As 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 (STA/LTA) 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 applications. Each detection approach has been applied to three years of seismic data recorded by stations in East Antarctica. Our results emphasize that the most appropriate detection approach depends on the data attributes and the study objectives. STA/LTA is well-suited for applications that require rapid results even if there is a greater likelihood for false positive detections, and correlation-based techniques work well for identifying events with a high degree of waveform similarity. Deep learning models offer the most adaptability if dealing with a range of seismic sources and noise, and their performance can be enhanced with transfer learning, if the detection parameters are fine-tuned to ensure the accuracy and reliability of the generated catalog. Our results in East Antarctic provide new insight into polar seismicity, highlighting both cryospheric and tectonic events, and demonstrate how automated event detection approaches can be optimized to investigate seismic activity in challenging environments.

We attempt to construct a timeline of The Hunga Tonga – Hunga Ha’apai eruption on 15 January 2022 through analyses of seismic, barometric, infrasonic, lightning, and satellite data. Satellite imagery at 04:00 UTC showed no ash in the air, but by 04:10 UTC, a plume had risen to 18 km. Over the next 20 minutes, the plume rose to 58 km. USGS determined that Mw5.8 volcanic earthquake of unknown mechanism had occurred at 04:14:45. Gravity waves were observed in satellite imagery, and barometric and infrasound stations around the world recorded ultra-low frequency pressure variations of more than 100 Pa, inducing ground-coupled airwaves around the globe, and meteo-tsunamis in the Caribbean Sea and Mediterranean Sea. Tsunami waves were recorded in coastal areas around the Pacific Ocean. From record sections, we determined speeds of 3.9 km/s and 299 m/s for the initial seismic and infrasound signals respectively, converging to an eruption onset time of ~0402 UTC ± 1 minute. The global pressure pulse has a speed of ~314 ± 3 m/s, consistent with theoretical models for Lamb waves (Bretherton, 1969), suggesting an origin time of ~0415 ± 2 minutes (consistent with the Mw5.8 volcanic earthquake, and sharp increases in lightning flash rates), and peaking around ~0429 ± 2 minutes. We suggest that Surtseyan volcanic activity commenced at ~04:02, building to a sub-Plinian eruption ~7 minutes later, before a phreato-Plinian eruption commenced at ~04:14. The peak Lamb wave amplitude at the closest station (757 km from HTHH) was 780 Pa. Assuming geometrical spreading like 1/√r (where r is the source-receiver distance), we estimate a lower bound of ~23 kPa for reduced pressure by extrapolation back to 1 km. Adding a near field term that decays like 1/r, we estimate an upper bound of 170 kPa for reduced pressure. Comparison of these values with those from other eruptions (McNutt et al. in this session) suggests the 15 January HTHH eruption was in the VEI 5-6 range.

An M5 earthquake occurred on November 7th, 2016, near the city of Cushing in Oklahoma, the largest crude oil storage site in the USA, after nearby disposal wells had been shut-in responding to three M4+ earthquakes in 2015. In this study, we investigated the rupture process of these M4+ events with finite fault model (FFM) inversions and computed Coulomb stress changes during this Cushing sequence. We found that the rupture processes of the four M4+ earthquakes are very complex, and they appeared to trigger one another, as evidenced by the inverted finite fault slip distribution and the calculated Coulomb stress change after each event. The foreshocks of the first M4 earthquake are probably triggered by Coulomb stress changes from previous earthquakes during 2014 and 2015 on unmapped faults several kilometers to the south. Fluid diffusion likely drives the bilateral seismic migration of the Cushing earthquake sequence after the foreshocks were triggered. In addition, fluid injection from the northwest of Cushing fault might have gradually increased the pore pressure on the Cushing fault, making the shallow part of the fault critically stressed.

Repeating earthquakes have been found at many faults around the world, and they provide valuable information on diverse faulting behavior at seismogenic depth. The Haiyuan fault is a major left-lateral strike-slip fault along the northeastern (NE) boundary of the Tibetan Plateau. Two great earthquakes (1920 Haiyuan, 1927 Gulang) have occurred on this fault system, but the section between the ruptures of the two earthquakes, also known as the Tianzhu seismic gap, remains unbroken. Shallow creep has been observed from geodetic data at the eastern end of the seismic gap. However, the driving mechanism and depth extent of shallow creep are not clear. Here we conduct a systematic search for repeating earthquakes in NE Tibet based on seismic data recorded by permanent stations in ten years (2009-2018). Based on waveform cross-correlations and subsequent relocations, we find several repeating earthquake clusters at Laohushan section. This is consistent with the shallow creep inferred from the geodetic data, indicating repeating earthquakes can be driven by nearby aseismic slip. ~300 repeaters were found within clusters of intense seismicity near the rupture zones of the 1927 M8.0 Gulang and 2016 M6.4 Menyuan earthquakes. Relocation of events in the cluster near the Gulang earthquake delineates two possible unmapped faults orthogonal to the Haiyuan fault. In addition, we also identify several repeating earthquakes generated by mining activities with different waveforms and occurrence patterns. Our study suggests that repeating earthquakes around the Haiyuan fault are mostly driven by postseismic relaxation process associated with 1920 Haiyuan and 1927 Gulang earthquakes.

We are engaging citizen scientists in an experiment to test if many human ears can replace the process of a professional seismologist in identifying dynamically triggered seismic events. Ordinarily, this process involves interactive data processing and visualization efforts on a volume of earthquake recordings (seismograms) that exploded during the recent big-data revolution, for example through EarthScope. In this citizen seismology project, we ask citizens to listen to relevant sections of seismograms that are accelerated to audible frequencies. This approach has five advantages: 1) The human ear implicitly performs a time-frequency analysis and is capable of discerning a wide range of different signals, 2) Many human ears listening to the same data provides statistics that rank seismograms in order of their likelihood to contain a recording of a triggered event, which is helpful to researchers’ analysis of this data as well as to 3) the ability of a deep-learning algorithm to model the boolean identifications or bulk statistics of the analyses, 4) the project has the potential to enhance informal learning because of the online platform that hosts the project, Zooniverse, is available to people of all identities and hosts many other citizen science projects, and 5) it helps prepare our team for diverse post-graduation careers as part of IDEAS, an NRT program at Northwestern University. The events we are asking citizens to help identify via listening are small seismic events such as local earthquakes and tectonic tremor, that occur in response to transient stresses from passing seismic surface waves from a large, distant earthquake. While much research progress has been made in understanding how these events are triggered, there is no reliable deterministic recipe for their occurrence. The aim of our project is to enlist the help of citizens to increase the data set of known triggered seismic events and known absences of triggered events in order to help researchers unravel key aspects of that recipe. A better understanding of triggered seismic events is expected to provide important clues towards a fundamental understanding of all seismic activity, including damaging earthquakes.

Machine learning algorithms have become a powerful tool in different areas of seismology, such as phase picking/earthquake detection, earthquake early warning and focal mechanism determination. Previously convolutional neural networks (CNN) have been applied to continuous seismic waveform recordings to perform efficient phase picking and event detection with good accuracy [Zhu et al., 2018]. However, the off-line training of current CNN requires at least a few thousands of accurately picked seismic phases, which makes it difficult to be applied to regions without sufficient picked phases. In this work, we will validate the transfer learning among different geographic regions. Our tests show that the phase picker trained on manually-labeled data acquired from Sichuan, China following the 2008 M7.9 Wenchuan earthquake [Zhu et al., 2018] works equally well on the continuous waveform acquired from Oklahoma, US [Zhu et al., 2018]. Specifically, using the CNN trained on the Wenchuan dataset, together with 895 local/regional catalog events recorded in central Oklahoma, we refine part of the networks to pick the arrival times of the local seismicity in Oklahoma. The refined CNN results are compatible with the matched filter results using the same catalog events as templates. Our next step is to extend our test to waveforms from different tectonic regions to demonstrate the generality of CNN-based phase picker. We plan to further use a New Zealand seismic dataset that includes more than 20 GeoNet stations in the North Island, where the matched-filter detected results are available to be compared with (Yao et al., 2018). Alternatively dataset include a subset of events in the waveform relocated catalog in Southern California. Updated results will be presented at the meeting.

Using P-wave seismograms by Barama et al. (2022), we trained a seismic source classifier using a Convolutional Neural Network. We trained for three classes: earthquake P-wave, nuclear P-wave, and noise. Seismograms with low signal to noise ratios (SNR) adversely affect the model performance, thus a threshold was applied, limiting the training set size. Our method can accurately characterize most events, finding over 95\% signals in the validation set, even with the SNR-limited training data. We applied the model on holdout datasets of the North Korean test blasts to evaluate the performance on unique region and station-source pairs. Additionally, we tested on the Source Physics Experiment events to investigate the potential for chemical blasts to act as a surrogate for nuclear blasts. We anticipate that machine-learning models like our classifier system can have broad application for other seismic signals including volcanic and non-volcanic tremor, anomalous earthquakes, ice-quakes or landslide-quakes.