Magmatic and tectonic processes at Kilauea Volcano

In 2018, Kilauea Volcano, Hawaii underwent a series of major events including more than a 500-meter collapse of the summit, lava eruptions, and a magnitude 6.9 earthquake and its numerous aftershocks. The consequences of these events were extensive damage to infrastructure and property. In early July 2018, a team funded by the NSF Grants for Rapid Response Research (RAPID) program deployed 12 ocean bottom seismometers (OBSs) offshore Kilauea. Eleven of the twelve OBSs were recovered in mid-September. All but one of the recovered OBSs had useful seismic and hydroacoustic data for the full recording period. During the two-month OBS recording periods, around 16,000 magnitude 1 and larger earthquakes were located by the U.S. Geological Survey (USGS) Hawaiian Volcano Observatory (HVO), many occurring offshore or near the coast. Because all HVO seismic stations are on land, many offshore earthquakes were likely missed or poorly located in the HVO catalog. Preliminary data analysis shows that the OBS deployment successfully captured a significant portion of abundant aftershock activity and spanned the termination of active lava eruption. It was the first time that such events are observed by concurrent onshore-offshore seismic arrays and these events provided an extraordinary opportunity to observe the interplay between magmatic and tectonic processes in exceptionally rich detail and dimension.

In this NSF-sponsored project, we will analyze the data collected by the offshore OBSs to gain new insights into the magmatic and tectonic processes at Kilauea Volcano, focusing on four main scientific questions: (1) What structures within Kilauea's south flank were activated during the 2018 events, and what deformation did they accommodate? (2) What are the physical properties of the main faults? (3) How does deep magma migrate in space and time? And (4) What are the stress states of the submarine south flank? These are long-standing questions about the magmatic and tectonic processes at Kilauea Volcano and oceanic shield volcanoes in general. The answers will provide constraints on flank deformation and deep magma migration and will have implications for seismic hazards, submarine landslides, and tsunami risks. The data analyses will include detecting and locating earthquakes and determining the physical properties of major faults and stress states of the Kilauea south flank. The expected results include a catalog of offshore earthquakes and refined offshore velocity models, which will enable the development, with support by the USGS, of new offshore seismicity monitoring capabilities with the potential to enhance societal preparedness and resilience against such impacts. The project will be accomplished in collaboration with scientists at the Hawaiian Volcano Observatory, Western Washington University, and Rice University.

Project Website (access limited to the project participants and collaborators)

Galapagos mantle plume - ridge interaction

There are three fundamental ways in which volcanoes form on Earth. This NSF-sponsored study will examine how two of these styles behave in proximity. The Galapagos archipelago is a well-known example of hotspot volcanism, fed by a rising plume of hot mantle, and this system is interacting with the Galapagos Spreading Center (GSC) where volcanism occurs in response to seafloor spreading at the boundary between two tectonic plates moving apart. Scientists have studied the Galapagos hotspot-GSC system for several decades and have long puzzled over persistent discrepancies between geophysical and geochemical observations and physical models for how the pair work together. This study will image how mass and temperature are transported, as well as how magma is generated beneath the Galapagos system, using a technique called seismic tomography. A network of instruments spanning a large area of the islands and adjacent seafloor will be deployed. These seismometers will record seismic waves traveling from distant earthquakes and ambient ground displacement over a period of 15 months. As the seismic waves pass beneath the study area, they respond to differences in mantle composition, temperature, deformation, and the presence of magma. Imaging these properties will allow many unanswered questions particular to the Galapagos system to be addressed. The study will also address the fundamental processes occurring in the shallow part of the mantle that is hot and weak and the interactions with the the cool, stiff overlying part that forms Earth's tectonic plates. This program will train three graduate students in marine geophysics. In addition, an Apply-to-Sail program will allow graduate students and early career scientists from other institutions and community college instructors to participate on the research cruises to gain sea-going training. Lastly, Ecuadorian scientists and graduate students will also participate on the cruises, bolstering science education and international research collaboration.

To produce the first mantle seismic view of how mantle plume-ridge interaction really works, an open-access seismic dataset will be collected around the Galapagos system. The experiment and subsequent analyses are designed to address three main scientific questions: (i) At what depths, in what geographic pattern, and by what mechanism does mantle plume material flow northward to the Galapagos Spreading Center and disperse along the ridge? (ii) Do the scale and nature of heterogeneity indicate small-scale, sub-lithospheric convection? and (iii) What is the spatial distribution of melting and volatile release, as well as the associated heterogeneity in composition and rheology due to plume-ridge interaction? The Galapagos system is exceptionally well-suited for such a study given the history of previous investigations of the surface manifestations, the evidence from mantle tomography below the Galapagos Archipelago, and the favorable azimuthal distribution of seismic sources. A large number (53) ocean-bottom, broadband seismometers will be deployed for 15 months in an array spanning the area between the Galapagos Islands and the Western Galapagos Spreading Center. Data from 7 broadband stations on the islands also will be used. The data will undergo initial processing, including ambient noise cross-correlation, and be archived in the IRIS-DMC for immediate public use. Tomography models of isotropic velocity will be produced from body waves, isotropic velocity from the combination of surface waves and ambient noise, as well as radial and azimuthal anisotropy from surface waves. Receiver functions will be analyzed to identify discontinuities related to the lithosphere and melting and shear wave splitting will be used to map anisotropy. Geodynamic models of plume-ridge interaction will be used for hypothesis testing by comparing modeled and observed seismic waveforms, and by using the geodynamic models as a priori information for the tomographic inversions. The project will also substantially advance a broad understanding of mantle plume processes, the asthenosphere, and their interactions with oceanic lithosphere; specifically, the deployment will function as a unit array within the Pacific Array Initiative.

This is a collaborative project with Garrett Ito of University of Hawaii (geodynamics), Emilie Hooft and Doug Toomey of University of Oregon (seismology), and Mario Ruiz of IG-EPN, Ecuado (seismology).

Machine Learning

Seafloor Geodesy Shallow slow slip events provide a mechanism for strain release at the shallow part of subduction zones, which is important for tsunami hazard assessment. For most subduction zones, the trench is far from the coast and it is unclear whether shallow slow slip events exist. Even in places where these events were detected, key quantities such as the duration and magnitude were not well constrained. As a result, the locking state of shallow subduction zones and the mechanism of shallow slow slip events is still unclear. To answer these questions, this NSF project will take advantage of recent advancement in machine learning and the accumulation of seafloor pressure datasets to improve our ability to detect shallow slow slip events in subduction zones. Preliminary analyses of seafloor pressure data from New Zealand have demonstrated that machine learning can successfully identify known slow slip events and further reduce ocean noise in seafloor pressure data. Using available data from several subduction zones, this project will further improve the machine-learning detector to estimate the duration, amplitude, and timing of shallow slow slip events. This project will also develop an improved way to reduce ocean noise in seafloor pressure data by using machine learning to capture the complex relationship of measurable quantities in the ocean. Collectively, this project will provide better tools to measure shallow slow slip events and assess the locking state of shallow subduction zones.

This is a collaborative project with Matt Wei (geodesy) and Marco Alvarez (computer science).

Earthquake Detection and Localization Motivated by the need to detect and locate thousands of earthquakes in Hawaii (see Kilauea Volcano), we developed a new deep learning method for automatic detection and 4D localization of earthquakes. Current methods are computationally expensive, ineffective under noisy environments, or labor-intensive. We leverage advances in machine learning to propose an improved solution - a convolutional neural network that uses continuous array data from a seismic network to seamlessly detect and localize events, without the intermediate steps of phase detection, association, travel time calculation, and inversion. Application to continuous records shows that our algorithm detects 690% as many earthquakes as the published catalog, 161% as many events than the unpublished catalog including greater than magnitude 0.1, and 125% as many events than the unpublished catalog with all earthquakes. Due to the enhanced detection sensitivity, localization granularity, and minimal computation costs, our solution is valuable, particularly for real-time earthquake monitoring.

Reverse-Time Migration Imaging

Reverse-time Migration imaging of the Upper Mantle Structures Beneath Southwest Japan (project website, access limited to the project participants)