The questions and priorities highlighted in this report are complex and multifaceted. They require perspectives on volcanism that span scales from individual crystals to entire volcanic arcs. Advances rely on instrumentation as varied as laser ablation mass spectrometers, broadband seismometers, and satellite sensors. New technology promises to provide critical insights on previously inaccessible parts of volcanoes and on eruptions.
Making new discoveries and improving understanding depend on the ability to undertake interdisciplinary research and provide interdisciplinary training. Shared infrastructure, resources, and data would accelerate the pace of progress in developing models and making critical measurements. Coordinated responses to eruptions globally would help overcome observational biases. Effective collaborations and partnerships among academia, volcano observatories, and government agencies would maximize the scientific return from monitoring data and improve eruption forecasts.
Volcano science is interdisciplinary. Addressing fundamental questions requires integrating diverse types of observations from geophysics, geology, geochemistry, geodynamics, and remote sensing. Efforts to improve forecasting involve research in statistics. And modern models for volcanic eruptions involve high-performance computing and collaboration with engineering science and applied mathematics. Research and understanding in each of these areas has advanced to the point that few individuals have expertise spanning more than one of these disciplines. Communication of research results is well supported through dedicated scientific journals, societies, and conferences. Supporting volcano scientists in collaborative interdisciplinary research is more challenging, however, and requires funding for cross-disciplinary research projects and for fostering sustained and substantive discussion and collaborations across disciplines.
Current core science funding programs at the National Science Foundation (NSF) are broken down into historical subdisciplines,1 making it challenging to support multi-investigator interdisciplinary research. A few funding programs at NSF have supported interdisciplinary research involving volcanoes, including the hazards aspects of eruptions (previously the hazard elements of SEES [Science, Engineering, and Education for Sustainability], currently PREVENTS [Prediction of and Resilience against Extreme Events])
1 Core science programs in NSF’s Division of Earth Sciences are EarthScope, Geobiology and Low-Temperature Geochemistry, Geomorphology and Land Use Dynamics, Geophysics, Hydrologic Sciences, Petrology and Geochemistry, Sedimentary Geology and Paleobiology, and Tectonics. See https://www.nsf.gov/funding/programs.jsp?org=EAR.
and collaborations with mathematics (previously through Collaboration in Mathematical Geosciences [CMG]). However, these programs were not aimed at advancing our understanding of the processes that govern the storage, ascent, and eruption of magma. A successful model for interdisciplinary research is NSF’s CSEDI (Cooperative Studies of the Earth’s Deep Interior) program, which supports collaboration between geochemistry, geodynamics, mineral physics, geomagnetism, and seismology (similar fields to those in volcano science) to understand the evolution and dynamics of Earth’s deep interior.
True collaboration between disciplines requires support for sustained exchange of ideas, challenges, and opportunities, beyond simply funding collaborative projects. Successful models for multidisciplinary problem solving often involve thematic meetings, centered around grand challenges. For examzple, the Southern California Earthquake Center is a collaboratory of geologists, seismologists, geodesists, modelers, and experimentalists who work together on specific goals defined each year, for understanding earthquake processes. GeoPRISMs (Geodynamic Processes at Rifting and Subducting Margins), an NSF multidisciplinary program to study continental margins, convenes Theoretical and Experiments Institutes that attack frontier problems. Gordon Conferences, research coordination networks, and summer institutes provide other avenues for interdisciplinary collaboration, with the added benefit of training for early career scientists. The payoffs include discoveries that would not otherwise emerge from a single perspective, new insights into complex processes, integration of data and models, and a community of scientists well versed in multiple fields and engaged in solving critical problems.
The volcano science community currently relies on a suite of analytical, computational, and experimental facilities. Community and multiuser facilities, in particular, provide opportunities and expertise to a broad range of users. Useful infrastructure improvements for volcano science range from analytical facilities to cyberinfrastructure, from satellites to long-lived field experiments. Infrastructure developed for complementary large Earth science projects (e.g., EarthScope and Subduction Zone Observatory) can also be leveraged.
Intrinsic properties of the magmatic systems that fuel volcanoes are measured using the tools of geochronology, geochemistry, rock physics, and petrology. Many facilities are hosted by single institutions, making access highly variable. These include geochemical and microanalytical facilities, and high-pressure, high-temperature experimental petrology and rock physics laboratories. Facilities to support geochronology, in particular, are critical for constraining the life cycles of volcanoes. Geochronology facilities in the United States are inadequate to meet the demand for all disciplines and are often inaccessible because of high costs (Harrison et al., 2015). Despite the scientific value (see Chapter 2), there is currently little U.S. emphasis on drilling to access the subsurface of volcanic systems, either for basic science studies or for deploying borehole instruments such as seismometers and strainmeters. This is in marked contrast to past community projects in the United States (e.g., Eichelberger, 1997; Eichelberger et al., 1984; Keller et al., 1979; Zablocki et al., 1974) and current international efforts (e.g., Bonaccorso et al., 2016; Elders et al., 2014; Sakuma et al., 2008).
Experimental facilities to study the dynamics of volcanic phenomena such as pyroclastic density currents, lava flows, and plumes are hosted by individual researchers. The large scale of many experimental models, however, could benefit from development and support of community user facilities (Valentine et al., 2011). Such large-scale experiments can provide a test bed for exploring new physical processes, validating codes, and testing new instrumentation, including in situ monitoring of flows.
Department of Energy-supported synchrotron beamlines have fueled rapid advances in spectroscopic and single-crystal measurements and microtomographic studies of volcanic materials. New and exciting applications include four-dimensional imaging of multiphase magma transport processes (e.g., Baker et al., 2012).
Access to high-performance computing, such as XSEDE (Extreme Science and Engineering Discovery Environment), NCAR (National Center for Atmospheric Research), and NERSC (National Energy Research Scientific Computing Center) facilities, are essential for state-of-the-art models. Communal cyberinfrastructure supports comprehensive data-
bases, model development, benchmarking, and implementation (e.g., Marzocchi et al., 2008; Sparks et al., 2012). VHub2 currently serves as a clearinghouse for such models. Community facilities, such as NSF-funded CIG (Computational Infrastructure for Geodynamics), could help improve the accessibility and user-friendliness of computational approaches by enhancing code efficiency and offering resources not available to individuals.
Finally, geophysical, geochemical, and geodetic data underpin research on active volcanoes. Key avenues for data collection include satellites that can be used for targeted observations of restless volcanoes, airborne instruments (including drones), and instrument pools for volcano-specific monitoring equipment. The NSF-funded PASSCAL (Portable Array Seismic Studies of the Continental Lithosphere) center provides seismometers for targeted campaigns, and UNAVCO provides engineering support and Global Positioning System (GPS) units for campaign measurements and permanent installation. A broader range of instruments and enhanced community coordination would maximize rapid response capabilities and permit innovative multisensor experiments on individual volcanoes.
Improving our ability to understand and forecast volcanic behavior requires a workforce capable of communicating and integrating information across the different fields represented within volcano science. The next generation of volcano scientists must not only develop core expertise but also acquire sufficient knowledge to incorporate results from, and communicate with, volcano scientists in other disciplines. Of these skills, only the first is a common goal of traditional graduate programs. Although most graduate programs also encourage some breadth in scientific understanding, the extent varies widely. Skills for communicating across disciplines are only indirectly a part of the training of most scientists. Improved training therefore requires new ways to expand the communication proficiency of scientists, and to foster opportunities and mechanisms for developing interdisciplinary research skills while still maintaining disciplinary rigor. Similarly, quantitative skills are increasingly important, requiring training in computation and statistics.
Training interdisciplinary volcano scientists poses several challenges. First is the breadth of volcano science—few institutions can cover all aspects internally. An exception is the U.S. Geological Survey (USGS), which is only peripherally involved in graduate student training. Second is the sheer size of the United States, which means that research institutions specializing in volcano science are spread across the country (particularly Hawaii and Alaska). The physical separation means that casual exchanges and interactions among volcano scientists are less than optimal. More formal interactions take place at specialized meetings, but they are difficult to maintain over the long term. Finally, the funding structure tends to be conservative, typically encouraging discipline-specific projects (Section 5.1) and thus indirectly discouraging exploratory work between disciplines.
A variety of programs in Europe address these challenges in volcano science by enabling joint training in different disciplines and across different institutions.3 These programs come with their own sets of challenges: They are expensive and time consuming for all participants, and they can be impractical for those with limited geographic mobility. Training networks, however, have the advantage of training a cohort of PhD students who, during their studies, will have developed what will hopefully be a lifelong network. An alternative approach is to develop a postdoctoral program that requires training in a field that is outside of, but complementary to, the PhD specialty.
On shorter time scales are focused summer schools, such as the Geophysical Fluid Dynamics Program4 (run by the Woods Hole Oceanographic Institute, now past its 50th year) and the CIDER (Cooperative Institute for Dynamic Earth Research)5 summer program (funded by NSF, now past its 10th year). Also important are training schools, internships, and volunteer programs that provide students the experience of working in an observatory environment. Examples of these
3 Some examples include the European Research Commission’s Innovative Training Networks (e.g., VERTIGO, NEMOH) and the European Science Foundation (e.g., MEMOVOLC).
programs include the international training course run by the Center for the Study of Active Volcanoes and the volunteer program run by the Hawaiian Volcano Observatory, both of which provide hands-on experience for students interested in Hawaiian volcanism.
5.4 DEVELOPING THE NEXT GENERATION OF INSTRUMENTATION AND BROADENING APPLICATIONS OF INSTRUMENTATION TO VOLCANO SCIENCE
Volcano science is grounded in a rich history of empirical observation, and new ground-based, airborne, and satellite technology are allowing many kinds of observations to be acquired more rapidly and in more detail than ever before. Substantial advances in our understanding of internal and surface processes have been achieved through acquisition and interpretation of seismic, magnetotelluric, deformation, gas, hydrologic, and thermal data, from space and on the ground (see Chapters 2 and 3). Evolving technology is permitting more interplay between measurements. For instance, seismic study has been expanded to seismoacoustics, incorporating infrasound recorded by low-frequency microphones (e.g., Arrowsmith et al., 2010). Radar is providing new insights into processes in explosive eruptions. Seismic and deformation monitoring now document a continuum of Earth motions. Thermal cameras, previously used to quantify stationary heat flow, now provide high-temporal-resolution imagery to measure both vent velocities and lava effusion rates (e.g., Patrick et al., 2014). A new generation of multispectral imaging cameras are able to measure gas concentrations in volcanic plumes at high enough spatial and temporal resolution to enable direct comparisons to seismic signals (e.g., Nadeau et al., 2011). Accompanying the benefits of expanding data sets and instrument capabilities are challenges posed by data that are heterogeneous in both space and time, making comparisons between precursory signals difficult.
Future advances in volcano science will be facilitated by sensor improvement and the deployment of global multiparameter sensor networks to capture the full range of temporal and spatial variability of volcanic activity. Experiments at model volcanoes such as Stromboli, Italy, or analog volcanoes such as geysers (Hurwitz and Manga, 2017), offer opportunities to interpret a host of geophysical signals in terms of processes and physical properties. Further developments in spectroscopy will enable higher-precision measurements of the isotopic and chemical composition of volcanic gases remotely, in situ, and in near real time. Advances in technology will make instruments smaller, cheaper, and more robust. Higher precision and spatial resolution of laboratory-based beam analytical techniques will provide finer temporal resolution of processes recorded by crystals and melt inclusions.
Several planned or proposed satellite missions would benefit the volcano science community (Davis et al., 2016). Distinguishing volcanic CO2 from anthropogenic emissions remains challenging with current sensors, but deployment of satellite-based technology with greater vertical sensitivity to CO2 (e.g., active laser instruments such as NASA’s ASCENDS [Active Sensing of CO2 Emissions over Nights, Days, & Seasons]) could lead to more timely detection of eruption precursors. This would be particularly effective in combination with more frequent (daily) repeat interferometric synthetic aperture radar (InSAR) measurements of ground deformation from a constellation of satellites. Regular acquisition of global, high-quality digital elevation models such as the TanDEM-X (TerraSAR-X add-on for Digital Elevation Measurement) WorldDEM6 would facilitate studies of dynamic volcano topography and permit more accurate simulations of volcanic mass flows and improved hazard mapping.
The scale of advances in data acquisition are illustrated by changing capabilities of volcano seismology (Table 5.1); however, all of the technologies summarized in Tables 1.1 and 1.2 have undergone a similar evolution or could be developed further with adequate resources. Improved database capabilities (Section 5.5) and software are needed to complement sensor improvements and the increased volume and quality of data. Efficient archiving and extraction of time series, spectral, and image data are crucial to improve data visualization and discovery. Finally, drone technology promises to revolutionize the capabilities for data and sample collection by allowing access to inaccessible or dangerous areas or by offering previously unanticipated perspectives.
TABLE 5.1 Advances in Volcano Seismology
|Year||Data Collection Capability|
|1980||Seismic data at volcanoes are analog and short period (>1 Hz). Recordings are displayed on paper helicorder plots.|
|1990||All seismic data are recorded digitally, but not continuously. Digital tape archiving is costly and inefficient.|
|2000||Seismic data are recorded continuously. Broadband seismology is revolutionizing volcano earthquake study.|
|2010||Digital networks have largely replaced analog systems. Large-N deployments allow development of high-resolution volcano tomographic images.|
|2020||What’s next? Thousands of very broadband geodetic seismometers rapidly deployed by drones and delivering data in real time.|
Data from many of the instruments used to monitor and study volcanoes provide insights into other hazards, such as earthquakes, landslides, and forest fires. They also yield information on subsurface processes, such as the evolution and structure of the crust, the development of geothermal systems, and the formation of ore deposits. There are thus opportunities to leverage instrumentation and networks to address a range of resource, hazard, and science questions.
Open data access has revolutionized some disciplines in Earth science. For example, the easy availability of waveform data has allowed for new and innovative analysis in seismology. This in turn has led to the discovery of phenomena such as tectonic tremor, as well as insights on the structure of Earth’s deep interior. Similar arguments can be made for readily accessible continuous GPS data.
Databases are playing an increasingly important role in volcano science. A summary of existing volcano databases is given in Appendix A. Flexible databases allow comparisons of parameters and phenomena across many volcanoes and eruptions. For example, it would be useful to know how often phreatic explosions are followed by magmatic eruptions as well as the distribution of time intervals between these events. Data access in volcano science is inherently more challenging than for seismology, because the field involves such disparate data types, including the following:
- Historical information on past eruptions, including Volcanic Explosivity Index, eruption rate, erupted products and volume, duration of eruption, events during eruptions (e.g., explosions, pyroclastic density currents, and lahars), and stratigraphic and field relations, including deposit thickness and extent.
- Data on potential eruption precursors, such as
- seismicity (earthquake locations, magnitudes, moment tensors, and frequency content;tremor amplitude; real-time seismic-amplitude measurement);
- deformation, including “snapshots” (InSAR interferograms) and time series (GPS, leveling, tilt, strain, and sets of InSAR images);
- gas and fluid measurements (ground based and remotely sensed);
- thermal measurements (ground based and remotely sensed); and
- Rock samples, including composition, phase assemblages, textures, and melt inclusion volatiles, ideally tied to specific eruptions.
- Physical volcanology parameters, such as deposit thickness, mass, density, and grain size and shape.
- Imaging data, including data from cameras (e.g., photos, video, and time lapse images) and geophysical techniques (e.g., seismic tomography images).
- Potential field measurements, including gravity and magnetotellurics.
Once an eruption has begun, a variety of data is generated through ground-based and remote sensing techniques, including eruption column heights and ash and gas distribution; pyroclastic density current and lava flow paths and volumes; petrology, geochemistry, and fluxes of erupting products and emitted gases; deformation; and seismicity. Accurate and frequent, possibly real-time or near-real-time, ingestion into databases would allow scientists at observatories to improve eruption modeling and forecasts of how the eruption will proceed and when it may end.
Despite the many existing databases, key information is not currently included, such as compositional information for eruptions over time; field data (e.g., maps and videos); ash fall, pyroclastic density current, lava flow, and lahar inundation maps; textural and sieve data; and samples including location information.
There are many challenges to moving toward a comprehensive volcano science database, including establishing standards for relevant data and metadata, linking databases through web services, and making a long-term commitment to maintenance. Significant efforts were required to develop standards for seismic networks, and these apply mainly to large networks such as the Global Seismic Network. Data from volcanoes are likely to be much more heterogeneous. Moreover, much of the relevant data are collected by volcano observatory staff, who have little time to disseminate them during heightened activity or may be concerned about public misunderstanding and alarm. The USGS will be making all published data publicly available, although mechanisms to share data with volcano scientists in real time have not yet been developed.
University researchers, who put enormous effort into data collection, may also be reluctant to make data freely available before publishing their interpretations. Different disciplines have different standards for data sharing. For example, there is a 2-year moratorium before seismic data must be made publicly available through the IRIS (Incorporated Research Institutions for Seismology) Data Management Center. Digital Object Identifiers provide one mechanism for acknowledging researchers’ contribution to data collections.
Additional challenges are specific to the USGS Volcano Disaster Assistance Program (VDAP). VDAP is deployed outside the United States following a formal request through the U.S. Agency for International Development. USGS employees are guests of the host countries and observatories. Part of their mission is to build capacity. The pattern has been for data to remain in the host country. There is thus an inherent tension between promoting the professional careers of local scientists by giving them primacy in publishing data, and ensuring that unique data are made available to the broader scientific community and archived in perpetuity.
Observatory and academic volcano scientists are currently well positioned to foster partnerships to take full advantage of rich data sets collected through monitoring and ensure that scientific gains from future major eruptions are maximized. Bringing together the different knowledge, expertise, and perspectives of these two groups will contribute to building a strong volcano science community capable of making new discoveries, developing and testing new instrumentation and monitoring techniques, and implementing more accurate and sophisticated forecasting models. USGS–academic partnerships can support the mission of the USGS by expanding the community of scientists studying volcanoes, and by training the next generation of professionals engaged in volcano science. The National Earthquake Hazards Reduction Program program7 has been successful in promoting such partnerships for earthquake science.
Volcano observatories excel at long-term volcano monitoring, as exemplified by the recent centennial of the USGS Hawaiian Volcano Observatory (Babb et al., 2011). These activities result in long-duration data sets that could not be collected by academic researchers through standard (e.g., NSF) funding mechanisms. While the primary use of these data is to assess the current state of a volcano and its potential for eruptive activity, collaborative retrospective analyses of monitoring data have led to many new insights on the fundamental processes that govern volcanic behavior (e.g., Carey et al., 2015). Such collaborative efforts lead to scientific advances while supporting the mission of the observatories as new instruments, analytical techniques, and models become incorporated into monitoring efforts.
Observatory–academic partnerships are an ideal vehicle for training the next generation of volcano scientists through graduate student internships and postdoc positions, taking advantage of the observatories’ proximity to active volcanoes, experienced and multidisciplinary staff, and exposure to the challenges of maintaining monitoring networks. Some of the fundamental questions highlighted in this report could
be tackled through research projects cosupervised by observatory and academic partners and supported through national funding agencies. Educational exchanges would benefit both groups. Moreover, observatory personnel from developing countries could attend graduate school in the United States, ideally leading to knowledge transfer, while access to a student workforce could help observatories fully mine data archives resulting from monitoring activities.
A major challenge in understanding volcanoes is that significant leaps in understanding volcanic processes tend to occur during and immediately following rare well-observed major or otherwise significant eruptions (e.g., Mount St. Helens in 1980, Pinatubo in 1991, and Eyjafjallajökull in 2010). It is thus critical that academic and observatory partners prepare in advance to maximize comprehensive and high-quality observations of the next major eruption (IAVCEI Task Group on Crisis Protocols, 2016). Careful long-term planning during “peacetime” is key for managing complementary objectives during the crisis: forecasting and hazard mitigation, and high-quality data and sample collection. Academic partners would need to provide observatory partners with a description of capabilities and/or resources they could contribute during a crisis or periods of volcanic unrest as well as a summary of data they need to advance scientific understanding. Furthermore, best practices for the collection of critical and ephemeral data and samples need to be established in advance to avoid loss. Finally, academic scientists would need to formulate an action plan to coordinate personnel and equipment and to disseminate the resulting data and samples within the community.
An effective volcano science community requires several elements, including the following:
- Support for interdisciplinary collaboration and training, which is essential to making discoveries and integrating models and measurements;
- Shared community infrastructure, which is necessary for state-of-the-art modeling, analytical facilities, monitoring, and field experiments;
- Databases that preserve and facilitate open exchange of information and hence enable exploration of the life cycle of volcanoes and improve forecasting;
- New technology and instruments that permit new detection, measurements, and sampling, including previously inaccessible parts of ongoing eruptions;
- A coordinated response by the research community to eruptions globally to overcome observational bias; and
- Volcano observatory–academic partnerships, which will accelerate the translation of basic science to applications and monitoring.