Mid-ocean ridge explosive eruptions

Benthic foraminifera with agglutinated limu o Pele (foram ~2mm across)
Image © MBARI 2003

The volcanic eruptions at mid-ocean ridges have been thought only to be quietly effusive, but because we find glassy fragments of lava bubbles (limu o Pele) at mid-ocean ridges, there must be gas-rich mildly explosive eruptions at mid-ocean ridges as well. We have now found these particles at many locations and depths along the earth’s mid-ocean ridge system, so mildly explosive eruptions must be quite common.

The particles are especially abundant where the lava erupted as jumbled sheet flows, so their formation is related to the rise and release of bubbles of magmatic gas that also drive the lava’s fast eruption rate. The particles take many forms, from curved plates of limu o Pele, to stretched rods and intricately folded and tack-welded thin sheets and hairs. They can be distributed quite a distance from the rocks extruded during the same eruptions, meaning that they must have been entrained in heated megaplumes that lifted the particles high into the water column.

Benthic foraminifera often glue particles to their tests, perhaps for protection from predators. These particles may be sponge spicules, sand grains, or other detritus, depending on the materials available and the “specialty” of the foram. In sediment cores from the Gorda Ridge, we found forams that “specialized” in volcanic glass grains and others that “specialized” in limu o Pele. They effectively concentrated the glass samples for us!

For other examples in this website of explosive eruptions occurring at depth see also

Our research on explosive eruptions at mid-ocean ridges

Distictly different explosive eruption styles recorded in sediment cores

AXIAL VOLCANO – A comprehensive understanding of explosive basaltic eruption processes in the deep-sea relies upon detailed analysis and comparison of the variety of volcaniclastic lithologies on the seafloor, which has been challenged by insufficient sample recovery. A dedicated ROV-based sampling approach using long push cores offers an unparalleled opportunity to fully characterize the diversity of unconsolidated volcaniclastic lithofacies on a recently active seamount. Lithofacies from Axial Seamount record two styles of pyroclastic eruptions, strombolian and phreatomagmatic, at 1.5 km water depth. Strombolian eruptions are represented by abundant fluidal and highly vesicular (up to 50%) vitriclasts within limu o Pele lapilli tuff and tuffaceous mud lithofacies. Lapilli-ash grain size, normal grading, good sorting, rip-up clasts and homogeneous glass geochemistry characterize individual limu o Pele lapilli tuff beds, and imply proximal deposition from a turbidity flow associated with a single eruption (i.e. event bed). Limu o Pele lapilli tuff beds are interbedded with poorly sorted, chemically heterogeneous and bioturbated tuffaceous mud units that preserve reworking and biologic habitation of more distal pyroclastic fallout and dilute turbidity flows. The phreatomagmatic eruption style is preserved by hydrothermal mineral-bearing muddy tuff that exhibits characteristics distinct from lapilli ash and tuffaceous mud lithofacies. Hydrothermal muddy tuff lithofacies are well-sorted and fine-grained with notable components of non-fluidal basaltic ash (∼45%), fluidal ash (∼30%) and accessory lithics (∼25%). Heterogeneous geochemistry of ash shards implies that juvenile components are minimal. The abundance, mineralogy and texture of lithic components (Fe–Mg clays, pyrite, epidote, actinolite, altered glass, basalt/diabase, hydrothermal breccia and agglutinate), and very fine-grain size of basaltic ash, are consistent with phreatomagmatic eruption deposits. A lack of bioturbation or other interbedded lithofacies, and presence of normal grading suggests prolonged eruption activity and deposition via turbidity flows or suspension fallout. The proximity of ancient hydrothermal muddy tuff lithofacies and active hydrothermal vents to caldera walls suggest that phreatomagmatic activity was linked to shallow circulation of fluids along caldera ring-faults rooted to underlying magma conduits and shallow reservoirs. This study provides evidence for two distinctly different pyroclastic eruption styles and provides a framework to further develop existing models of deep-sea explosive volcanism.

Reference: Portner, R.A., D.A. Clague, C. Helo, B.M. Dreyer, J.B. Paduan (2015) Contrasting styles of deep-marine pyroclastic eruptions revealed from Axial Seamount push core records. Earth Planet. Sci. Lett., 423: 219-231. doi: 10.1016/j.epsl.2015.03.043.

Extremely rapid cooling rates of limu o Pele

AXIAL VOLCANO – We present a calorimetric analysis of pyroclastic glasses and glassy sheet lava flow crusts collected on Axial Seamount, Juan de Fuca Ridge, NE Pacific Ocean, at a water depth of about 1400 m. The pyroclastic glasses, subdivided into thin limu o Pele fragments and angular, blocky clasts, were retrieved from various stratigraphic horizons of volcaniclastic deposits on the upper flanks of the volcanic edifice. Each analysed pyroclastic sample consists of a single type of fragment from one individual horizon. The heat capacity (cp) was measured via differential scanning calorimetry (DSC) and analysed using relaxation geospeedometry to obtain the natural cooling rate across the glass transition. The limu o Pele samples (1 mm grain size fraction) and angular fragments (0.5 mm grain size fraction) exhibit cooling rates of 104.3 to 106.0 K s− 1 and 103.9 to 105.1 K s− 1, respectively. A coarser grain size fraction, 2 mm for limu o Pele and 1 mm for the angular clasts yields cooling rates at the order of 103.7 K s− 1. The range of cooling rates determined for the different pyroclastic deposits presumably relates to the size or intensity of the individual eruptions. The outer glassy crusts of the sheet lava flows were naturally quenched at rates between 63 K s− 1 and 103 K s− 1. By comparing our results with published data on the very slow quenching of lava flow crusts, we suggest that (1) fragmentation and cooling appear to be coupled dynamically and (2) ductile deformation upon the onset of cooling is restricted due to the rapid increase in viscosity. Lastly, we suggest that thermally buoyant plumes that may arise from rapid heat transfer efficiently separate clasts based on their capability to rise within the plume and as they subsequently settle from it.

Reference: Helo, C., D.A. Clague, D.B. Dingwell, and J. Stix (2013). High and highly variable cooling rates during pyroclastic eruptions on Axial Seamount, Juan de Fuca Ridge. Journal of Volcanology and Geothermal Research, 253: 54-64. doi: 10.1016/j.jvolgeores.2012.12.004. [Article]

Elevated fluxes of CO2 can drive explosive eruptions

AXIAL VOLCANO – The abundance of volatile compounds, and particularly CO2, in the upper oceanic mantle affects the style of volcanic eruptions. At mid-ocean ridges, eruptions are generally dominated by the gentle effusion of basaltic lavas with a low volatile content. However explosive volcanism has been documented at some ocean spreading centres, indicative of abundant volatile compounds. Estimates of the initial CO2 concentration of primary magmas can be used to constrain the CO2 content of the upper oceanic mantle, but these estimates vary greatly. Here we present ion microprobe measurements of the CO2 content of basaltic melt trapped in plagioclase crystals. The crystals are derived from volcanic ash deposits erupted explosively at Axial Seamount, Juan de Fuca Ridge, in the northeast Pacific Ocean. We report unusually high CO2 concentrations of up to 9,160 ppm, which indicate that the upper oceanic mantle is more enriched in carbon than previously thought. We furthermore suggest that CO2 fluxes along mid-ocean ridges vary significantly. Our results demonstrate that elevated fluxes of CO2 from the upper oceanic mantle can drive explosive eruptions at mid-ocean ridges.

Reference: Helo, C., Longpre, M.-A., Shimizu, N., Clague, D.A., Stix, J. (2011) Explosive eruptions at mid-ocean ridges driven by CO2-rich magmas, Nature Geoscience, doi:10.1038/NGEO1104. [Supplementary information]

Strombolian eruptions are widespread at mid-ocean ridges

MID-OCEAN RIDGES – Glassy lava fragments were collected in pushcores or using a small suction-sampler from over 450 sites along the Juan de Fuca Ridge, Blanco Transform Fault, Gorda Ridge, northern East Pacific Rise, southern East Pacific Rise, Fiji back-arc basin, and near-ridge seamounts in the Vance, President Jackson, Taney, and a seamount off southern California. The samples consist of angular glass fragments, limu o Pele, Pele’s hair, and other fluidal fragments formed during pyroclastic eruptions. Since many of the sites are deeper than the critical point of seawater, fragmentation cannot be hydrovolcanic and caused by expansion of seawater to steam.

The glass fragments have a wide range of MORB compositions, ranging from fractionated to primitive and from depleted to enriched. Enriched magmas, which have higher volatile contents, may form more abundant pyroclasts than depleted magmas. Eruptions with high effusion rates produce sheet flows and abundant pyroclasts whereas those with low effusion rates produce pillow ridges and few pyroclasts. This relation suggests that high effusion and conduit rise rates are coupled to high magmatic gas contents. The eruptions are mainly effusive with a minor strombolian bubble burst component. We propose that the gas phase is an added component of variable amounts of magmatic foam from the top of the magma reservoir. As the mixture of resident magma and foam rises in the conduit, the larger bubbles in the foam rise more quickly and sweep up the smaller bubbles nucleating and growing from the resident magma. On eruption, the process of bubble coalescence is more complete for the slower rising, gas-poor lavas that erupt as pillow lavas whereas the limu o Pele associated with sheet flow eruptions commonly contain several percent vesicles that avoided coalescence during ascent. The spatter erupted at the vent is quench granulated in seawater above the vent, reducing the pyroclast grainsize. The granulated spatter and limu o Pele fragments are then entrained in a rising plume of seawater heated by the eruption, which disperses them to distances as great as 5 km from the vent.

Reference: D.A. Clague, J.B. Paduan, A.S. Davis (2009) Widespread strombolian eruptions of mid-ocean ridge basalt, Journal of Volcanology and Geophysical Research, 180: 171-188, doi:10.1016/j.jvolgeores.2008.08.007. [Article]

Submarine strombolian eruptions

GORDA RIDGE – Compositionally variable limu o Pele (lava bubble-wall fragments) occurs in widely distributed sediments collected during ROV Tiburon dives along the Gorda Ridge axis. The fragments were formed deeper than the critical depth of seawater so are unlikely to have been generated by supercritical expansion of seawater upon heating in contact with hot lava. Discharge of CO2 through erupting lava is the most likely way to make such bubbles at >298 bars pressure. The distribution and composition of limu o Pele fragments indicate that low-energy strombolian activity is a common, although minor, component of eruptions along mid-ocean ridges.

Combined dissolved and exsolved volatile contents of N-MORB from the Gorda Ridge with 12.8-15.6% spherical vesicles are about 0.78% CO2 and 0.18 wt% H2O and exceed estimates of primary CO2 of only 0.07 to 0.095 wt% calculated from whole rock Nb concentrations. This discrepancy suggests that the magmas accumulated an exsolved volatile phase prior to eruption. The evidence that a separated volatile phase drives strombolian eruptions on the seafloor also implies that gas bubbles coalesce during storage or transport to the surface. The combination of large bubbles in otherwise dense magma suggests nearly complete coalescence of small bubbles and is most consistent with accumulation of the exsolved volatile phase, most likely near the tops of crustal magma chambers, prior to upward transport in shallow conduits to the eruptive vents on the seafloor. A portion of this CO2-rich separated fluid phase is released in brief bursts during eruptions where it becomes part of event plumes.

Reference: D.A. Clague, A.S. Davis, J. E. Dixon (2003) Submarine strombolian eruptions on the Gorda mid-ocean ridge, In: Explosive Subaqueous Volcanism, J.D.L. White, J.L. Smellie, and D.A. Clague (eds), Geophysical Monograph 140, American Geophysical Union, 111-128.

Glass grains offer clues

GORDA RIDGE – Volcanic glass fragments recovered from sediment collected in the Escanaba Trough and on traverses up the west walls of the Gorda Ridge greatly extend information on the petrologic diversity and volcanic evolution of the region, compared to that obtained from traditional samples from lava outcrops. Glass fragments were found loose in sediments and cemented on benthic foraminifera that constructed their tests from detrital grains. Many of the glass grains have compositions that closely match those of nearby lava outcrops, indicating that they formed by spalling off pillow rims and quench granulation. However, the glass compositions also provide information on compositions of unexposed flows that were covered by later eruptions and/or sediment. Bubble-wall glass shards (limu o Pele), found only in the sediment, give evidence of explosive eruptions that occurred despite great depth.

Some of the glass fragments from the walls of the axial valley have unusually high K2O (to >0.7%) that is not represented by any sampled outcrop; such enriched mid-oceanic-ridge basalt compositions have not previously been reported from the Gorda Ridge. Analyzing glass grains from the sediment in addition to samples from lava outcrops allows a more comprehensive characterization of the volcanic history of the region.

Reference: A.S. Davis, D.A. Clague (2003) Got glass? Glass from sediment and foraminifera tests contribute clues to volcanic history, Geology, 31(2): 103-106. [Abstract] [Article]


Upper-ocean systems
Biological oceanography
Biological oceanography research
Publication—Global modes of sea surface temperature
Chemical sensors
Chemical data
Land/Ocean Biogeochemical Observatory in Elkhorn Slough
Listing of floats
SOCCOM float visualization
Periodic table of elements in the ocean
Profiling float
Marine microbes
Population dynamics of phytoplankton
Microbial predators
Microbe-algae interactions
Targeted metagenomics
In the news
Upcoming events and lab news
Past talks and presentations
Join the lab
Molecular ecology
Molecular systematics
SIMZ Project
Bone-eating worms
Gene flow and dispersal
Molecular-ecology expeditions
Interdisciplinary field experiments
Genomic sensors
Ocean observing system
Midwater research
Midwater ecology
Deep-sea squids and octopuses
Food web dynamics
Midwater time series
Respiration studies
Zooplankton biodiversity
Seafloor processes
Biology and ecology
Effects of humans
Ocean acidification, warming, deoxygenation
Lost shipping container study
Effects of upwelling
Faunal patterns
Past research
Technology development
High-CO2 / low-pH ocean
Benthic respirometer system
Climate change in extreme environments
Monitoring instrumentation suite
Sargasso Sea research
Antarctic research
Long-term time series
Geological changes
Arctic Shelf Edge
Continental Margins and Canyon Dynamics
Coordinated Canyon Experiment
Monterey Canyon: Stunning deep-sea topography revealed
Ocean chemistry of greenhouse gases
Emerging science of a high CO2/low pH ocean
Submarine volcanoes
Mid-ocean ridges
Magmatic processes
Volcanic processes
Explosive eruptions
Hydrothermal systems
Back arc spreading ridges
Near-ridge seamounts
Continental margin seamounts
Non-hot-spot linear chains
Eclectic seamounts topics
Margin processes
Hydrates and seeps
California borderland
Hot spot research
Hot-spot plumes
Magmatic processes
Volcanic processes
Explosive eruptions
Volcanic hazards
Hydrothermal systems
Flexural arch
Coral reefs
ReefGrow software
Eclectic topics
Submarine volcanism cruises
Volcanoes resources
Areas of study
Microscopic biology research
Open ocean biology research
Seafloor biology research
Automated chemical sensors
Methane in the seafloor
Volcanoes and seamounts
Hydrothermal vents
Methane in the seafloor
Submarine canyons
Earthquakes and landslides
Ocean acidification
Physical oceanography and climate change
Ocean circulation and algal blooms
Ocean cycles and climate change
Research publications
Full publications list