Seismic Applications Research Articles

Temperature-dependent elastic properties of bitumen-bearing sands and its seismic application in Mcmurray Formation, Alberta

Seismic signatures of bitumen-bearing sand reservoir would be critically affected by the change of petrophysical parameters during the steam-assisted gravity drainage (SAGD) production. We investigated the temperature evolution of the steam chamber in a bitumen-bearing sands reservoir from the 4D time-lapse seismic by merging geological characterizations, laboratory measurements, and temperature observation wells focusing on a bitumen sands reservoir in Alberta, Canada. We have characterized both ultrasonic P- and S-wave velocities of heavy oil rocks and bitumen itself of the McMurray Formation. The Vp and Vs of oil sands vary with temperatures ranging from 10 to 175 ?C. Overall, Vp gradually decreases as temperature increases for different decrease slopes. Meanwhile, bitumen saturation, bury depth, and clay content mutually control elastic characteristics of oil sands. In addition, the viscosity and moduli of heavy oil behave differently at varying temperatures, leading to ultrasonic velocities varying temperature-dependent. A frequency-dependent impedance-temperature relation bridges the ultrasonic P-impedance and petrophysical temperature parameter through 4D time-lapse seismic, which allows us to estimate the temperature distribution of the steam chamber in sands reservoir, thereby optimizing the steam injection strategy ultimately during production process. Field test results demonstrate that the temperature is the highest in the vicinity of the steam injection well within the steam injection zone. Then, the hot steam (e.g., gt;200 ?C) gradually diffuses toward the surrounding cool (e.g., 8 ?C) bitumen-cemented sands, which, consequently, are warmed up and bitumen becomes movable. In addition, it is because of the heterogeneity of oil sand reservoirs, that the temperature map distributes inhomogeneously. More importantly, by comparison of temperature mapping in first and second monitoring lines, it implies that hot steam tends to spread in the reservoir with a similar path at different steam injection stages, reminding us that the optimized steam injection strategy plays a substantial role.

Velocity model building via combining seismic slope tomography and supervised deep learning

AbstractSeismic slope tomography is an effective method to build macro velocity model. In order to improve the accuracy and resolution of the slope tomography, we proposed a novel approach that combines slope tomography with supervised deep learning. First, the slope tomography is used to obtain the macro velocity model and the positions of reflection points. Subsequently, the slope tomographic model, positions of reflection points and the corresponding observed traveltimes are used as inputs simultaneously for a neural network, whereas the actual velocity models are used as the labels. Through training the neural network with sufficient samples, the mapping from the inputs to the real velocity model is established. The neural network learns the background velocity of the real model from the smooth tomographic model, the velocity details from the traveltimes and the formation interface information from the positions of reflection points. Consequently, a high‐accuracy and high‐resolution velocity model is obtained on the basis of the slope tomographic model. Both tests on synthetic seismic data and applications to field seismic data demonstrate the effectiveness of the proposed method.

Superelastic behavior of novel Ni–Ti–Co shape memory alloys for seismic applications

Abstract Ni–Ti–Co is a new shape memory alloy (SMA) composition that has a higher strength and a lower superelastic temperature range than the traditional binary Ni–Ti SMA composition. In this paper, the superelastic properties of Ni–Ti–Co bars that are important for seismic applications were studied and compared with those of Ni–Ti bars. The superelastic behavior, strength, strain recovery, low-cycle fatigue characteristics, and fracture strains of Ni–Ti–Co and Ni–Ti SMA with different heat treatment strategies and testing temperatures from −40 °C to 50 °C were investigated with seismic applications in mind. The effect of low-cycle fatigue loading and temperature variations on the superelasticity degradation of Ni–Ti–Co SMA were evaluated. The results showed that the yield stress of Ni–Ti–Co SMA at room temperature was 1.41–1.74 times that of the Ni–Ti SMA. Ni–Ti–Co SMA exhibited 100% strain recovery when unloaded from 5% strain in a temperature range from −40 °C to room temperature; while at 50 °C, a residual strain of 0.6% was observed when unloaded from 5% strain. For comparison, the Ni–Ti SMA lost superelasticity when the temperature was reduced to 0 °C. When subjected to low-cycle fatigue loading, the stability of the superelastic behavior (maintaining yield stress, energy dissipation and strain recovery) of Ni–Ti–Co SMA at −40 °C was better than that at room temperature and 50 °C. In particular, the superelasticity of Ni–Ti–Co SMA at −40 °C was superior to that of the Ni–Ti SMA at 0 °C. At −40 °C, the Ni–Ti–Co SMA showed almost no loss in yield stress, damping ratio and recovery strain during the first 100 cycles of fatigue loading. Moreover, from 100th to 471st cycle, at which the fracture occurred, the strain recovery of Ni–Ti–Co SMA showed almost no degradation.

A rupture directivity adjustment model and its application in seismic hazard

This article contains a rupture directivity adjustment model for strike-slip earthquakes that can be applied to a traditional ground-motion model (GMM; one without explicit treatment of rupture directivity) to incorporate rupture directivity effects in either deterministic or probabilistic seismic hazard analyses. Application of the directivity model requires adjustments to both the GMM median and aleatory variability. The model described herein supersedes the previous models for strike-slip earthquakes developed by the authors of this article (Abrahamson, 2000; Bayless et al., 2020; Somerville, 2003; and Somerville et al., 1997; chapter 2 of Spudich et al., 2013). A key feature of the directivity model is that it is centered because it does not alter the magnitude and distance scaling of the GMM when averaged over uniformly distributed sites at a given rupture distance. Additionally, we address a longstanding issue regarding the directivity condition contained in the NGA-W2 dataset (Ancheta et al., 2014) and conclude that the biases in the mean and standard deviation of directivity effects are small enough to ignore for the purposes of modeling directivity. We provide guidance on the directivity model implementation, including a review of deterministic and probabilistic seismic hazard applications, and recommend methods for modeling hypocenter locations and multisegment ruptures. The model addresses the RotD50 (Boore et al., 2010) horizontal component of 5% damped spectral acceleration. A future update will address directionality and directivity effects for other styles of faulting. Implementations of the model are provided in the electronic supplement.

Real-time hybrid testing validation of a stiffness-controllable mass damper for reducing vibrations in nonlinear structures

The design of passive tuned mass dampers (TMDs) and the evaluation of their vibration control performance typically rest on the assumption that the primary structure is a linear system. The optimized TMD is then tuned to the natural frequency of the primary structure, enabling the transfer of vibrational energy from the structure to the TMD and its absorption through the damping mechanism of the TMD, thereby reducing vibrations in the primary structure. However, if the structure exhibits nonlinear behavior during an earthquake, causing the natural frequency of the primary structure to vary, the passive TMD may experience a detuning effect, leading to significant degradation in its control performance. To address this challenge, this paper introduces a novel mass damper with controllable stiffness (MDCS) that dynamically adjusts its stiffness to adapt to changes in structural frequency due to nonlinear responses. This feature enables the MDCS to maintain optimal performance under nonlinear conditions. First, a mathematical model of a nonlinear structure with an MDCS was established, and a least input energy method (LIEM) control law suitable for nonlinear structures was developed to minimize the total energy of the nonlinear structure and the MDCS. This study verified the theoretical model and the LIEM control algorithm by conducting real-time hybrid testing (RTHT) using a physical lever-type MDCS mechanism in combination with a nonlinear numerical (substructure) model. The RTHT results demonstrated that the MDCS provides superior vibration reduction and reduced stroke in nonlinear structures compared to passive TMDs, marking a significant advancement in vibration control for seismic applications.

Experimental investigation of damping ratio of sand-laponite and sand-bentonite mixtures using one-dimensional bender elements

Damping ratio and dynamic shear modulus are fundamental parameters for assessing the seismic response of geotechnical structures such as retaining walls, dams, tunnels, foundations, landfill covers, and embankments. Numerous seismic wave sources can substantially impact the stability and integrity of geotechnical structures, influencing design choices and the implementation of alternative solutions. In this study, the damping ratio of sand mixed with small quantities of laponite was determined by an experimental set-up using bender elements meticulously constructed for this study. Comparative tests were conducted with pure sand (i.e., control test) and sand mixed with bentonite to evaluate the effects of two types of nanoparticles. The results revealed that the damping ratio (ξ) of pure sand was approximately 7.48 %, which is generally compatible with values reported in the literature, taking into account the variations of sand and errors associated with laboratory equipment and electronic devices. Over time, the damping ratio of pure sand gradually decreased, reaching equilibrium at 0.99 % after 3–4 days of continuous shaking. The highest observed damping ratios for sand-laponite mixtures were 59 %, 69.7 %, and 98.6 % for sand+1 % laponite, sand+2 % laponite, and sand+3 % laponite, respectively. After reaching the peak, the damping ratio gradually decreased to equilibrium at 11 %, 16 %, and 19.4 %, respectively. The higher damping values for sand-laponite specimens reflect a viscous damping contribution from the presence of laponite at sand grain contacts, with higher laponite content resulting in increased damping. In comparison, the peak damping ratios for sand-bentonite mixtures were 21.9 %, 42.7 %, and 67.9 % for sand+1 %, +2 %, and +3 % bentonite, respectively. The findings highlight the potential of laponite to enhance the damping capacity of sands, which could be valuable for seismic design applications requiring improved energy dissipation.

Experimental, analytical, and numerical study of a wall-type self-centering slip friction damper

This paper proposes a novel wall-type self-centering slip friction damper (WSCSFD), which mainly comprises of the grooved T-shape and L-shape plates that are combined by stacked disc springs. The damper resists shear forces arising from the inter-story drifts under earthquakes. By leveraging the variable friction mechanism and the pre-compressed disc springs, the WSCSFD can provide self-centering capability and energy dissipation capacity. The configuration and working mechanism of WSCSFD were first discussed. And then, the theoretical equations governing the force–displacement relationship were derived. One reduced-scale damper specimen was fabricated to conduct the proof-of-concept tests. The test results validated the deformation mode and cyclic behavior of the WSCSFD. The hysteretic parameters related to seismic applications were quantified and discussed. To complement the experimental observations and gain a further understanding of the local behavior, three-dimensional finite element (FE) models were established in ABAQUS. Finally, to address the deficiencies of the damper specimen, some potential improvement approaches were suggested and evaluated through the validated FE models.

A semantic augmented approach to FEMA P-58 based dynamic regional seismic loss estimation application

Regional seismic loss estimation (RSLE) is a crucial process in both immediate post-earthquake emergency response and long-term reconstruction endeavors. Over the years, significant progress has been made in RSLE: sensing approaches such as field investigation and remote sensing offers a comprehensive overview necessary for real-time disaster response, while simulation techniques such as the methodology proposed in Federal Emergency Management Agency (FEMA) P-58 series provide insights into the mechanism of disaster development and its potential long-term impacts on urban assets. Nonetheless, challenges persist in the realm of practical RSLE applications. Firstly, a dynamic understanding of a disaster event and its influence is deemed important for effective emergency responses while challenging to achieve in current approaches. Secondly, stakeholders with varying roles, including administrators, rescue teams and ordinary citizens have distinct information requirements for RSLE. Thirdly, the complexity of seismic loss estimation, involving diverse data sources such as building information models, performance models, fragility data and sensor observations, poses interoperability issue. To tackle these issues, this article introduces a dynamic, multi-granularity, ontological representation scheme tailored for RSLE decision-supporting systems. This scheme operates across various scales, from individual building components to broader regional scales by synergistically employing FEMA P-58 guidelines and Semantic Web technologies. Upon the corresponding semantics, a question-and-answer agent powered by large language model is further developed to facilitate interaction requirements within the RSLE process via FEMA P-58 pipeline. The practical efficacy of this approach is validated through a prototype deployed under a real earthquake event, demonstrating its value in real-world scenarios.

Seismic vulnerability assessment and strengthening of an existing typical Indian unreinforced masonry building in a high-risk seismic zone-a comprehensive application

Seismic vulnerability assessment and strengthening of an existing typical Indian unreinforced masonry building in a high-risk seismic zone-a comprehensive application

Seismic Isolation via I-Shaped and T-Shaped Large-Scale Phononic Metamaterials

In this work, the attenuation of surface seismic waves from large-scale phononic metamaterials is numerically studied. The proposed metamaterials consist of rectangular trenches that form either I-shaped or T-shaped cavities embedded at the ground surface. The numerical investigation includes the study of the response of the proposed structures for different values of their geometric parameters. In addition, modifications of the proposed structures where heavy cores coated with a soft material were considered in the cavities were also numerically studied. For a more realistic numerical approach, the transmission spectrum of a selected large-scale phononic metamaterial was also investigated in a suitable half-space numerical scheme. The results of the present research showed that the studied large-scale metastructures could be a very promising potential candidate for seismic shielding applications for the protection of existing urban or countryside structures. The proposed metamaterials are low in cost and easy to construct for the protection of existing buildings, critical infrastructures, or even entire urban areas without need for any kind of intervention at them, therefore providing an effective solution in the field of seismic isolation.

Hybrid-self-centring damper for industrial structure: Development and experimental validation

The seismic resilience of industrial structures has become an increasingly important issue considering the significant role played by industrial structures in social operation and post-earthquake rescue. This study presented a hybrid-self-centring damper (HSCD) incorporating the variable friction device (VFD) and tapered strip device (TSD) for tunable hysteretic behaviour to improve the seismic resilience of industrial structures. The notion of the proposed HSCD was first illustrated, followed by an experimental study on eight test specimens to understand the behaviour of the proposed damper. The test results confirmed that the proposed HSCD exhibited the expected energy dissipation sequence with excellent self-centring ability. The energy dissipation sequence of the proposed HSCD can be achieved by successive inelasticity consisting of the VFD and TSD, and the hysteretic behaviour can be flexibly adjusted by modulating the essential parameters. Detailed finite element models validated by the experimental results were developed to take insights into the behaviour of the proposed damper. To facilitate the design in the seismic application of the proposed HSCD, the hysteretic model and nonlinear-spring-based simplified model verified by test results were developed to quantify the hysteretic behaviour of the proposed damper.

Probabilistic pre-conditioned compound landslide hazard assessment framework: integrating seismic and precipitation data and applications

While landslides have been extremely researched, there is a notable gap in the literature regarding the combined impact of precipitation-induced and earthquake-induced landslide events on a large scale. This study presents an approach to evaluate pre-conditioned compound hazards, examining the individual and combined effects of seismic and precipitation-induced landslides. Utilizing a diverse dataset comprising precipitation, seismic, geological, and geotechnical data, the analysis includes assessments of seismic sliding displacements and precipitation-induced slope stability over a wide geographic area (Iran with ~ 1.7 million km2). We conducted discrete and joint hazard analyses to gain insights into combined seismic and precipitation-induced landslide hazards. A total of over 39,000 analyses were conducted to portray the proposed framework. The analysis indicated a higher likelihood of slope failure during earthquakes compared to precipitation-induced events. However, the combined impacts of both hazards result in significantly elevated hazard levels according to our assessments. Specifically, the joint analysis revealed that the sequence order of events can influence hazard levels. When an earthquake follows heavy precipitation, the landslide hazard level significantly increases compared to when precipitation follows an earthquake. These findings suggest that a discrete hazard analysis may underestimate hazards compared to a joint hazard analysis, especially when events occur sequentially. Comparisons between predicted and observed hazards for historical cases support the effectiveness of our proposed approach in predicting hazard levels. Overall, our proposed compound landslide hazard analysis provides a valuable tool for risk assessment and preparedness, aiding in mitigating the impact of pre-conditioned landslides.

Seismic shear wave noise suppression and application to well tie

Abstract Seismic shear waves have been used in oil and gas exploration for decades. A 2D seismic shear wave inline was acquired in the Sanhu area, located in the Qaidam Basin in western China. Although the acquired shear wave data showed high resolution and comparable bandwidth to the compressional wave, it was contaminated by various types of noise, including linear noise, single-frequency noise, and especially internal multiples. Internal multiples seriously compromise the primary reflections at both near offset and far offset and are difficult to suppress. This paper presents a case study of noise attenuation for the seismic shear wave. First, single-frequency noise and linear noise are attenuated through filtering methods. Then, two methods (the Radon transform and the frequency-wavenumber (F-K) filtering) are evaluated for their effectiveness in multiple suppression and amplitude preservation. The results indicate that both methods successfully reduce long-period multiples at the far offset, enhancing the signal-to-noise ratio. We show that F-K filtering retains the characteristic ‘strong–weak–strong’ amplitude variation in the SV-SV-wave data gather, making it preferable for subsequent amplitude variation with offset analysis and inversion. Finally, a wave equation-based multiple suppression inversion method is used to suppress near-offset internal multiples. This involves iteratively predicting internal multiples and adaptively subtracting them from the original data. Stacked sections of different offsets are compared to demonstrate the de-multiple result, and the result is also validated by the improvement of well calibration with the seismic shear wave data.

Performance of an advanced sand constitutive model in modelling soil and soil-structure interaction under seismic excitation

There is a growing number of available advanced soil constitutive models aimed at capturing soil cyclic behaviour and their subsequent use in seismic applications. Nevertheless, detailed validation studies of these soil constitutive models on benchmark experimental works including seismic soil-structure interaction are still rare. This work presents a short validation study of the seismic performance of an advanced elastoplastic sand constitutive model on a boundary value problem including kinematic and inertial soil-structure interaction. The results of the finite element numerical model for the free field and structural responses are compared with the experimental work on a group of piles analysed in a flexible soil container filled with dry sand and subjected to simplified seismic loading. In general, the comparisons show a satisfactory match between the results of the simulations and the experiments, with the exception of the numerical predictions of settlements. The computed results are discussed based on: i) the dominant stress-paths in soil; ii) parametric studies on the settlement evaluation; iii) the origin of the high frequency motion oscillations to simple sinusoidal input motions; all with respect to potential improvements in the formulation of the elastic behaviour of the constitutive model in the future.

Experimentally Calibrated Numerical Modeling for Performance-Based Seismic Design of Low-Frequency Rocking Electrical Cabinets

ABSTRACT In this paper, the performances of two electrical cabinet specimens during shake table tests are described in detail in terms of the observed damage and their measured peak transient responses. The measured responses of the specimens are then used to calibrate simplified numerical models of the two specimens. The calibrated numerical models are used to derive fragility functions to illustrate their usefulness in performance-based seismic design applications. The calibration process and the basic concept of the numerical models can be used for calibrating similar numerical models for other typologies of similar floor anchored non-structural elements that have a rocking failure mode.

Distributed Vibration Sensing Based on a Forward Transmission Polarization-Generated Carrier.

For distributed fiber-optic sensors, slowly varying vibration signals down to 5 mHz are difficult to measure due to low signal-to-noise ratios. We propose and demonstrate a forward transmission-based distributed sensing system, combined with a polarization-generated carrier for detection bandwidth reduction, and cross-correlation for vibration positioning. By applying a higher-frequency carrier signal using a fast polarization controller, the initial phase of the known carrier frequency is monitored and analyzed to demodulate the vibration signal. Only the polarization carrier needs to be analyzed, not the arbitrary-frequency signal, which can lead to hardware issues (reduced detection bandwidth and less noise). The difference in arrival time between the two detection ends obtained through cross-correlation can determine the vibration position. Our experimental results demonstrate a sensitivity of 0.63 mrad/με and a limit of detection (LoD) of 355.6 pε/Hz1/2 at 60 Hz. A lock-in amplifier can be used on the fixed carrier to achieve a minimal LoD. The sensing distance can reach 131.5 km and the positioning accuracy is 725 m (root-mean-square error) while the spatial resolution is 105 m. The tested vibration frequency range is between 0.005 Hz and 160 Hz. A low frequency of 5 mHz for forward transmission-based distributed sensing is highly attractive for seismic monitoring applications.

A Benchmarking Method to Rank the Performance of Physics-Based Earthquake Simulations

Abstract Physics-based earthquake simulators are an increasingly popular modeling tool in earthquake forecasting for seismic hazard as well as fault rupture behavior studies. Their popularity comes from their ability to overcome completeness limitations of real catalogs, and also because they allow reproducing complex fault rupture and interaction patterns via modeling the physical processes involved in earthquake nucleation and propagation. One important challenge of these models revolves around selecting the physical input parameters that will yield the better similarity to earthquake relationships observed in nature, for instance, the frictional parameters of the rate-and-state law—a and b—or the initial normal and shear stresses. Because of the scarcity of empirical data, such input parameters are often selected by trial–error exploration and predominantly manual model performance analyses, which can overall be time consuming. We present a new benchmarking approach to analyze and rank the relative performance of simultaneous earthquake simulation catalogs by quantitatively scoring their combined fit to three reference function types: (1) earthquake-scaling relationships, (2) the shape of the magnitude–frequency distributions, and (3) the rates of the surface ruptures from paleoseismology or paleoearthquake occurrences. The approach provides an effective and potentially more efficient approximation to easily identify the models and input parameter combinations that fit more closely to earthquake relations and behavior. The approach also facilitates the exhaustive analysis of many input parameter combinations, identifying systematic correlations between parameters and model outputs that can potentially improve the overall understanding of the physics-based models. Finally, we demonstrate how the method results agree with the published findings in other earthquake simulation evaluations, a fact that reinforces its overall usefulness. The model ranking outputs can be useful for subsequent analyses, particularly in seismic hazard applications, such as the selection of appropriate earthquake occurrence rate models and their weighting for a logic tree.

A new hold-down device for seismic applications in CLT buildings: Design, testing, analytical and numerical assessment

An hold-down device was designed and tested experimentally for seismic applications (Seismic Hold-Down, SHD). The magnitude of axial strength as well as the up-lift displacement capacity is mainly oriented to cross-laminated-timber (CLT) buildings with medium-to-high ductility. There is consensus in the design practice of hold-down devices for promoting, as dissipative mechanism, flexural yielding of the dowel-type connectors (e.g. nails, screws) combined to timber’s embedment. Few are the applications aiming at hold-down’s optimization as it emerges from up-to-date literature review. In this framework, SHD was conceived such that tension break-out of steel was promoted at a reduced segment (fuse) of the vertical flange. In the context of capacity design, a convenient overstrength was assumed for other failure mechanisms, e.g. anchor’s pullout, laterally-loaded steel-to-timber connection, shear-plug of timber. The mechanical response under axial load was consistently predicted by an application of the component method, i.e. global stiffness and strength were analytically-derived using simplified series of non-linear mechanical springs. Three-dimensional finite elements model was employed to investigate buckling of the fuse which might occur, during cycles, when up-lift displacement is recovering. In practical circumstances, it is recommend to design SHD devices of a CLT wall similarly to diagonal elements of steel bracings, i.e. neglecting the contribution of the compressed element. Finally, SHD is compared to traditionally-designed hold-down, which mostly dissipates (and fails) at steel-to-timber connection. Overall hysteretic response is somehow similar, although the inelastic mechanisms involved are different. For example, both the devices show pinching during cycles which for SHD is mainly caused by buckling of the fuse while, for traditional hold-down, is dominated by timber’s embedment.

Phase unwrapping error identification and suppression method in φ-OTDR systems based on PELT-VMD-ARIMA.

In phase-sensitive optical time-domain reflectometer (φ-OTDR) systems, phase unwrapping errors can distort vibration information. To address this issue, a phase unwrapping error identification and suppression method combining pruned exact linear time (PELT) changepoint detection, variational mode decomposition (VMD), and autoregressive integrated moving average (ARIMA) models, termed PELT-VMD-ARIMA, is proposed. Firstly, the principle of the proposed method is introduced, and its effectiveness is verified through a series of numerical simulation experiments. Next, piezoelectric transducers (PZTs) are employed as seismic sources in experiments involving single-frequency and chirp signals. Compared to the mean-shift method, the proposed method reduces the average root mean square error (RMSE) by 70.36% within 2δ range around the changepoints. Finally, the proposed method was validated through an active source seismic application. The results demonstrate the efficacy of the proposed method in identifying and suppressing phase unwrapping errors, thereby enhancing signal quality. This method enhances the vibration recognition capability of φ-OTDR systems, which facilitates precise distributed acoustic sensing applications.

Evaluating Automated Seismic Event Detection Approaches: An Application to Victoria Land, East Antarctica

AbstractAs seismic data collection continues to grow, advanced automated processing techniques for robust phase identification and event detection are becoming increasingly important. However, the performance, benefits, and limitations of different automated detection approaches have not been fully evaluated. Our study examines how the performance of conventional techniques, including the Short‐Term Average/Long‐Term Average method and cross‐correlation approaches, compares to that of various deep learning models. We also evaluate the added benefits that transfer learning may provide to machine learning‐based seismic applications. Each detection approach has been applied to 3 years of data recorded by stations in East Antarctica, which contain both cryospheric and tectonic‐related seismic events. Our results demonstrate that by integrating different event detection approaches with transfer learning, the strengths of each approach can maximize seismic detection accuracy and reliability while also enhancing the completeness of associated event catalogs. That said, model performance can vary, depending on the quality of the data set as well as the data augmentation schemes and training strategies employed. Our results in East Antarctica provide new insight into polar seismicity and demonstrate how automated event detection approaches can be optimized to investigate seismic activity in challenging environments.