ALBERT

Description: Monowai is an active submarine volcanic center in the Kermadec Arc, Southwest Pacific Ocean. During May 2011, it erupted over a period of 5 days, with explosive activity directly linked to the generation of seismoacoustic T phases. We show, using cross-correlation and time-difference-of-arrival techniques, that the eruption is detected as far as Ascension Island, equatorial South Atlantic Ocean, where a bottom moored hydrophone array is operated as part of the International Monitoring System of the Comprehensive Nuclear-Test-Ban Treaty Organization. Hydroacoustic phases from the volcanic center must therefore have propagated through the Sound Fixing and Ranging channel in the South Pacific and South Atlantic Oceans, a source-receiver distance of ~15,800 km. We believe this to be the furthest documented range of a naturally occurring underwater signal above 1 Hz. Our findings, which are consistent with observations at regional broadband stations and long-range, acoustic parabolic equation modeling, have implications for submarine volcano monitoring.

Description: Only little is known on active volcanism in the ocean. As eruptions are attenuated by seawater and fallout does not regularly reach the sea surface, eruption rates and mechanisms are poorly understood. Estimations on the number of active volcanoes across the modern seas range from hundreds to thousands, but only very few active sites are known. Monowai is a submarine volcanic centre in the northern Kermadec Arc, Southwest Pacific Ocean. During May 2011, it erupted over a period of five days, with explosive activity directly linked to the generation of seismoacoustic tertiary waves (‘T-phases’), recorded at three broadband seismic stations in the region. We show, using windowed cross-correlation and time-difference-of-arrival techniques, that T-phases associated with this eruption are detected as far as Ascension Island, South Atlantic Ocean, where two bottom-moored hydrophone arrays are operated as part of the International Monitoring System (IMS) of the Comprehensive Nuclear-Test-Ban Treaty Organization (CTBTO). We observe a high incidence of T-phase arrivals during the time of the eruption, with the angle of arrival stabilizing at the geodesic azimuth between the IMS arrays and Monowai. T-phases from the volcanic centre must therefore have propagated through the Sound Fixing And Ranging (SOFAR) channel in the South Pacific and South Atlantic Oceans and over a total geodesic range of approximately 15,800 km, one of the longest source-receiver distances of any naturally occurring underwater signal ever observed.

Description: Submarine volcanic activity is difficult to detect, because eruptions at depth are strongly attenuated by seawater. With increasing depth the ambient water pressure increases and limits the expansion of gas and steam such that volcanic eruptions tend to be less violent and less explosive with depth. Furthermore, the thermal conductivity and heat capacity of water causes rapid cooling of ejected products and hence erupted magma cools much more quickly than during subaerial eruptions. Therefore, reports on submarine volcanism are restricted to those sites where erupted products – like the presence of pumice rafts, gas bubbling on the sea surface, and local seawater colour changes - reach the sea surface. However, eruptions cause sound waves that travel over far distances through the Sound-Fixing-And-Ranging (SOFAR) channel, so called T-waves. Seismic networks in French Polynesia recorded T-waves since the 1980’s that originated at Monowai Volcano, Kermadec Arc, and were attributed to episodic growth and collapse events. Repeated swath-mapping campaigns conducted between 1998 and 2011 confirm that Monowai volcano is a highly dynamic volcano.

Description: The Kamchatka Peninsula in far eastern Russia represents the most volcanically active arc in the world in terms of magma production and the number of explosive eruptions. We investigate large-scale silicic volcanism in the past several million years and present new geochronologic results from major ignimbrite sheets exposed in Kamchatka. These ignimbrites are found in the vicinity of morphologically-preserved rims of partially eroded source calderas with diameters from ∼ 2 to ∼ 30 km and with estimated volumes of eruptions ranging from 10 to several hundred cubic kilometers of magma. We also identify and date two of the largest ignimbrites: Golygin Ignimbrite in southern Kamchatka (0.45 Ma), and Karymshina River Ignimbrites (1.78 Ma) in south-central Kamchatka. We present whole-rock geochemical analyses that can be used to correlate ignimbrites laterally. These large-volume ignimbrites sample a significant proportion of remelted Kamchatkan crust as constrained by the oxygen isotopes. Oxygen isotope analyses of minerals and matrix span a 3‰ range with a significant proportion of moderately low-δ18O values. This suggests that the source for these ignimbrites involved a hydrothermally-altered shallow crust, while participation of the Cretaceous siliceous basement is also evidenced by moderately elevated δ18O and Sr isotopes and xenocryst contamination in two volcanoes. The majority of dates obtained for caldera-forming eruptions coincide with glacial stages in accordance with the sediment record in the NW Pacific, suggesting an increase in explosive volcanic activity since the onset of the last glaciation 2.6 Ma. Rapid changes in ice volume during glacial times and the resulting fluctuation of glacial loading/unloading could have caused volatile saturation in shallow magma chambers and, in combination with availability of low-δ18O glacial meltwaters, increased the proportion of explosive vs effusive eruptions. The presented results provide new constraints on Pliocene–Pleistocene volcanic activity in Kamchatka, and thus constrain an important component of the Pacific Ring of Fire.

Description: Wide-angle seismic data have been used to determine the velocity and density structure of the crust and uppermost mantle beneath the Cape Verdes mid-plate swell. Seismic modelling reveals a ‘standard’ oceanic crust, ∼8 km in thickness, with no direct evidence for low-density bodies at the base of the crust. Gravity anomaly modelling within the constraints and resolution provided by the seismic model, does not preclude, however, a layer of crustal underplate up to 3 km thick beneath the swell crest. The modelling shows that while the seismically constrained crustal structure accounts for the short-wavelength free-air gravity anomaly, it fails to fully explain the long-wavelength anomaly. The main discrepancy is over the swell crest where the gravity anomaly, after correcting for crustal structure, is higher by about 30 mGal than it is over its flanks. The higher gravity can be explained if the top 100 km of the mantle beneath the swell crest is less dense than its surroundings by 30 kg m−3. The lack of evidence for low densities and velocities in the uppermost mantle, and high densities and velocities in the lower crust, suggests that neither a depleted swell root or crustal underplate are the origin of the observed shallower-than-predicted bathymetry and that, instead, the swell is most likely supported by dynamic uplift associated with an anomalously low density asthenospheric mantle.

Description: The Cape Verde Islands are located on a mid-plate topographic swell and are thought to have formed above a deep mantle plume. Wide-angle seismic data have been used to determine the crustal and uppermost mantle structure along a ~ 440 km long transect of the archipelago. Modelling shows that ‘normal’ oceanic crust, ~ 7 km in thickness, exists between the islands and is gently flexed due to volcano loading. There is no direct evidence for high density bodies in the lower crust or for an anomalously low density upper mantle. The observed flexure and free-air gravity anomaly can be explained by volcano loading of a plate with an effective elastic thickness of 30 km and a load and infill density of 2600 kg m− 3. The origin of the Cape Verde swell is poorly understood. An elastic thickness of 30 km is expected for the ~ 125 Ma old oceanic lithosphere beneath the islands, suggesting that the observed height of the swell and the elevated heat flow cannot be attributed to thermal reheating of the lithosphere. The lack of evidence for high densities and velocities in the lower crust and low densities and velocities in the upper mantle, suggests that neither a crustal underplate or a depleted swell root are the cause of the shallower than expected bathymetry and that, instead, the swell is supported by dynamic uplift associated with the underlying plume.

Description: Seismicity and tectonic structure of the Alboran Sea were derived from a large amphibious seismological network deployed in the offshore basins and onshore in Spain and Morocco, an area where the convergence between the African and Eurasian plates causes distributed deformation. Crustal structure derived from local earthquake data suggests that the Alboran Sea is underlain by thinned continental crust with a mean thickness of about 20 km. During the 5 months of offshore network operation, a total of 229 local earthquakes were located within the Alboran Sea and neighboring areas. Earthquakes were generally crustal events, and in the offshore domain, most of them occurred at crustal levels of 2 to 15 km depth. Earthquakes in the Alboran Sea are poorly related to large-scale tectonic features and form a 20 to 40 km wide NNE-SSW trending belt of seismicity between Adra (Spain) and Al Hoceima (Morocco), supporting the case for a major left-lateral shear zone across the Alboran Sea. Such a shear zone is in accord with high-resolution bathymetric data and seismic reflection imaging, indicating a number of small active fault zones, some of which offset the seafloor, rather than supporting a well-defined discrete plate boundary fault. Moreover, a number of large faults known to be active as evidenced from bathymetry, seismic reflection, and paleoseismic data such as the Yusuf and Carboneras faults were seismically inactive. Earthquakes below the Western Alboran Basin occurred at 70 to 110 km depth and hence reflected intermediate depth seismicity related to subducted lithosphere.

Description: New marine geophysical data recorded across the Tonga-Kermadec subduction zone are used to image deformation and seismic velocity structures of the forearc and Pacific Plate where the Louisville Ridge seamount chain subducts. Due to the obliquity of the Louisville Ridge to the trench and the fast 128 mm yr−1 south–southwest migration of the ridge-trench collision zone, post-, current and pre-seamount subduction deformation can be investigated between 23°S and 28°S. We combine our interpretations from the collision zone with previous results from the post- and pre-collision zones to define the along-arc variation in deformation due to seamount subduction. In the pre-collision zone the lower-trench slope is steep, the mid-trench slope has ∼3-km-thick stratified sediments and gravitational collapse of the trench slope is associated with basal erosion by subducting horst and graben structures on the Pacific Plate. This collapse indicates that tectonic erosion is a normal process affecting this generally sediment starved subduction system. In the collision zone the trench-slope decreases compared to the north and south, and rotation of the forearc is manifest as a steep plate boundary fault and arcward dipping sediment in a 12-km-wide, ∼2-km-deep mid-slope basin. A ∼3 km step increase in depth of the middle and lower crustal isovelocity contours below the basin indicates the extent of crustal deformation on the trench slope. At the leading edge of the overriding plate, upper crustal P-wave velocities are ∼4.0 km s−1 and indicate the trench fill material is of seamount origin. Osbourn Seamount on the outer rise has extensional faulting on its western slope and mass wasting of the seamount provides the low Vp material to the trench. In the post-collision zone to the north, the trench slope is smooth, the trench is deep, and the crystalline crust thins at the leading edge of the overriding plate where Vp is low, ∼5.5 km s−1. These characteristics are attributed to a greater degree of extensional collapse of the forearc in the wake of seamount subduction. The northern end of a seismic gap lies at the transition from the smooth lower-trench slope of the post-collision zone, to the block faulted and elevated lower-trench slope in the collision zone, suggesting a causative link between the collapse of the forearc and seismogenesis. Along the forearc, the transient effects of a north-to-south progression of ridge subduction are preserved in the geomorphology, whereas longer-term effects may be recorded in the ∼80 km offset in trench strike at the collision zone itself.