. The time-lapse compliant VSP workflow will help inform shallow monitoring procedures, while the CO2 delineation will contribute to future multi-disciplinary studies at CaMI.FRS.