Processes Across Scales:
From Crystals to Continents
Session A (1 pm)
Embedded Files
A01 - Source mechanism study based on 3D model using high-frequency waveforms from earthquakes in the central Macquire Ridge Complex, Zhi Wei, Postdoc
Source Mechanism Study Based on 3D Model Using High-Frequency Waveforms from Earthquakes in the Central Macquarie Ridge Complex
Zhi Wei1, Hrvoje Tkalčić1, Jinyin Hu1, Nicholas Rawlinson2, Caroline Eakin1, Sheng Wang1, Millard F. Coffin3, Robert Pickle1, Thanh-Son Pham1, Joann Stock4 and MRC team
1Research School of Earth Sciences, Australian National University, Canberra, 2601, Australia
2Department of Earth Sciences, University of Cambridge, Cambridge, CB3 0EZ, United Kingdom
3Institute for Marine and Antarctic Studies, University of Tasmania, Hobart, TAS 7001, Australia
4Seismological Laboratory, California Institute of Technology, Pasadena, CA 91125, USA
The Macquarie Ridge Complex (MRC) evolved from a spreading mid-ocean ridge and is now dominated by a transpressional plate boundary. The MRC is unique: Macquarie Island represents the only exposure of mantle rock above sea level in the ocean, the region hosts some of the largest underwater strike-slip earthquakes on the planet, and it raises the fundamental question of whether this region is experiencing underthrusting or at the stage of incipient subduction. To investigate this question, our team deployed temporary land stations and a temporary network of ocean-bottom seismometers (OBSs) in the Furious Fifties near Macquarie Island during 2020–2021. The final seismic dataset consists of recordings from six land stations and fifteen recovered OBSs, from which thousands of local earthquakes have been detected, including several moderate events (M4.0–5.5) that were well located. Because earthquake source mechanisms provide key insights into tectonic motion, we focus on these moderate events to investigate their source characteristics and evaluate the state of the ridge. For this purpose, we analyzed 5–20 s waveforms using a Bayesian inversion framework that explicitly accounts for noise to obtain reliable source mechanism estimates based on a 3D velocity model of the region.
A02 - Systematic Detection of Glacial Earthquakes in Thwaites Glacier, West Antarctica, by Regional Surface Waves, Thanh-Son Pham, Postdoc
Systematic Detection of Glacial Earthquakes in Thwaites Glacier, West Antarctica, by Regional Surface Waves
Thanh-Son Phạm1
1Research School of Earth Sciences, The Australian National University, Canberra, ACT, Australia
Glacial earthquakes are a class of seismic sources generated by the capsize of icebergs calved from glacier termini. Although being teleseismically observed in Greenland glaciers, seismic detection of such events in Antarctica has been elusive. Here, we develop an automatic detection algorithm employing the coherence of relatively short-period Rayleigh waves recorded by the regional seismic network available in Antarctica. The application to the 2010–2023 dataset resulted in a catalog of 368, largely uncatalogued, Ms 2–3, seismic events, associated with Thwaites and Pine Island Glaciers. The correlation of the occurrence frequency of Thwaites events with the episodic speed-ups of its frontal ice tongue between 2018 and 2020 infers their capsizing glacial earthquake nature. However, the Pine Island Glacier events, mainly located near the grounding line, remain puzzling. This study proved the popularity of glacial earthquakes in Antarctica with different characteristics from their Greenland counterparts, which warrant further investigation.
A03 - Heterogeneous Excitation of Primary Microseisms Under Cyclone Forcing, Abhay Pandey, Student
Heterogeneous Excitation of Primary Microseisms Under Cyclone Forcing
Abhay Pandey, Thanh-Son Phạm & Hrvoje Tkalčić
Primary microseisms are long-period seismic waves generated by the interaction of ocean surface gravity waves with the seafloor, yet the spatial distribution and excitation mechanisms of their sources remain poorly resolved. Here, we investigate the heterogeneous generation of primary microseisms associated with tropical cyclone Ita (2014). Combining continuous seismic data from a regional network with high-resolution (30 m) bathymetric maps, we show that primary microseism excitation is highly localized in both time and space, and critically dependent on fine-scale seafloor roughness. Our analysis reveals that the most coherent and energetic Rayleigh wave bursts arise when the cyclone traverses shallow region with pronounced bathymetric variability, where the coupling between ocean waves and the solid Earth is most efficient. The finding provide observational evidence to confirm theoretical conjecture that topographic undulations at scales comparable to ocean wave wavelengths govern the strength of microseism sources in 0.05 – 0.1 Hz frequency band. Collectively, this study highlights the critical role of nearshore bathymetric roughness in shaping the spatial coherence of primary microseism excitation, emphasizing its importance in source modeling and the identification of regions with efficient ocean–atmosphere–solid Earth coupling for the strategic deployment of seismic sites for long-term, climate-sensitive ocean wave monitoring.
A04 - A next-gen catalogue of global moderate earthquakes (Mw6.0-7.0) incorporating source time functions and uncertainty estimates, Jinyin Hu, Postdoc
A next-gen catalogue of global moderate earthquakes (Mw6.0–7.0) incorporating source time functions and uncertainty estimates
Jinyin Hu1, *, Hrvoje Tkalčić1, Thanh-Son Phạm1 and Babak Hejrani1,2
1Research School of Earth Sciences, The Australian National University, Canberra, ACT, Australia
2Geoscience Australia, Australian Capital Territory, Australia
Earthquake catalogues have been established for decades at regional and global scales, routinely providing earthquake locations and source mechanisms. The latter is commonly represented by a point-source moment tensor (a symmetric 3×3 matrix) combined with the event location, known as the centroid moment tensor (CMT). CMTs are determined through seismic source inversion, in which synthetic waveforms predicted from trial CMTs within an Earth model are matched to recorded seismograms. With advances in inversion techniques, several catalogues (most notably the Global CMT) have achieved great success. Recently, increasing attention has shifted toward quantifying the uncertainties of CMT solutions, arising from both measurement error in recorded waveforms and theory error in synthetic waveforms due to the imperfect Earth model. In this study, we aim to determine CMT solutions and their associated uncertainties for moderate-size earthquakes (Mw 6.0–7.0). These events are critical for understanding earthquake source physics and regional structures, yet are often less well constrained than larger earthquakes. In addition, we simultaneously estimate source time functions (STFs) to investigate the energy release history of these events. As a first step, we analyze four earthquakes in diverse tectonic settings to demonstrate the improved solutions compared to existing catalogues for CMTs and STFs after explicitly incorporating uncertainties from measurement error and theory error. Our findings highlight the importance of treating uncertainties in seismic source inversions and represent a step toward the development of a next-generation earthquake catalogue with robust CMT, STF, and uncertainty estimates.
A05 - Ambient Noise Tomography Reveals High Velocity Zone in the Mid-to-lower Crust along the Southwestern Edge of the Yilgarn Craton, Australia, Ping Zhang, Postdoc
Ambient Noise Tomography Reveals High Velocity Zone in the Mid-to-lower Crust along the Southwestern Edge of the Yilgarn Craton, Australia
Ping Zhang, Meghan S. Miller, Fabrizio Magrini, Robert Pickle, Huaiyu Yuan, Ruth Murdie, Klaus Gessner, and Raphael Quentin de Gromard
The evolution of southwest Australia, including Precambrian cratonization and multiple episodes of supercontinental assembly and breakup from Proterozoic to Phanerozoic, remains debated, in part due to previously poorly constrained crustal structures. This study presents a new 3-D shear wave velocity (Vs) model in the crust and uppermost mantle of southwest Australia using ambient noise tomography with newly collected data by SWAN array. The model features a high-velocity zone in the middle-to-lower crust along the southwestern edge of the Yilgarn Craton but east of the Darling Fault, a structural feature imaged for the first time. This anomaly is likely related to magmatic underplating processes during continental breakup and rifting localized on a pre-existing weak zone. The imaged heterogeneity in the deep crust across the western margin of the Yilgarn Craton implies that the tectonomagmatic processes that have shaped craton margins can affect larger parts of continental crust at depth compared to what is apparent at Earth’s surface.
A06 - Advancing the Understanding of Polar Ice Structure Using Passive Seismic Studies, Ammu Sanjayan, Student
Advancing the Understanding of Polar Ice Structure Using Passive Seismic Studies
Ammu Sanjaya1, Thanh-Son Phạm1, Hrvoje Tkalčić1
1Research School of Earth Sciences, The Australian National University, Canberra, ACT, Australia
Accurate characterization of the P-to-S wave speed ratio (Vp/Vs) within the Antarctic Ice Sheet (AIS) is crucial for understanding its internal structure, mechanical stability, and vulnerability to climate change. In this study, we apply autocorrelation to steeply arriving P- and S-waves and their subsequent reverberations from distant earthquakes, i.e., teleseismic coda, to reveal their reflectivity within the ice sheet, which enables the estimation of Vp/Vs from P- and S-wave reflection time picks. By incorporating shear-wave reflections retrieved from S-wave coda, in addition to those obtained from P-wave coda autocorrelation, we significantly expand the Vp/Vs dataset across Antarctica. This enhancement leads to 28 new, reliable Vp/Vs estimates, nearly doubling the number of previously resolved sites to a total of 61, and improving spatial coverage including the remote interiors, particularly with increased resolution in East Antarctica. The updated Vp/Vs map reveals distinct regional contrasts: East Antarctica shows relatively stable ratios close to the standard value of 2 for isotropic ice, while West Antarctica exhibits greater variability, with elevated Vp/Vs values around 2.10 to 2.2 in central regions and lower values toward the coastal margins. These variations correlate broadly with independent geothermal heat flow estimates—ranging from 40–70 mW m⁻² in East Antarctica to over 120 mW m⁻² in parts of West Antarctica. We interpret the elevated Vp/Vs ratios in these regions as evidence of possibly higher englacial temperature, warmer or soft sediment layers that reduce shear-wave velocities. Synthetic modeling supports this interpretation by showing that such conditions significantly influence both Vp/Vs ratios and reflection timing.
A07 - M-ICE-STERIOUS SIGNALS: Investigating Anomalous Glacial Earthquakes at the Pine Island Glacier, Antarctica, Alexandra Vickery, Student
M-ICE-STERIOUS SIGNALS:
Investigating Anomalous Glacial Earthquakes at the Pine Island Glacier, Antarctica
Alexandra Vickery, Thanh-Son Phạm, Malcolm Sambridge
In a rapidly changing world, few places are changing more rapidly than West Antarctica. Pine Island Glacier, one of Antarctica's biggest - and one of its most rapidly melting - is a highly active cryosphere environment. However, its dynamics are not well understood. Reasonably large, low frequency seismic events known as "glacial earthquakes" (GEQ) have been detected and investigated at a number of (mostly Greenlandic) glacier sites, and associated with glacial calving processes. A number of such events have been detected at the Pine Island Glacier (Phạm 2025); however, they appear to occur tens of kilometres inland and thus are likely not generated from this same process.
This investigation aims to characterise these anomalous GEQ-like Pine Island Glacier events, through analysis of the low-frequency signals the events were identified with, and the high-frequency signals collected locally and regionally prior to and during the event (including from on the glacier itself). This analysis is supplemented with an examination of satellite data and an application of the newly-developed “trans-conceptual Bayesian sampling” technique (Sambridge et al, 2025), towards characterising the event focal mechanisms.
The analysis suggests that while the Pine Island Glacier events share many superficial characteristics with other GEQ events, they do not appear to follow the same generation mechanism, perhaps instead being related to larger glacial movements and events. The Pine Island Glacier events thus may represent a distinct category, and provide further insight into the cryosphere dynamics active in the region.
A08 - Distinct lithospheric anisotropic fabrics across Southwestern Australia and the Yilgarn Craton revealed by Phase 1 of the WA Array, Miriam Gauntlett, Postdoc
Distinct lithospheric anisotropic fabrics across Southwestern Australia and the Yilgarn Craton revealed by Phase 1 of the WA Array
Miriam Gauntlett, Caroline Eakin, Nitarani Bishoyi, Ping Zhang, J.-P. O’Donnell, Ruth Murdie, Meghan Miller, Robert Pickle, Reza Ebrahimi
The southwest region of Western Australia (WA) is one of the oldest continental regions on Earth, hosting the Archean Yilgarn Craton, first formed over two billion years ago. The subsequent deformation undergone by the craton is recorded in large-scale rock fabric, which can manifest in the directional dependence of seismic velocity, also known as seismic anisotropy. Mantle flow in the asthenosphere may also contribute to seismic anisotropy. We use new broadband seismic arrays across the southwest of WA to calculate seismic anisotropy parameters. We find evidence for coherent seismic anisotropy, with magnitudes comparable to global averages. The direction of anisotropy shows variation across the area but is not aligned with current plate motion and the expected mantle flow direction. Instead, we propose that the anisotropy we observe is reflective of past deformation and perhaps linked to the formation of this ancient continent.
A09 - A Three-Dimensional Model of Submarine Groundwater Discharge for Australian Coastal Regions, Sruthy Sajeev, Student
A Three-Dimensional Model of Submarine Groundwater Discharge for Australian Coastal Regions
Sruthy Sajeev, Louis Moresi, Juan Carlos Graciosa, Neng Lu
Submarine groundwater discharge (SGD), the flow of fresh and saline groundwater into the ocean is an important but often under-quantified component of the global water and chemical budgets. The fresh fraction of SGD is particularly significant because of its high solute and nutrient loads. It has been estimated to contribute up to 10% of global river discharge and to rival riverine inputs for solutes such as carbon, iron, silica, and strontium. Moreover, fresh SGD may help buffer ocean acidification through the delivery of groundwater alkalinity.
Here, we present the first steps toward developing a spatially resolved numerical model of coastal groundwater discharge. As an initial stage, we performed numerical benchmarks of groundwater flow and contaminant transport using Underworld3. We evaluated model performance through L2-norm error calculations and validated against analytical solutions. In addition, we numerically benchmarked the Henry problem and calculated velocity root-mean-square (Vrms) to check whether the solution has attained steady-state. These benchmarks lay the groundwork for robust large-scale simulations of SGD.
A10 - Imaging Lake George Fault Zone with Traffic Noise, Chengxin Jiang, Postdoc
Imaging Lake George Fault Zone with Traffic Noise
Chengxin Jiang and Meghan Miller
Mapping fault‐zone properties is crucial for mitigating seismic hazards, particularly in urban settings. This process often requires high‐resolution seismic imaging, which depends on dense data coverage and high‐frequency seismic energy sensitive to shallow structures, with traffic noise providing an ideal source. However, extracting coherent phases from traffic noise remains challenging due to the complex conditions of variable sources and array configurations. Although array seismology techniques enhance coherence, they can limit model resolution. In this study, we demonstrate that high‐quality surface Rayleigh‐wave dispersions (2.5–10 Hz) can be extracted from single‐station‐pair cross correlations using a meticulously designed dense nodal array near a highway in the Lake George fault zone, southeast Australia. By analyzing the complete nine‐component cross‐correlation tensor, we find that radial–radial correlations, rather than conventional vertical–vertical correlations, yield the strongest dispersion signals. These high‐frequency dispersion measurements enable surface‐wave tomography, providing the first detailed velocity structure of the top 800 m of the fault zone. The mapped Lake George fault zone displays seismic characteristics similar to those of several major active fault systems worldwide. It represents a long‐lived damage zone capable of hosting significant seismic events. These results provide new insights into the application of traffic noise for near‐surface imaging and monitoring in urban environments, and they have significant implications for considering the directionality of distributed acoustic sensing data in future work.
A11 - Avalanche Localization with Distributed Acoustic Sensing Near Milford Sound/Piopiotahi, New Zealand, Konstantinos Michailos, Postdoc
Avalanche Localization with Distributed Acoustic Sensing Near Milford Sound/Piopiotahi, New Zealand
Konstantinos Michailos, Voon Hui Lai, Sebastian Konietzny, Meghan Miller, John Townend, Simon Morris, Hans-Peter Anderson, Graham Clarke
Snow avalanches are natural hazards that pose significant risks to alpine environments. Milford Sound attracts over 400,000 visitors annually via State Highway 94 (SH94), the only road connecting Milford Sound to the rest of New Zealand. Milford Road Alliance (MRA) manages the avalanche risk along SH94 using weather forecasts, meteorological data, and web-cameras. Although valuable for hazard management, MRA’s avalanche records have limited temporal and spatial accuracy, especially for distant or nighttime events.
To complement MRA’s monitoring efforts, we use seismic data by repurposing a dark telecommunication fiber running along SH94 into a 31 km long DAS array. We recorded tens of avalanches, both natural and triggered, between June and September 2024. Signal processing techniques were applied to detect, locate, and characterise these avalanche events. We investigate any potential connections between avalanche activity and weather conditions. This analysis aims to enhance our understanding of avalanche-triggering processes in the region.
A12 - Inner core velocity tomography from a deep-learning generated PKIKP catalogue, Jiarun Zhou, Student
Inner core velocity tomography from a deep-learning generated PKIKP catalogue
Jiarun Zhou, Hrvoje Tkalčić, Son Phạm
Understanding of Earth’s inner core has advanced greatly in recent decades, supported by the expansion of seismic networks and the growth of global datasets. Although previous studies have combined various inner-core–sensitive phases, their resolution has remained largely confined to the upper inner core. Meanwhile, the manual processing of large datasets continues to pose a bottleneck, restricting deeper investigations and more robust interpretations.
Deep-learning algorithms have proven effective for inner core datasets, by achieving human-level precision and allowing comprehensive analysis of datasets beyond manual processing. Here, we employ a deep-learning–generated PKIKP catalogue to conduct inner core velocity tomography, focusing on absolute travel times of PKIKP phases traversing the deep part of the inner core. Our approach demonstrates the potential to unlock hidden seismic information and to improve resolution of Earth’s deepest structure. This framework also paves the way for extending tomography to other core-sensitive phases and to next-generation seismic datasets.
A13 - Imaging Upper Mantle Structure beneath Earth's Hidden Continent of Northern Zealandia with Ambient Noise Tomography, Shixian Dong, Student
Imaging Upper Mantle Structure beneath Earth's Hidden Continent of Northern Zealandia with Ambient Noise Tomography
Shixian Dong, Chengxin Jiang, Meghan S. Miller
Zealandia, the Earth’s youngest and thinnest continent, lies in the southwestern Pacific with about 94% of its ~4.9 million km² area submerged, earning the name the “hidden continent.” Northern Zealandia has undergone intense continental rifting and crustal thinning, yet its upper mantle structure and relation to extensional mechanisms remain poorly understood. Apart from New Zealand’s North and South Islands, deep structural studies of this submarine region are scarce, lacking direct constraints on mantle structure and tectonic evolution. Sparse seismicity, mostly along the Tonga-Kermadec and New Hebrides trenches, further limits traditional body-wave tomography, making ambient noise tomography (ANT) an effective alternative.
In this study, we apply ANT to investigate the shear-wave velocity structure beneath northern Zealandia. We use seismic noise data recorded between 2010 and 2025 from 131 stations distributed across eastern Australia, New Zealand, New Caledonia, Fiji, and the Vanuatu region. Ambient noise cross-correlations are computed using the NoisePy package, with spectral whitening and phase-weighted stacking. Fundamental-mode Rayleigh wave dispersion curves at 20–100 s periods are extracted through frequency–time analysis (FTAN), and surface wave tomographic inversion is performed to obtain a new upper mantle shear-wave velocity model.
Our results reveal detailed upper mantle velocity structures beneath northern Zealandia and highlight significant differences from the Australian continental mantle. This model provides new constraints on the geodynamic processes of this hidden continent, particularly extreme continental extension and the deep driving forces behind Zealandia’s separation from Gondwana.
A14 - Alpine Fault zone structure revealed using Distributed Acoustic Sensing (DAS), Voon Hui Lai, Postdoc
Alpine Fault zone structure revealed using Distributed Acoustic Sensing (DAS)
Voon Hui Lai1, Meghan Miller1, Xin Wang2, Yujie Wang2, Chengxin Jiang1, John Townend3
1Research School of Earth Sciences, Australian National University
2Institute of Geology and Geophysics, Chinese Academy of Science
3School of Geography, Environment and Earth Sciences, Victoria University of Wellington
The Alpine Fault, which forms the boundary between the Pacific Plate and the Australian Plate, poses substantial seismic hazard to southern New Zealand, with magnitude ~8 earthquakes occurring on <300-year timescales, most recently in 1717 CE. The fault zone architecture is of particular interest, as it is hypothesized to influence coseismic fault rheology and fluid-rock interaction, which in turn affects local fault slip rates and rupture propagation patterns.
Resolving the fault zone’s architecture, including its geometry and velocity properties, requires high spatial resolution observations that can capture multiple length scales. Distributed acoustic sensing (DAS), with its dense sensors, provides a unique opportunity to investigate and image the shallow fault zone architecture.
Here we used DAS data targeting the Alpine Fault near Haast, South Westland and revealed a complex shallow velocity structure across the fault zone characterized by sedimentary deposits from a rich post-glacial depositional history, glacial erosional features, and fault damage zone structure. Waveform modelling further explains the observed fault zone trapped waves due to distinct fault strands and secondary scattering due to shallow bedrock irregularities. This high-resolution view of the Alpine Fault zone architecture near a paleoseismologically recognized segment boundary enhances our ability to assess seismic hazards and inform mitigation strategies’ late in the fault’s typical interseismic phase.
A15 - Novel application of unsupervised machine learning for characterization of subsurface seismicity, tectonic dynamics and stress distribution, Mohammad Salama, Student
Novel application of unsupervised machine learning for characterization of subsurface seismicity, tectonic dynamics and stress distribution
Mohammad Salama∗, Muhammad Tahir Iqbala, Raja Adnan Habiba, Amna Tahira,Aamir Sultana and Talat Iqbala
Centre for Earthquake Studies (CES), National Centre for Physics (NCP), Islamabad-44000, Pakistan
Our study pioneers an innovative use of unsupervised machine learning, a powerful tool for navigating unclassified data, to unravel the complexities of subsurface seismic activities and extract meaningful patterns. Our central objective is to comprehensively characterize seismicity within an active region by identifying distinct seismic clusters in spatial distribution, thereby gaining a deeper understanding of subsurface stress distribution and tectonic dynamics. Employing a diverse range of clustering algorithms, with particular emphasis on Fuzzy C-Means (FCM), our research meticulously dissects the intricate physical processes that govern a complex tectonic zone. This technique effectively delineates distinct tectonic zones, aligning seamlessly with established seismological knowledge and underscoring the transformative potential of Artificial Intelligence (AI) in analyzing regional subsurface phenomena, even under conditions of data scarcity. Moreover, associating earthquakes with specific seismogenic structures significantly enhances seismic hazard analyses, potentially paving the way for autonomous insights that inform engineering hazard assessments.
Page updated
Google Sites
Report abuse