National Tsunami Warning Center

Abstract Low-frequency (0.01 to 0.2 Hz) seismic noise, arising from pelagic storms, is commonly observed as microseisms in seismic records from land and ocean bottom detectors. One principal research objective, in the study of microseisms, has been to locate their sources. This article reports on an analysis of primary and secondary microseisms (i.e., near and double the frequency of ocean swell) recorded simultaneously on three land-based long-period arrays (Alaskan Long Period Array, Montana Large Aperture Seismic Array, and Norwegian Seismic Array) during the early 1970s. Reliable microseism source locations are determined by wide-angle triangulation, using the azimuths of approach obtained from frequency-wave number analysis of the records of microseisms propagating across these arrays. Two near-shore sources of both primary and secondary microseisms appear to be persistent in the sense that they are associated with essentially constant near-shore locations. Secondary microseisms are observed to emanate from wide-ranging pelagic locations in addition to the same near-shore locations determined for the primary microseisms.

We document through the analysis of 2002–2005 observational data the recent Atlantic Water (AW) warming along the Siberian continental margin due to several AW warm impulses that penetrated into the Arctic Ocean through Fram Strait in 1999–2000. The AW temperature record from our long‐term monitoring site in the northern Laptev Sea shows several events of rapid AW temperature increase totaling 0.8°C in February–August 2004. We hypothesize the along‐margin spreading of this warmer anomaly has disrupted the downstream thermal equilibrium of the late 1990s to earlier 2000s. The anomaly mean velocity of 2.4–2.5 ± 0.2 cm/s was obtained on the basis of travel time required between the northern Laptev Sea and two anomaly fronts delineated over the Eurasian flank of the Lomonosov Ridge by comparing the 2005 snapshot along‐margin data with the AW pre‐1990 mean. The magnitude of delineated anomalies exceeds the level of pre‐1990 mean along‐margin cooling and rises above the level of noise attributed to shifting of the AW jet across the basin margins. The anomaly mean velocity estimation is confirmed by comparing mooring‐derived AW temperature time series from 2002 to 2005 with the downstream along‐margin AW temperature distribution from 2005. Our mooring current meter data corroborate these estimations.

The ocean is key to understanding societal threats including climate change, sea level rise, ocean warming, tsunamis, and earthquakes. Because the ocean is difficult and costly to monitor, we lack fundamental data needed to adequately model, understand, and address these threats. One solution is to integrate sensors into future undersea telecommunications cables. This is the mission of the SMART subsea cables initiative (Science Monitoring And Reliable Telecommunications). SMART sensors would “piggyback” on the power and communications infrastructure of a million kilometers of undersea fiber optic cable and thousands of repeaters, creating the potential for seafloor-based global ocean observing at a modest incremental cost. Initial sensors would measure temperature, pressure, and seismic acceleration. The resulting data would address two critical scientific and societal issues: the long-term need for sustained climate-quality data from the under-sampled ocean (e.g., deep ocean temperature, sea level, and circulation), and the near-term need for improvements to global tsunami warning networks. A Joint Task Force (JTF) led by three U.N. agencies (ITU/WMO/UNESCO-IOC) is working to bring this initiative to fruition. This paper explores the ocean science and early warning improvements available from SMART cable data, and the societal, technological, and financial elements of realizing such a global network. Simulations show that deep ocean temperature and pressure measurements can improve estimates of ocean circulation and heat content, and cable-based pressure and seismic-acceleration sensors can improve tsunami warning times and earthquake parameters. The technology of integrating these sensors into fiber optic cables is discussed, addressing sea and land-based elements plus delivery of real-time open data products to end users. The science and business case for SMART cables is evaluated. SMART cables have been endorsed by major ocean science organizations, and JTF is working with cable suppliers and sponsors, multilateral development banks and end-users to incorporate SMART capabilities into future cable projects. By investing now, we can build up a global ocean network of long-lived SMART cable sensors, creating a transformative addition to the global ocean observing system.

The magnitude 9.0 earthquake centered off the west coast of northern Sumatra (3.307°N, 95.947°E) on December 26, 2004 at 00:59 UTC (United States Geological Survey (USGS) (2005), USGS Earthquake Hazards Program‐Latest Earthquakes, Earthquake Hazards Program, http://earthquake.usgs.gov/eqinthenews/2004/usslav/ , 2005) generated a series of tsunami waves that devastated coastal areas throughout the Indian Ocean. Tide gauges operated on behalf of national and international organizations recorded the wave form at a number of island and continental locations. This report summarizes the tide gauge observations of the tsunami in the Indian Ocean (available as of January 2005) and provides a recommendation for the use of the basin‐wide tide gauge network for future warnings.

To establish how well the new Community Climate System Model, version 4 (CCSM4) simulates the properties of the Arctic sea ice and ocean, results from six CCSM4 twentieth-century ensemble simulations are compared here with the available data. It is found that the CCSM4 simulations capture most of the important climatological features of the Arctic sea ice and ocean state well, among them the sea ice thickness distribution, fraction of multiyear sea ice, and sea ice edge. The strongest bias exists in the simulated spring-to-fall sea ice motion field, the location of the Beaufort Gyre, and the temperature of the deep Arctic Ocean (below 250 m), which are caused by deficiencies in the simulation of the Arctic sea level pressure field and the lack of deep-water formation on the Arctic shelves. The observed decrease in the sea ice extent and the multiyear ice cover is well captured by the CCSM4. It is important to note, however, that the temporal evolution of the simulated Arctic sea ice cover over the satellite era is strongly influenced by internal variability. For example, while one ensemble member shows an even larger decrease in the sea ice extent over 1981–2005 than that observed, two ensemble members show no statistically significant trend over the same period. It is therefore important to compare the observed sea ice extent trend not just with the ensemble mean or a multimodel ensemble mean, but also with individual ensemble members, because of the strong imprint of internal variability on these relatively short trends.

• Results of a community benchmarking exercise for tsunami currents are presented. • Model tests focus on tsunami-induced currents that are shear and separation driven. • Areas affected by large-scale eddies have the greatest inter-model variability. • The need for a statistical modeling approach for eddy-affected areas is discussed. To help produce accurate and consistent maritime hazard products, the National Tsunami Hazard Mitigation Program organized a benchmarking workshop to evaluate the numerical modeling of tsunami currents. Thirteen teams of international researchers, using a set of tsunami models currently utilized for hazard mitigation studies, presented results for a series of benchmarking problems; these results are summarized in this paper. Comparisons focus on physical situations where the currents are shear and separation driven, and are thus de-coupled from the incident tsunami waveform. In general, we find that models of increasing physical complexity provide better accuracy, and that low-order three-dimensional models are superior to high-order two-dimensional models. Inside separation zones and in areas strongly affected by eddies, the magnitude of both model-data errors and inter-model differences can be the same as the magnitude of the mean flow. Thus, we make arguments for the need of an ensemble modeling approach for areas affected by large-scale turbulent eddies, where deterministic simulation may be misleading. As a result of the analyses presented herein, we expect that tsunami modelers now have a better awareness of their ability to accurately capture the physics of tsunami currents, and therefore a better understanding of how to use these simulation tools for hazard assessment and mitigation efforts.

The real time W phase source inversion algorithm was independently running at three organizations (USGS, PTWC and IPGS) at the time of the 2011 off the Pacific coast of Tohoku Earthquake. Valuable results for tsunami warning purposes were obtained 20 min after the event origin time. Within the next hour, as more data became available, the W phase solutions improved, and converged to a common result (Mw ≈ 9.0, dip ≈ 14°). A post-mortem W phase analysis using data selection based on pre-event noise confirmed the Mw = 9.0 result and yielded a best double couple given by (strike/dip/rake = 196°/12°/85°). We also ran the algorithm with increasingly longer periods (T ≈ 1500 sec) to test for the possibility of additional slow slip. The seismic moment remained stable, confirming the prior results.

Abstract The launches of the Sentinel‐1 synthetic aperture radar satellites in 2014 and 2016 started a new era of high‐resolution velocity and strain rate mapping for the continents. However, multiple challenges exist in tying independently processed velocity data sets to a common reference frame and producing high‐resolution strain rate fields. We analyze Sentinel‐1 data acquired between 2014 and 2019 over the northeast Tibetan Plateau, and develop new methods to derive east and vertical velocities with ∼100 m resolution and ∼1 mm/yr accuracy across an area of 440,000 km 2 . By implementing a new method of combining horizontal gradients of filtered east and interpolated north velocities, we derive the first ∼1 km resolution strain rate field for this tectonically active region. The strain rate fields show concentrated shear strain along the Haiyuan and East Kunlun Faults, and local contractional strain on fault junctions, within the Qilianshan thrusts, and around the Longyangxia Reservoir. The Laohushan‐Jingtai creeping section of the Haiyuan Fault is highlighted in our data set by extremely rapid strain rates. Strain across unknown portions of the Haiyuan Fault system, including shear on the eastern extension of the Dabanshan Fault and contraction at the western flank of the Quwushan, highlight unmapped tectonic structures. In addition to the uplift across most of the lowlands, the vertical velocities also contain climatic, hydrological or anthropogenic‐related deformation signals. We demonstrate the enhanced view of large‐scale active tectonic processes provided by high‐resolution velocities and strain rates derived from Sentinel‐1 data and highlight associated wide‐ranging research applications.

A numerical model for the global tsunami computation constructed by Kowalik et al. (2005, 2007a) is applied to the tsunami of November 15, 2006 in the northern Pacific with spatial resolution of one minute. Numerical results are compared to sea level data collected by Pacific DART buoys. The tide gauge at Crescent City (CC) recorded an initial tsunami wave of about 20 cm amplitude and a second larger energy packet arriving 2 hours later. The first energy input into the CC harbor was the primary (direct) wave traveling over the deep waters of the North Pacific. Interactions with submarine ridges and numerous seamounts located in the tsunami path were a larger source of tsunami energy than the direct wave. Travel time for these amplified energy fluxes is longer than for the direct wave. Prime sources for the larger fluxes at CC are interactions with Koko Guyot and Hess Rise. Tsunami waves travel next over the Mendocino Escarpment where the tsunami energy flux is concentrated owing to refraction and directed toward CC. Local tsunami amplification over the shelf break and shelf are important as well. In many locations along the North Pacific coast, the first arriving signal or forerunner has lower amplitude than the main signal, which often is delayed. Understanding this temporal distribution is important for an application to tsunami warning and prediction. As a tsunami hazard mitigation tool, we propose that along with the sea level records (which are often quite noisy), an energy flux for prediction of the delayed tsunami signals be used.

Importance: Convalescent plasma (CP) has been generally unsuccessful in preventing worsening of respiratory failure or death in hospitalized patients with COVID-19 pneumonia. Objective: To evaluate the efficacy of CP plus standard therapy (ST) vs ST alone in preventing worsening respiratory failure or death in patients with COVID-19 pneumonia. Design, Setting, and Participants: This prospective, open-label, randomized clinical trial enrolled (1:1 ratio) hospitalized patients with COVID-19 pneumonia to receive CP plus ST or ST alone between July 15 and December 8, 2020, at 27 clinical sites in Italy. Hospitalized adults with COVID-19 pneumonia and a partial pressure of oxygen-to-fraction of inspired oxygen (Pao2/Fio2) ratio between 350 and 200 mm Hg were eligible. Interventions: Patients in the experimental group received intravenous high-titer CP (≥1:160, by microneutralization test) plus ST. The volume of infused CP was 200 mL given from 1 to a maximum of 3 infusions. Patients in the control group received ST, represented by remdesivir, glucocorticoids, and low-molecular weight heparin, according to the Agenzia Italiana del Farmaco recommendations. Main Outcomes and Measures: The primary outcome was a composite of worsening respiratory failure (Pao2/Fio2 ratio <150 mm Hg) or death within 30 days from randomization. Results: Of the 487 randomized patients (241 to CP plus ST; 246 to ST alone), 312 (64.1%) were men; the median (IQR) age was 64 (54.0-74.0) years. The modified intention-to-treat population included 473 patients. The primary end point occurred in 59 of 231 patients (25.5%) treated with CP and ST and in 67 of 239 patients (28.0%) who received ST (odds ratio, 0.88; 95% CI, 0.59-1.33; P = .54). Adverse events occurred more frequently in the CP group (12 of 241 [5.0%]) compared with the control group (4 of 246 [1.6%]; P = .04). Conclusions and Relevance: In patients with moderate to severe COVID-19 pneumonia, high-titer anti-SARS-CoV-2 CP did not reduce the progression to severe respiratory failure or death within 30 days. Trial Registration: ClinicalTrials.gov Identifier: NCT04716556.

Crescent City (CC) is particularly susceptible to tsunami attack. The combination of tsunami signal enhancement by both distant bathymetric features and local nearshore resonance seems to be responsible for this behavior. Application of global tsunami propagation models to the Kuril Islands Tsunami of November 2006 in the companion paper (Kowalik et al., 2008) delineated the importance of distant bathymetric features for both tsunami wave amplification and for increased duration of the tsunami arriving at CC. These distant features were responsible for delivering the late, high‐energy signal, which was delayed by 2 hours from the predicted arrival time. Additional local amplification and increased duration are caused by the shelf bathymetry adjacent to CC. As the initial tsunami arrives at CC, the shelf trapping mechanism tends to excite the gravest of the natural modes. This oscillation is slowly shifted toward the natural mode closest to the period of the incident tsunami wave packet. Short‐duration wave trains tend to influence the tsunami response at CC harbor in such a way that 1 hour after the first wave arrival a much stronger wave will be generated owing to the tendency of the shelf to initially ring at the frequency of the gravest mode. Practical implications of distant and local amplification relate to the potential hazard near CC, highlighting the need for awareness of the tsunami signal duration. In this respect, two timescales are important for this event: the 2‐hour time delay due to distant bathymetric features and the 1‐hour delay due to local offshore bathymetry.

Seismic imaging methods have provided detailed three-dimensional constraints on the physical properties of magmatic systems leading to invaluable insight into the storage, differentiation and dynamics of magma. These constraints have been crucial to the development of our modern understanding of magmatic systems. However, there are still outstanding knowledge gaps resulting from the challenges inherent in seismic imaging of volcanoes. These challenges stem from the complex physics of wave propagation across highly heterogeneous low-velocity anomalies associated with magma reservoirs. Ray-based seismic imaging methods such as travel-time and surface-wave tomography lead to under-recovery of such velocity anomalies and to under-estimation of melt fractions. This review aims to help the volcanologist to fully utilize the insights gained from seismic imaging and account for the resolution limits. We summarize the advantages and limitations of the most common imaging methods and propose best practices for their implementation and the quantitative interpretation of low-velocity anomalies. We constructed and analysed a database of 277 seismic imaging studies at 78 arc, hotspot and continental rift volcanoes. Each study is accompanied by information about the seismic source, part of the wavefield used, imaging method, any detected low-velocity zones, and estimated melt fraction. Thirty nine studies attempted to estimate melt fractions at 22 different volcanoes. Only five studies have found evidence of melt storage at melt fractions above the critical porosity that separates crystal mush from mobile magma. The median reported melt fraction is 13% suggesting that magma storage is dominated by low-melt fraction crystal mush. However, due to the limits of seismic resolution, the seismological evidence does not rule out the presence of small (&amp;lt;10 km 3 ) and medium-sized (&amp;lt;100 km 3 ) high-melt fraction magma chambers at many of the studied volcanoes. The combination of multiple tomographic imaging methods and the wider adoption of methods that use more of the seismic wavefield than the first arriving travel-times, promise to overcome some of the limitations of seismic tomography and provide more reliable constraints on melt fractions. Wider adoption of these new methods and advances in data collection are needed to enable a revolution in imaging magma reservoirs.

Abstract The broadband moment magnitude Mwp (Tsuboi, et al., 1995) allows for the effective determination of earthquake magnitude by using broadband P waveforms. It was developed to determine moment magnitude of shallow earthquakes around the Japanese Islands for early tsunami warnings. Tsuboi et al. (1995) demonstrated that Mwp shows good agreement with the Mw from Harvard centroid moment tensor (CMT) solutions. In the present study, we show that Mwp is also applicable to deep earthquakes and earthquakes recorded at teleseismic distances. The Mwp proves to be useful for quick, accurate size estimates of earthquakes recorded at both regional and teleseismic distances occurring at any depth.

We study a database of more than 119,000 measurements of the mantle magnitude Mm introduced by Okal and Talandier (1989), obtained since 1999 as part of the operational procedures at the Pacific Tsunami Warning Center. The perfor- mance of this method is significantly affected by the seismic instrumentation at the recording station, with the very-broadband STS-1 and KS54000 systems offering the lowest residuals between measured values of Mm and those predicted from the Har- vard Centroid Moment Tensor (CMT) catalog, and also by the period at which spec- tral amplitudes are measured, with the best results between 70 and 250 sec. With such mild restrictions, estimates of seismic moments can be obtained in real time by retaining either the maximum value of Mm measured on each record, or its average over the various mantle frequencies, with the resulting residuals, on the order of 0.1 0.2 moment magnitude units. Mm deficiencies in the case of the two large earthquakes of Peru (2001) and Hokkaido (2003) are attributed to azimuthal bias from an excess of stations (principally in North America) in directions nodal for the focal mechanism and directivity patterns. We further study a group of more than 3000 measurements of the energy-to-moment ratio H introduced by Newman and Okal (1998), which allows the real-time identification of teleseismic sources violating scaling laws and, in particular, of so-called tsunami The use of a sliding window of analysis in the computation of H allows the separation of late earthquakes, characterized by a delayed but fast moment release, from truly slow earthquakes. Many such events are recognized, notably on major oceanic and con- tinental strike-slip faults. Introduction and Background We present detailed analyses of a large dataset resulting from implementation, as part of the operational procedures of the Pacific Tsunami Warning Center (PTWC), of the real- time calculation of the mantle magnitude Mm and of the slowness parameter H, as defined and introduced by Okal and Talandier (1989) and Newman and Okal (1998), re- spectively. During the past five years, more than 119,000 individual measurements of Mm and 3000 values of H have been computed through automated algorithms. These values contribute to the estimation of the source characteristic of distant earthquakes in the framework of the real-time as- sessment of their tsunamigenic potential. This study offers a progress report on the performance of the algorithms and, in particular, discusses the influence of instrumentation at the various reporting stations. As the final revision of this article was being prepared, the occurrence of the great Sumatra earthquake on 26 De- cember 2004 provided an opportunity to extend the concepts of Mm and H to a range of magnitudes unexplored for the

It is possible that no catastrophe has mobilized the global ocean science and coastal emergency management communities more than the 2004 Indian Ocean tsunami. Though the Pacific tsunami threat was recognized, and a warning system had been in place since 1965, there was no warning system in the Indian Ocean, and almost 230,000 people perished. More broadly, the event highlighted critical gaps in global tsunami science and observation systems. In 2004, real-time coastal and deep-ocean observation systems were almost nonexistent. Tsunami sources were inferred based on rough seismic parameters. Since then, tremendous strides have been made under the auspices of IOC/UNESCO toward better understanding tsunami mechanisms, deploying advanced real-time tsunami observation systems, and establishing tsunami warning and mitigation systems for the four main ocean basins at risk from tsunamis. Nevertheless, significant detection, measurement, and forecast uncertainties remain to meet emergency response and community needs. A new generation of ocean sensing capabilities presents an opportunity to address several of these uncertainties. Ocean bottom pressures can be measured over dense, multisensor grids linking stand-alone buoy systems with emerging capabilities like fiber-optic cables. The increasing number of coastal sea-level stations can better verify forecasts and account for local variability. In addition, GNSS sensors may be able to provide solid-earth data needed to define seismic tsunami sources more precisely in the short timescales required. When combined with advances in seismology, other emerging techniques, and state-of-the-art modeling and computational resources, these capabilities will enable more timely and accurate tsunami detection and measurement to meet identified user needs and save lives. Because of these advances in detection and measurement, the opportunity exists to greatly reduce and/or quantify uncertainties associated with forecasting tsunamis. Providing more timely and accurate information related to tsunami location, arrival time, height, inundation, and duration would improve public trust and confidence and fundamentally alter tsunami emergency response. Additionally, this capability could be integrated with related fields (e.g., storm surge, sea-level rise, tide predictions, and ocean forecasting) to develop and deploy one continuous, real-time, accurate depiction of the always moving boundary that separates ocean from coast, wet from dry, and, sometimes, life from death.

The Joint Task Force, Science Monitoring And Reliable Telecommunications (JTF SMART) Subsea Cables, is working to integrate environmental sensors for ocean bottom temperature, pressure, and seismic acceleration into submarine telecommunications cables. The purpose of SMART Cables is to support climate and ocean observation, sea level monitoring, observations of Earth structure, and tsunami and earthquake early warning and disaster risk reduction, including hazard quantification. Recent advances include regional SMART pilot systems that are the first steps to trans -ocean and global implementation. Examples of pilots include: InSEA wet demonstration project off Sicily at the European Multidisciplinary Seafloor and water column Observatory Western Ionian Facility; New Caledonia and Vanuatu; French Polynesia Natitua South system connecting Tahiti to Tubaui to the south; Indonesia starting with short pilot systems working toward systems for the Sumatra-Java megathrust zone; and the CAM-2 ring system connecting Lisbon, Azores, and Madeira. This paper describes observing system simulations for these and other regions. Funding reflects a blend of government, development bank, philanthropic foundation, and commercial contributions. In addition to notable scientific and societal benefits, the telecommunications enterprise’s mission of global connectivity will benefit directly, as environmental awareness improves both the integrity of individual cable systems as well as the resilience of the overall global communications network. SMART cables support the outcomes of a predicted, safe, and transparent ocean as envisioned by the UN Decade of Ocean Science for Sustainable Development and the Blue Economy. As a continuation of the OceanObs’19 conference and community white paper ( Howe et al., 2019 , doi: 10.3389/fmars.2019.00424 ), an overview of the SMART programme and a description of the status of ongoing projects are given.

Tsunamis rank among the most devastating and unpredictable natural hazards to affect coastal areas. Just 3 years ago, in December 2004, the Indian Ocean tsunami caused more than 225,000 deaths. Like many extreme events, however, destructive tsunamis strike rarely enough that written records span too little time to quantify tsunami hazard and risk. Tsunami deposits preserved in the geologic record have been used to extend the record of tsunami occurrence but not the magnitude of past events. To quantify tsunami hazard further, we asked the following question: Can ancient deposits also provide guidance on the expectable water depths and speeds for future tsunamis?

Abstract The Mw 7.1 47 km deep earthquake that occurred on 30 November 2018 had deep societal impacts across southcentral Alaska and exhibited phenomena of broad scientific interest. We document observations that point to future directions of research and hazard mitigation. The rupture mechanism, aftershocks, and deformation of the mainshock are consistent with extension inside the Pacific plate near the down‐dip limit of flat‐slab subduction. Peak ground motions &amp;gt;25%g were observed across more than 8000 km2, though the most violent near‐fault shaking was avoided because the hypocenter was nearly 50 km below the surface. The ground motions show substantial variation, highlighting the influence of regional geology and near‐surface soil conditions. Aftershock activity was vigorous with roughly 300 felt events in the first six months, including two dozen aftershocks exceeding M 4.5. Broad subsidence of up to 5 cm across the region is consistent with the rupture mechanism. The passage of seismic waves and possibly the coseismic subsidence mobilized ground waters, resulting in temporary increases in stream flow. Although there were many failures of natural slopes and soils, the shaking was insufficient to reactivate many of the failures observed during the 1964 M 9.2 earthquake. This is explained by the much shorter duration of shaking as well as the lower amplitude long‐period motions in 2018. The majority of observed soil failures were in anthropogenically placed fill soils. Structural damage is attributed to both the failure of these emplaced soils as well as to the ground motion, which shows some spatial correlation to damage. However, the paucity of instrumental ground‐motion recordings outside of downtown Anchorage makes these comparisons challenging. The earthquake demonstrated the challenge of issuing tsunami warnings in complex coastal geographies and highlights the need for a targeted tsunami hazard evaluation of the region. The event also demonstrates the challenge of estimating the probabilistic hazard posed by intraslab earthquakes.

The Hawaiian Islands' location in the middle of the Pacific Ocean is threatened by tsunamis from great earthquakes in nearly all directions. Historical great earthquakes M w &gt; 8.5 in the last 100 years have produced large inundations and loss of life in the islands but cannot account for a substantial (≤ 600 m 3 ) paleotsunami deposit in the Makauwahi sinkhole on the Island of Kaua‘i. Using high‐resolution bathymetry and topography we model tsunami inundation of the sinkhole caused by an earthquake with a moment magnitude of M w ~9.25 located in the eastern Aleutians. A preponderance of evidence indicates that a giant earthquake in the eastern Aleutian Islands circa 1425–1665 A.D.—located between the source regions of the 1946 and 1957 great tsunamigenic earthquakes—created the paleotsunami deposit in Kaua‘i. A tsunami deposit in the Aleutians dated circa 1530–1660 A.D. is consistent with this eastern Aleutian source region.

Abstract Arc volcanoes are underlain by complex systems of molten‐rock reservoirs ranging from melt‐poor mush zones to melt‐rich magma chambers. Petrological and satellite data indicate that eruptible magma chambers form in the topmost few kilometres of the crust. However, very few chambers have ever been definitively located, suggesting that most are too short‐lived or too small to be imaged, which has direct implications for hazard assessment and modeling of magma differentiation. Here we use a high‐resolution technology based on inverting full seismic waveforms to image a small, high‐melt‐fraction magma chamber that was not detected with standard seismic tomography. The melt reservoir extends from ∼2 to at least 4 km below sea level (b.s.l.) at Kolumbo—a submarine volcano near Santorini, Greece. The chamber coincides with the termination point of the recent earthquake swarms and may be a missing link between a deeper melt reservoir and the high‐temperature hydrothermal system venting at the crater floor. The chamber poses a serious hazard as it could produce a highly explosive, tsunamigenic eruption in the near future. Our results suggest that similar reservoirs (relatively small but high‐melt‐fraction) may have gone undetected at other active volcanoes, challenging the existing eruption forecasts and reactive‐flow models of magma differentiation.