Search for: All records

Abstract Volcanic tremor is a semi‐continuous seismic and/or acoustic signal that occurs at time scales ranging from seconds to years, with variable amplitudes and spectral features. Tremor sources have often been related to fluid movement and degassing processes, and are recognized as a potential geophysical precursor and co‐eruptive geophysical signal. Eruption forecasting and monitoring efforts need a fast, robust method to automatically detect, characterize, and catalog volcanic tremor. Here we develop VOlcano Infrasound and Seismic Spectrogram Network (VOISS‐Net), a pair of convolutional neural networks (one for seismic, one for acoustic) that can detect tremor in near real‐time and classify it according to its spectral signature. Specifically, we construct an extensive data set of labeled seismic and low‐frequency acoustic (infrasound) spectrograms from the 2021–2022 eruption of Pavlof Volcano, Alaska, and use it to train VOISS‐Net to differentiate between different tremor types, explosions, earthquakes and noise. We use VOISS‐Net to classify continuous data from past Pavlof Volcano eruptions (2007, 2013, 2014, 2016, and 2021–2022). VOISS‐Net achieves an 81.2% and 90.0% accuracy on the seismic and infrasound test sets respectively, and successfully characterizes tremor sequences for each eruption. By comparing the derived seismoacoustic timelines of each eruption with the corresponding eruption chronologies compiled by the Alaska Volcano Observatory, our model identifies changes in tremor regimes that coincide with observed volcanic activity. VOISS‐Net can aid tremor‐related monitoring and research by making consistent tremor catalogs more accessible. 

Abstract Knowledge of aquifer dynamics, including groundwater storage changes, is key to effective groundwater resource and reservoir management. Resolving and accurate modeling of these processes requires knowledge of subsurface poroelastic properties and lateral heterogeneity within units of interest. Computationally demanding methods for determining lateral heterogeneity in poroelastic properties exist but remain difficult to practically employ. The InSAR-based detection of uplift over a New Mexico well with a casing breach provides an opportunity to determine poroelastic properties using a tractable 2D analytical plane strain solution for surface uplift created by a pressurized reservoir with overburden. Using a Bayesian inversion framework, we calculate poroelastic properties under deep (depth of well-screen) and shallow (depth of well-breach) conditions. We find that shallow injection is necessary to produce the observed deformation. However, pressure-varying forward solutions for uplift are required to reproduce the temporal evolution of deformation. For this we use realistic shallow poroelastic properties and well dynamics, which reflect the evolving injection conditions at the well breach as the casing further erodes. Analysis of individual interferograms or InSAR time series may provide insights into shallow subsurface heterogeneity or anomalous injection conditions at operating wells more rapidly than scheduled field inspections. 

Abstract We developed an open source, extensible Python‐based framework, that we call the Versatile Modeling Of Deformation (VMOD), for forward and inverse modeling of crustal deformation sources. VMOD abstracts from specific source model implementations, data types and inversion methods. We implement the most common geodetic source models which can be combined to model and analyze multi‐source deformation. VMOD supports Global Navigation Satellite System (GNSS), InSAR, electronic distance measurement, Leveling and tilt data. To infer source characteristics from observations, VMOD implements non‐linear least squares and Markov Chain Monte‐Carlo Bayesian inversions, including joint inversions using different sources of data. VMOD's structure allows for easy integration of new geodetic models, data types, and inversion strategies. We benchmark the forward models against other published results and the inversion approaches against other implementations. We apply VMOD to analyze deformation at Unimak Island, Alaska, observed with continuous and campaign GNSS, and ascending and descending InSAR time series generated from Sentinel‐1 satellite radar acquisitions. These data show an inflation pattern at Westdahl volcano and subsidence at Fisher Caldera. We use VMOD to test a range of source models by jointly inverting the GNSS and InSAR data sets. Our final model simultaneously constrains the parameters of two sources. Our results reveal a depressurizing spheroid under Fisher Caldera ∼4–6 km deep, contracting at a rate of ∼2–3 Mm³/yr, and a pressurizing spherical source underneath Westdahl volcano ∼6–8 km deep, inflating at ∼5 Mm³/yr. This and past applications of VMOD to volcanic unrest benefit from an extensible framework which supports jointly inversions of data sets for parameters of easily composable multi‐source models. 

Abstract In April 2022, a seismic swarm near Mt. Edgecumbe in southeast Alaska suggested renewed activity at this transform fault volcano, which was last active ≈800 years ago. Previously, thin rhyolitic tephras were deposited 5 and 4 ka. Satellite radar data from 2014 to 2022 resolves line‐of‐sight rapid inflation up to 7.1 cm/yr beginning in August 2018. Bayesian modeling suggests a transcrustal system of a deflating (−0.528 km³) dipping sill at 20 km depth recharging a magma chamber at 10 km (0.222 km³). A near‐vertical conduit could capture the volume difference without noticeable surface deformation. Reanalyzed seismicity, recorded 25 km away, shows increases since July 2019. Magma ascent through ductile material and brittle strain release in a stressed overburden could explain the time delay. Cloud‐native open data and workflows enabled discovery and analysis of this signal within days after going unnoticed for >3 years. 

Abstract Morphological processes often induce meter‐scale elevation changes. When a volcano erupts, tracking such processes provides insights into the style and evolution of eruptive activity and related hazards. Compared to optical remote‐sensing products, synthetic aperture radar (SAR) observes surface change during inclement weather and at night. Differential SAR interferometry estimates phase change between SAR acquisitions and is commonly applied to quantify deformation. However, large deformation or other coherence loss can limit its use. We develop a new approach applicable when repeated digital elevation models (DEMs) cannot be otherwise retrieved. Assuming an isotropic radar cross‐section, we estimate meter‐scale vertical morphological change directly from SAR amplitude images via an optimization method that utilizes a high‐quality DEM. We verify our implementation through simulation of a collapse feature that we modulate onto topography. We simulate radar effects and recover the simulated collapse. To validate our method, we estimate elevation changes from TerraSAR‐X stripmap images for the 2011–2012 eruption of Mount Cleveland. Our results reproduce those from two previous studies; one that used the same dataset, and another based on thermal satellite data. By applying this method to the 2019–2020 eruption of Shishaldin Volcano, Alaska, we generate elevation change time series from dozens of co‐registered TerraSAR‐X high‐resolution spotlight images. Our results quantify previously unresolved cone growth in November 2019, collapses associated with explosions in December–January, and further changes in crater elevations into spring 2020. This method can be used to track meter‐scale morphology changes for ongoing eruptions with low latency as SAR imagery becomes available. 

An increase in volcanic thermal emissions can indicate subsurface and surface processes that precede, or coincide with, volcanic eruptions. Space-borne infrared sensors can detect hotspots—defined here as localized volcanic thermal emissions—in near-real-time. However, automatic hotspot detection systems are needed to efficiently analyze the large quantities of data produced. While hotspots have been automatically detected for over 20 years with simple thresholding algorithms, new computer vision technologies, such as convolutional neural networks (CNNs), can enable improved detection capabilities. Here we introduce HotLINK: the Hotspot Learning and Identification Network, a CNN trained to detect hotspots with a dataset of −3,800 satellite-based, Visible Infrared Imaging Radiometer Suite (VIIRS) images from Mount Veniaminof and Mount Cleveland volcanoes, Alaska. We find that our model achieves an accuracy of 96% (F1-score 0.92) when evaluated on −1,700 unseen images from the same volcanoes, and 95% (F1-score 0.67) when evaluated on −3,000 images from six additional Alaska volcanoes (Augustine Volcano, Bogoslof Island, Okmok Caldera, Pavlof Volcano, Redoubt Volcano, Shishaldin Volcano). In comparison with an existing threshold-based hotspot detection algorithm, MIROVA (Coppola et al., Geological Society, London, Special Publications, 2016, 426, 181–205), our model detects 22% more hotspots and produces 12% fewer false positives. Additional testing on −700 labeled Moderate Resolution Imaging Spectroradiometer (MODIS) images from Mount Veniaminof demonstrates that our model is applicable to this sensor’s data as well, achieving an accuracy of 98% (F1-score 0.95). We apply HotLINK to 10 years of VIIRS data and 22 years of MODIS data for the eight aforementioned Alaska volcanoes and calculate the radiative power of detected hotspots. From these time series we find that HotLINK accurately characterizes background and eruptive periods, similar to MIROVA, but also detects more subtle warming signals, potentially related to volcanic unrest. We identify three advantages to our model over its predecessors: 1) the ability to detect more subtle volcanic hotspots and produce fewer false positives, especially in daytime images; 2) probabilistic predictions provide a measure of detection confidence; and 3) its transferability, i.e., the successful application to multiple sensors and multiple volcanoes without the need for threshold tuning, suggesting the potential for global application. 

Volcanic earthquake catalogs are an essential data product used to interpret subsurface volcanic activity and forecast eruptions. Advances in detection techniques (e.g., matched-filtering, machine learning) and relative relocation tools have improved catalog completeness and refined event locations. However, most volcano observatories have yet to incorporate these techniques into their catalog-building workflows. This is due in part to complexities in operationalizing, automating, and calibrating these techniques in a satisfactory way for disparate volcano networks and their varied seismicity. In an effort to streamline the integration of catalog-enhancing tools at the Alaska Volcano Observatory (AVO), we have integrated four popular open-source tools: REDPy, EQcorrscan, HypoDD, and GrowClust. The combination of these tools offers the capability of adding seismic event detections and relocating events in a single workflow. The workflow relies on a combination of standard triggering and cross-correlation clustering (REDPy) to consolidate representative templates used in matched-filtering (EQcorrscan). The templates and their detections are then relocated using the differential time methods provided by HypoDD and/or GrowClust. Our workflow also provides codes to incorporate campaign data at appropriate junctures, and calculate magnitude and frequency index for valid events. We apply this workflow to three datasets: the 2012–2013 seismic swarm sequence at Mammoth Mountain (California), the 2009 eruption of Redoubt Volcano (Alaska), and the 2006 eruption of Augustine Volcano (Alaska); and compare our results with previous studies at each volcano. In general, our workflow provides a significant increase in the number of events and improved locations, and we relate the event clusters and temporal progressions to relevant volcanic activity. We also discuss workflow implementation best practices, particularly in applying these tools to sparse volcano seismic networks. We envision that our workflow and the datasets presented here will be useful for detailed volcano analyses in monitoring and research efforts.