Nearly 3,500 methane bubble streams, clustered into more than 1,300 methane emission sites, have been identified along the US Cascadia margin, derived both from archived published data and 2011, 2016–2018 dedicated multibeam surveys using co-registered seafloor and water column data. In this study, new multibeam sonar surveys systematically mapped nearly 40% of the US Cascadia margin, extending from the Strait of Juan de Fuca in the north to the Mendocino fracture zone in the south, and bounded East–West by the coast and the base of the accretionary prism. The frequency-depth histogram of the bubble emission sites has a dominant peak at the 500 m isobar, which extends laterally along much of the Cascadia margin off Oregon and Washington. Comparisons with published seismic data on the distribution of bottom simulating reflectors (BSR), which is the acoustic impedance boundary between methane hydrate (solid phase) and free gas phase below, correlates the bottom simulating reflectors upward termination of the feather edge of methane hydrate stability (FEMHS) zone and the newly identified bubble emission sites off Oregon and Washington. The Cascadia margin off northern California, where the BSR ends seaward of the FEMHS, has fewer sites centered on the 500 m isobaths, although data are more limited there. We propose that the peak in bubble emission sites observed near the 500 m isobath results from migration of free gas from beneath the solid phase of the BSR upslope to the FEMHS termination zone, and suggest that this boundary will be useful to monitor for a change in methane release rate potentially related to a warming ocean.

The considerable challenges of accessing unpredictable events at remote seafloor locations make submarine eruptions difficult to study in real time. The serendipitous discovery of two persistently active sites (NW Rota-1 in the Mariana arc, at ~550 m, and West Mata in the NE Lau basin at ~1200 m) resulted in multi-year, multi-parameter studies that included water column plume surveys and direct (ROV) observations. Intense magmatic-hydrothermal plumes rose buoyantly above both sites, while deep particle plume layers, dominated by fine ash and devoid of hydrothermal tracers, were found dispersing laterally on isopycnal surfaces at variable depths below the eruptive vents and above the seafloor. The presence or absence of deep ash plumes was directly correlated with explosive activity or quiescence, respectively. An estimated 0.4-14.6 x 105 m3/yr of fine ash entered the water column surrounding these volcanoes and remained suspended at distances exceeding 10’s of km. We show that deep ash plume layers in the water column are a common feature of explosive submarine eruptions at other sites as well, and that they demonstrate a syn-eruptive mode of transport for fine ash that will result in deposition as “hidden” cryptotephra or fallout deposits in marine sediments at distances greater than previously predicted. Cruise FK171110 extended the time series of observations at West Mata, and resulted in discovery of new lava flows emplaced after September 2012, with one constrained between March 2016 and November 2017. ROV dives confirmed that West Mata was quiescent during this expedition, but widespread deep ash plumes were present. Turbidity in the deep ash plumes decreased by 80% over a 25-day period, with an average loss of 3% (0.15-0.6 g/m2) per day, suggesting the eruption that formed the 2016-2017 eruptive deposits had occurred within 8-121 days prior to the FK171110 expedition. Future studies of submarine volcanic processes will depend on improved exploration and event detection capabilities. In addition to recognizing the characteristic hydrothermal event plumes rising into the water column above actively erupting sites, widespread ash plumes dispersing at depths deeper than eruptive vents can also be diagnostic of ongoing, or very recent, eruptions. We infer the eruptive status at other sites based on these criteria.

A recent compilation of methane plumes detected offshore Washington State includes 1,772 individual bubble streams issuing from 491 discrete vent sites. The majority of these plume sites form a narrow 10‐km‐wide band located shallower than 250‐m water depth, with most sites located near the 175‐m‐deep continental shelf break that tracks the head scarps of large submarine canyons. Archive multichannel seismic profiles over the Cascadia shelf and uppermost margin that were co‐located within a few hundred meters with active emission sites show that methane bubble streams arise from listric/normal faults and triangular‐shaped regions of disturbed seismic reflectors that intersect the seafloor and extend several kilometers into the subsurface. Geological processes were evaluated for producing the narrow emission site depths including nonuniform distribution of methane within the Cascadia accretionary sediment wedge and horizontal transfer of groundwater from onshore subaerial sources. A model of enhanced sediment permeability arising from a contrasting response between the inner and outer portions of the accretionary wedge deformation during a megathrust earthquake cycle appears the most likely mechanism. This faulting is generated during extension of the overriding plate during megathrust earthquake cycles, with semicontinuous permeability enhancement of the fluid pathways from excitation by contemporary incident seismic waves.

West Mata is a submarine volcano located in the SW Pacific Ocean between Fiji and Samoa in the NE Lau Basin. West Mata was discovered to be actively erupting at its summit in September 2008 and May 2009. Water-column chemistry and hydrophone data suggest it was probably continuously active until early 2011. Subsequent repeated bathymetric surveys of West Mata have shown that it changed to a style of frequent but intermittent eruptions away from the summit since then. We present new data from ship-based bathymetric surveys, high-resolution bathymetry from an autonomous underwater vehicle, and observations from remotely operated vehicle dives that document four additional eruptions between 2012-2018. Three of those eruptions occurred between September 2012 and March 2016; one near the summit on the upper ENE rift, a second on the NE flank away from any rift zone, and a third at the NE base of the volcano. The latter intruded a sill into a basin with thick sediments, uplifted them, and then extruded lava onto the seafloor around them. The most recent of the four eruptions occurred between March 2016 and November 2017 along the middle ENE rift zone and produced pillow lava flows with a shingled morphology and tephra as well as clastic debris that mantled the SE slope. ROV dive observations show that the shallower recent eruptions at West Mata include a substantial pyroclastic component, based on thick (>1m) tephra deposits near eruptive vents. The deepest eruption sites lack these near-vent tephra deposits, suggesting that pyroclastic activity is minimal below ~2500 mbsl. The multibeam sonar re-surveys constrain the timing, thickness, area, morphology, and volume of the new eruptions. The cumulative erupted volume since 1996 suggests that eruptions at West Mata are volume-predictable with an average eruption rate of 7.8 x 106 m3/yr. This relatively low magma supply rate and the high frequency of eruptions (every 1-2 years) suggests that the magma reservoir at West Mata is relatively small. With its frequent activity, West Mata continues to be an ideal natural laboratory for the study of submarine volcanic eruptions.

Volcanic eruptions are important events in Earth's cycle of magma generation and crustal construction. Over durations of hours to years, eruptions produce new deposits of lava and/or fragmentary ejecta, transfer heat and magmatic volatiles from :Earth's interior to the overlying air or seawater, and significantly modify the landscape and perturb local ecosystems. Today and through most of geological history, the greatest number and volume of volcanic eruptions on Earth have occurred in the deep ocean along mid-ocean ridges, near subduction zones, on oceanic plateaus, and on thousands of mid-plate seamounts. However, deep-sea eruptions (> 500 m depth) are much more difficult to detect and observe than subaerial eruptions, so comparatively little is known about them. Great strides have been made in eruption detection, response speed, and observational detail since the first recognition of a deep submarine eruption at a mid-ocean ridge 25 years ago. Studies of ongoing or recent deep submarine eruptions reveal information about their sizes, durations, frequencies, styles, and environmental impacts. Ultimately, magma formation and accumulation in the upper mantle and crust, plus local tectonic stress fields, dictate when, where, and how often submarine eruptions occur, whereas eruption depth, magma composition, conditions of volatile segregation, and tectonic setting determine submarine eruption style.

New field observations reveal that extensive (up to ~ 402 km2) aphyric, glassy dacite lavas were erupted at multiple sites in the recent past in the NE Lau basin, located about 200 km southwest of Samoa. This discovery of volumetrically significant and widespread submarine dacite lava flows extends the domain for siliceous effusive volcanism into the deep seafloor. Although several lava flow fields were discovered on the flank of a large silicic seamount, Niuatahi, two of the largest lava fields and several smaller ones (“northern lava flow fields”) were found well north of the seamount. The most distal portion of the northernmost of these fields is 60 km north of the center of Niuatahi caldera. We estimate that lava flow lengths from probable eruptive vents to the distal ends of flows range from a few km to more than 10 km. Camera tows on the shallower, near-vent areas show complex lava morphology that includes anastomosing tube-like pillow flows and ropey surfaces, endogenous domes and/or ridges, some with “crease-like” extrusion ridges, and inflated lobes with extrusion structures. A 2 × 1.5 km, 30-m deep depression could be an eruption center for one of the lava flow fields. The Lau lava flow fields appear to have erupted at presumptive high effusion rates and possibly reduced viscosity induced by presumptive high magmatic water content and/or a high eruption temperature, consistent with both erupted composition (~ 66% SiO2) and glassy low crystallinity groundmass textures. The large areal extent (236 km2) and relatively small range of compositional variation (σ = 0.60 for wt% Si02%) within the northern lava flow fields imply the existence of large, eruptible batches of differentiated melt in the upper mantle or lower crust of the NE Lau basin. At this site, the volcanism could be controlled by deep crustal fractures caused by the long-term extension in this rear-arc region. Submarine dacite flows exhibiting similar morphology have been described in ancient sequences from the Archaean through the Miocene and in small batches on present-day seafloor spreading centers. This study shows that extensive siliceous lavas can erupt on the modern seafloor under the right conditions.

We present acoustic records of methane bubble streams recorded similar to 10 km southwest of Heceta Bank on the Oregon continental margin using an autonomous hydrophone. The hydrophone was deployed at 1228 m water depth via a Remotely Operated Vehicle (ROV) during the E/V Nautilus expedition (NA072) in June 2016. Bubble sound is produced by detachment of the gas bubble from the end of a tube or conduit which causes the bubble to oscillate, producing a sound signal. Despite persistent ship propeller and ROV noise, the acoustic signature of the overall bubble seep site can be seen in the hydrophone record as a broadband (1.0 - 45 kHz) series of short duration (similar to 10-20 ms) oscillatory signals that occur in clusters lasting 2-3 s. The frequency of an individual bubble's oscillation is proportional to the bubble's radius; estimates here of bubble radii are consistent with bubble sizes observed in ROV still images. Acoustic signal loss models imply bubble sounds might be recorded over an area of seafloor from similar to 300 - 3.2 x 10(4) m(2). This study represents a first-step in attempting to identify and quantify deep-ocean bubble stream sounds using passive acoustic techniques.

During E/V Nautilus NA072 expedition, multibeam sonar surveys located over 800 individual bubble streams rising from the Cascadia Margin between the Strait of Juan de Fuca and Cape Mendocino at depths between 104 and 2,073 m. Gas bubbles were collected directly at the seafloor using gastight sampling bottles. These bubbles were consistently composed of over 99% methane with traces of carbon dioxide, oxygen, nitrogen, noble gases, and more rarely higher hydrocarbons. A common previous view was that a biogenic source was responsible for seeps from within the gas hydrate stability zone (upper limit near 500‐m isobath) and a thermogenic source was responsible for seeps from the upper slope and the shelf. Higher hydrocarbons in deep seeps with a biogenic methane signature, as well as the lack of higher hydrocarbons in some shallower seeps with a thermogenic methane signature, show that the origin of the gas cannot simply be attributed to seep location on the margin. Instead, mixing and oxidation processes play an integral role. 3He/4He ratios at Coquille SW point to a contribution of 30% mantle helium, whereas all the other investigated sites are characterized by a crustal helium signature. Hence, the Coquille SW seeps are directly or indirectly connected to the mantle or to very young oceanic crust. The detection of mantle helium in these seeps can thus be used as a tracer for deep‐reaching fracture systems and their changing pathways.

Arc volcanoes are important to our understanding of submarine volcanism because at some sites frequent eruptions cause them to grow and collapse on human timescales. This makes it possible to document volcanic processes. Active submarine eruptions have been observed at the summit of NW Rota-1 in the Mariana Arc. We use remotely operated vehicle videography and repeat high-resolution bathymetric surveys to construct geologic maps of the summit of NW Rota-1 in 2009 and 2010 and relate them to the geologic evolution of the summit area over a 10year period (2004-2014). We find that 2009 and 2010 were characterized by different eruptive styles, which affected the type and distribution of eruptive deposits at the summit. Year 2009 was characterized by ultraslow extrusion and autobrecciation of lava at a single eruptive vent, producing a large cone of blocky lava debris. In 2010, higher-energy explosive eruptions occurred at multiple closely spaced vents, producing a thin blanket of pebble-sized tephra overlying lava flow outcrops. A landslide that occurred between 2009 and 2010 had a major effect on lithofacies distribution by removing the debris cone and other unconsolidated deposits, revealing steep massive flow cliffs. This relatively rapid alternation between construction and destruction forms one end of a seamount growth and mass wasting spectrum. Intraplate seamounts, which tend to grow larger than arc volcanoes, experience collapse events that are orders of magnitude larger and much less frequent than those occurring at subduction zone settings. Our results highlight the interrelated cyclicity of eruptive activity and mass wasting at submarine arc volcanoes.