skip to main content

Title: Automatic Detection of Volcanic Surface Deformation Using Deep Learning

Interferometric Synthetic Aperture Radar (InSAR) provides subcentimetric measurements of surface displacements, which are key for characterizing and monitoring magmatic processes in volcanic regions. The abundant measurements of surface displacements in multitemporal InSAR data routinely acquired by SAR satellites can facilitate near real‐time volcano monitoring on a global basis. However, the presence of atmospheric signals in interferograms complicates the interpretation of those InSAR measurements, which can even lead to a misinterpretation of InSAR signals and volcanic unrest. Given the vast quantities of SAR data available, an automatic InSAR data processing and denoising approach is required to separate volcanic signals that are cause of concern from atmospheric signals and noise. In this study, we employ a deep learning strategy that directly removes atmospheric and other noise signals from time‐consecutive unwrapped surface displacements obtained through an InSAR time series approach using an end‐to‐end convolutional neural network (CNN) with an encoder‐decoder architecture, modified U‐net. The CNN is trained with simulated synthetic unwrapped surface displacement maps and is then applied to real InSAR data. Our proposed architecture is capable of detecting dynamic spatio‐temporal patterns of volcanic surface displacements. We find that an ensemble‐average strategy is recommended to stabilize detected results for varying deformation rates and signal‐to‐noise ratios (SNRs). A case study is also presented where this method is applied to InSAR data covering Masaya volcano, Nicaragua and the results are validated using continuous GPS data. The results confirm that our network can indeed efficiently suppress atmospheric and other noise to reveal the noise‐free surface deformation.

more » « less
Award ID(s):
Author(s) / Creator(s):
; ; ; ; ; ;
Publisher / Repository:
Date Published:
Journal Name:
Journal of Geophysical Research: Solid Earth
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    We use Sentinel‐1 and ALOS‐2 Interferometric Synthetic Aperture Radar (InSAR), and Global Navigation Satellite System (GNSS) data to investigate the mechanisms of coseismic and postseismic deformation due to the 2021 M7.4 Maduo (China) earthquake. We present a refined coseismic slip model constrained by the rupture trace and precisely located aftershocks. The InSAR time series corrected for the atmospheric and decorrelation noise reveal postseismic line of sight displacements up to ∼0.1 m. The displacements are discontinuous along the fault trace, indicating shallow afterslip and velocity‐strengthening friction in the top 2–3 km of the upper crust. The magnitude of shallow afterslip is however insufficient to compensate for the coseismic slip deficit, implying substantial off‐fault yielding. The observed surface deformation does not exhibit obvious features that could be attributed to poroelastic effects. We developed a fully coupled model that accounts for both stress‐driven creep on a deep localized shear zone and viscoelastic relaxation in the bulk of the lower crust. The mid‐ to near‐field data can be reasonably well explained by deep afterslip and/or non‐Maxwellian visco‐elasticity. Our results suggest a power‐law stress exponent of ∼4–4.5 assuming a power‐law rheology, and transient and steady‐state viscosities of 1018and 1019 Pa s, respectively, assuming a bi‐viscous (Burgers) rheology. However, a good fit to the GNSS data cannot be achieved assuming the bulk viscoelastic relaxation alone, and requires a contribution of deep afterlip and/or a localized shear zone extending through much of the lower crust.

    more » « less
  2. Artificial intelligence applications within the geosciences are becoming increasingly common, yet there are still many challenges involved in adapting established techniques to geoscience data sets. Applications in the realm of volcanic hazards assessment show great promise for addressing such challenges. Here, we describe a Jupyter Notebook we developed that ingests real-time Global Navigation Satellite System (GNSS) data streams from the EarthCube CHORDS (Cloud-Hosted Real-time Data Services for the geosciences) portal TZVOLCANO, applies unsupervised learning algorithms to perform automated data quality control (“noise reduction”), and explores autonomous detection of unusual volcanic activity using a neural network. The TZVOLCANO CHORDS portal streams real-time GNSS positioning data in 1[Formula: see text]s intervals from the TZVOLCANO network, which monitors the active volcano Ol Doinyo Lengai in Tanzania, through UNAVCO’s real-time GNSS data services. UNAVCO’s real-time data services provide near-real-time positions processed by the Trimble Pivot system. The positioning data (latitude, longitude and height) are imported into the Jupyter Notebook presented in this paper in user-defined time spans. The positioning data are then collected in sets by the Jupyter Notebook and processed to extract a useful calculated variable in preparation for the machine learning algorithms, of which we choose the vector magnitude for further processing. Unsupervised K-means and Gaussian Mixture machine learning algorithms are then utilized to locate and remove data points (“filter”) that are likely caused by noise and unrelated to volcanic signals. We find that both the K-means and Gaussian Mixture machine learning algorithms perform well at identifying regions of high noise within tested GNSS data sets. The filtered data are then used to train an artificial intelligence neural network that predicts volcanic deformation. Our Jupyter Notebook has promise to be used for detecting potentially hazardous volcanic activity in the form of rapid vertical or horizontal displacement of the Earth’s surface. 
    more » « less
  3. null (Ed.)
    Ensemble based data assimilation approaches, such as the Ensemble Kalman Filter (EnKF), have been widely and successfully implemented to combine observations with dynamic forecast models. In this study the EnKF is adapted to assimilate ground deformation observations from interferometric synthetic-aperture radar (InSAR) and GPS into thermomechanical finite element models (FEM) to evaluate volcanic unrest. Two eruption hindcasts are investigated: the 2008 eruption of Okmok volcano, Alaska and the 2018 eruption of Sierra Negra volcano, Galápagos, Ecuador. At Okmok, EnKF forecasts tensile failure and the lateral movement of the magma from a central pressure source in the lead up to its 2008 eruption indicating potential for diking. Alternatively, at Sierra Negra, the EnKF forecasts significant shear failure coincident with a Mw 5.4 earthquake that preceded the 2018 eruption. These successful hindcasts highlight the flexibility and potential of the volcano EnKF approach for near real time monitoring and hazard assessment at active volcanoes worldwide. 
    more » « less
  4. With approximately 800 million people globally living within 100 km of a volcano, it is essential that we build a reliable observation system capable of delivering early warnings to potentially impacted nearby populations. Global Navigation Satellite System (GNSS) and satellite Synthetic Aperture Radar (SAR) document comprehensive ground motions or ruptures near, and at, the Earth’s surface and may be used to detect and analyze natural hazard phenomena. These datasets may also be combined to improve the accuracy of deformation results. Here, we prepare a differential interferometric SAR (DInSAR) time series and integrate it with GNSS data to create a fused dataset with enhanced accuracy of 3D ground motions over Hawaii island from November 2015 to April 2021. We present a comparison of the raw datasets against the fused time series and give a detailed account of observed ground deformation leading to the May 2018 and December 2020 volcanic eruptions. Our results provide important new estimates of the spatial and temporal dynamics of the 2018 Kilauea volcanic eruption. The methodology presented here can be easily repeated over any region of interest where an SAR scene overlaps with GNSS data. The results will contribute to diverse geophysical studies, including but not limited to the classification of precursory movements leading to major eruptions and the advancement of early warning systems. 
    more » « less
  5. SUMMARY InSAR displacement time-series are emerging as a valuable product to study a number of Earth processes. One challenge to current time-series processing methods, however, is that when large earthquakes occur, they can leave sharp coseismic steps in the time-series. These discontinuities can cause current atmospheric correction and noise smoothing algorithms to break down, as these algorithms commonly assume that deformation is steady through time. Here, we aim to remedy this by exploring two methods for correcting earthquake offsets in InSAR time-series: a simple difference offset estimate (SDOE) process and a multiparameter offset estimate (MPOE) parametric time-series inversion technique. We apply these methods to a 2-yr time-series of Sentinel-1 interferograms spanning the 2019 Ridgecrest, CA earthquake sequence. Descending track results indicate that the SDOE method precisely corrects for only 20 per cent of the coseismic offsets at 62 study locations included in our scene and only partially corrects or sometimes overcorrects for the rest of our study sites. On the other hand, the MPOE estimate method successfully corrects the coseismic offset for the majority of sites in our analysis. This MPOE method allows us to produce InSAR time-series and data-derived estimates of deformation during each phase of the earthquake cycle. In order to better isolate and estimate the signal of post-seismic lithospheric deformation in the InSAR time-series, we apply a GNSS-based correction to our interferograms. This correction ties the interferograms to median-filtered weekly GNSS displacements and removes additional atmospheric artefacts. We present InSAR-based estimates of post-seismic deformation for the area around the Ridgecrest rupture, as well as a 2-yr coseismic-corrected, GNSS-corrected InSAR time-series data set. This GNSS-corrected InSAR time-series will enable future modelling of post-seismic processes such as afterslip in the near field of the rupture, poroelastic deformation at intermediate distances and viscoelastic deformation at longer timescales in the far field. 
    more » « less