An official website of the United States government
Antarctica is almost completely covered by the world’s largest ice sheet, and its hidden geologic structure partially controls the behavior of the ice layer. Recent advances in geophysical and remote sensing tools have allowed us to observe various transient phenomena, such as tectonic earthquakes, glacial bed slip events, and iceberg calving signals, all of which can be used to investigate solid Earth – cryosphere interactions. We analyzed seismic data collected by the TAMNNET temporary deployment as well as other stations in East Antarctica to identify and locate local icequakes, earthquakes, and other seismic events that occurred between 2012-2015. We employ two event detection approaches. The first is based on phase match filtering and waveform cross-correlation, which uses known events as templates to search through continuous data and to identify similar seismic signals. The second uses EQtransformer, a deep-learning-based event signal detector and phase picker. Event detections identified with both approaches will be compared to assess the effectiveness of these methods in East Antarctica. We also plan to use the combined constraints from our initial approaches to train a new machine-learning model and to assess its performance. Ultimately, our results will be used to evaluate automated event detection approaches for polar environments and to address fundamental questions related to tectonic-cryospheric interactions.
more »
« less
Comparison of Automatic Event Detection Algorithms: A Case Study in East Antarctica
As seismic data availability increases, the necessity for automated processing techniques has become increasingly evident. Expanded geophysical datasets collected over the past several decades across Antarctica provide excellent resources to evaluate different event detection approaches. We have used the traditional Short-Term Average/Long-Term Average (STA/LTA) algorithm to catalogue seismic data recorded by 19 stations in East Antarctica between 2012 and 2015. However, the complexities of the East Antarctic dataset, including low magnitude earthquakes and other types of seismic events such as icequakes or firnquakes, warrant more advanced automated detection techniques. Therefore, we have also applied template matching as well as several deep learning algorithms, including Generalized Phase Detection (GPD), PhaseNet, BasicPhaseAE, and EQTransformer (EQT), to identify seismic phases within our dataset. Our goal is not only to increase the volume of detectable seismic events but also to gain insights into the effectiveness of these different automated approaches. Our assessment evaluates the completeness of the newly generated catalogs, the precision of identified event locations, and the quality of the picks. The performance of these different event detection techniques applied to continuous seismic data from a polar environment will be highlighted. We will also identify potential limitations and necessary adjustments for deep learning algorithm training, which is essential for their reliable application to specific datasets.
more »
« less
- Award ID(s):
- 1914698
- PAR ID:
- 10488350
- Publisher / Repository:
- American Geophysical Union
- Date Published:
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract We apply a machine learning (ML) earthquake detection technique on over 21 yr of seismic data from on-continent temporary and long-term networks to obtain the most complete catalog of seismicity in Antarctica to date. The new catalog contains 60,006 seismic events within the Antarctic continent for 1 January 2000–1 January 2021, with estimated moment magnitudes (Mw) between −1.0 and 4.5. Most detected seismicity occurs near Ross Island, large ice shelves, ice streams, ice-covered volcanoes, or in distinct and isolated areas within the continental interior. The event locations and waveform characteristics indicate volcanic, tectonic, and cryospheric sources. The catalog shows that Antarctica is more seismically active than prior catalogs would indicate, examples include new tectonic events in East Antarctica, seismic events near and around the vicinity of David Glacier, and many thousands of events in the Mount Erebus region. This catalog provides a resource for more specific studies using other detection and analysis methods such as template matching or transfer learning to further discriminate source types and investigate diverse seismogenic processes across the continent.more » « less
-
The structure of the Antarctic crust is important to our understanding of processes occurring within the Antarctic cryosphere as well as to the Earth’s response to ice mass loss. With the increase in geophysical studies of Antarctica, crustal structure has become much better defined beneath many regions. Several crustal models have been created from seismic-derived and/or gravity-derived data, and some of these models incorporate sets of crustal receiver functions either as a priori constraints or to validate model results. However, receiver function constraints do not exist throughout large regions of Antarctica due to a lack of seismic coverage; given this, we search for additional metrics by which we can compare and contrast Earth models. One approach that has been utilized for other continents is to forward model accurate synthetic waveforms through existing seismic velocity models to identify which models most accurately reproduce seismic waveform datasets. Such waveform datasets may come from accurately determined seismic events or from ambient seismic noise. In an effort to assess existing Antarctic crustal models using a different metric to identify regions where crustal structure is still most uncertain, we have collected a suite of available seismic- and gravity-derived Antarctic crustal models. In the absence of accurately determined ‘ground-truthed’ seismic events in Antarctica, we use a frequency-time normalization approach to extract Rayleigh waves from ambient seismic noise, with periods of 15-55 seconds that are sensitive to crustal structure. We split the observations into two separate validation datasets. The first dataset includes all station-station cross-correlations, with at least one seismic station in each pair that has not been previously used to constrain prior tomographic inversions (a true validation dataset), and the second dataset includes all available station-station cross-correlations, including those that may have been used to constrain some of the models we are testing. We construct sets of Earth models from the available crustal models underlain by two different upper mantle models. We forward model synthetic waveforms using a finite difference approach through each of the Earth models and measure the phase delays between the synthetic waveforms and the ambient seismic noise dataset. Results from our waveform validation study and identification of the poorly characterized regions of Antarctic crust are forthcoming and will be presented.more » « less
-
Many seismic tomography investigations have imaged the East Antarctic lithosphere as a thick and continuous cratonic structure that is separated from the thinner lithosphere of the adjacent West Antarctic Rift System by the Transantarctic Mountains. However, recent studies have painted a more complicated picture, suggesting, for instance, a separate cratonic fragment beneath Dronning Maud Land and possible lithospheric delamination beneath the southern Transantarctic Mountains. In addition, patterns of intracratonic seismicity have been identified near the Gamburtsev Subglacial Mountains in East Antarctica, indicating possible rift zones in this region. That said, detailed imaging of the subsurface structure has remained challenging given the sparse distribution of seismic stations and the generally low seismicity rate throughout the interior of East Antarctica. Therefore, new approaches that can leverage existing seismic datasets to elucidate the Antarctic cratonic structure are vital. We are utilizing records of ambient seismic noise recorded by numerous temporary, moderate-term, and long-term seismic networks throughout Antarctica to improve the imaging of the lithospheric structure. Empirical Green’s Functions with periods of 40-340 seconds have been extracted using a frequency-time normalization approach, and these data are being used to constrain our full-waveform inversion. A finite-difference approach with a continental-scale, spherical grid is employed to numerically model synthetic seismograms, and a scattering integral method is used to construct the associated sensitivity kernels. Our initial results suggest that some portions of East Antarctica, particularly those beneath the Wilkes Subglacial Basin and the Aurora Basin, may have reduced shear-wave velocities that potentially indicate regions of thinner lithosphere. Further, possible segmentation may be present in the vicinity of the Gamburtsev Subglacial Mountains. Our new tomographic results will allow for further assessment of the East Antarctic tectonic structure and its relation to local seismicity.more » « less
-
Abstract Advances in visual perceptual tasks have been mainly driven by the amount, and types, of annotations of large-scale datasets. Researchers have focused on fully-supervised settings to train models using offline epoch-based schemes. Despite the evident advancements, limitations and cost of manually annotated datasets have hindered further development for event perceptual tasks, such as detection and localization of objects and events in videos. The problem is more apparent in zoological applications due to the scarcity of annotations and length of videos-most videos are at most ten minutes long. Inspired by cognitive theories, we present a self-supervised perceptual prediction framework to tackle the problem of temporal event segmentation by building a stable representation of event-related objects. The approach is simple but effective. We rely on LSTM predictions of high-level features computed by a standard deep learning backbone. For spatial segmentation, the stable representation of the object is used by an attention mechanism to filter the input features before the prediction step. The self-learned attention maps effectively localize the object as a side effect of perceptual prediction. We demonstrate our approach on long videos from continuous wildlife video monitoring, spanning multiple days at 25 FPS. We aim to facilitate automated ethogramming by detecting and localizing events without the need for labels. Our approach is trained in an online manner on streaming input and requires only a single pass through the video, with no separate training set. Given the lack of long and realistic (includes real-world challenges) datasets, we introduce a new wildlife video dataset–nest monitoring of the Kagu (a flightless bird from New Caledonia)–to benchmark our approach. Our dataset features a video from 10 days (over 23 million frames) of continuous monitoring of the Kagu in its natural habitat. We annotate every frame with bounding boxes and event labels. Additionally, each frame is annotated with time-of-day and illumination conditions. We will make the dataset, which is the first of its kind, and the code available to the research community. We find that the approach significantly outperforms other self-supervised, traditional (e.g., Optical Flow, Background Subtraction) and NN-based (e.g., PA-DPC, DINO, iBOT), baselines and performs on par with supervised boundary detection approaches (i.e., PC). At a recall rate of 80%, our best performing model detects one false positive activity every 50 min of training. On average, we at least double the performance of self-supervised approaches for spatial segmentation. Additionally, we show that our approach is robust to various environmental conditions (e.g., moving shadows). We also benchmark the framework on other datasets (i.e., Kinetics-GEBD, TAPOS) from different domains to demonstrate its generalizability. The data and code are available on our project page: https://aix.eng.usf.edu/research_automated_ethogramming.htmlmore » « less
