A submarine eruption on the Northeast Lau Spreading Center: A rapid response effort
While most ocean expeditions are planned well in advance, scientists try to deploy rapidly when an unusual event is detected in progress. In May 2009, MBARI volcanologist David Clague and his team were given the rare opportunity to observe a volcano erupting in the deep sea. This provided the chance for the researchers to confirm their ideas about how such eruptions behave. Until this expedition, most volcanologists had been limited to interpretations based on lava flows and deposits that sometimes erupted thousands to millions of years ago. Deep submarine eruptions have only been detected a few times, and only Northwest Rota-1, a volcano in the Marianas Arc, had been observed in action, so this opportunity was highly unusual and very exciting.
Chemical signatures were detected in the water column in November 2008 that indicated active eruptions at two places in the Lau Basin, at West Mata Volcano and at a site 65 kilometers away at the Northeast Lau Spreading Center (NELSC). This Lau rapid response expedition was primarily funded by the National Science Foundation and the National Oceanic and Atmospheric Administration. The cruise combined ROV Jason II, MBARI mapping AUV D. Allan B., and conductivity-temperature-depth (CTD) measurements. The field program was highly multidisciplinary, focusing primarily on water chemistry and microbiology, but included experts in biology, geology, petrology, and volcanology.
The team set sail in May to the Lau Basin on the R/V Thompson to map, observe, and collect samples of the active eruptions. The NELSC eruption had ceased, but at West Mata Volcano, two active volcanic vents were found near the summit. MBARI’s team deployed the mapping AUV to collect high-resolution seafloor mapping data to understand the geologic setting, and used the ROV to assess the physical volcanology of the active and recent eruptions.
A single 17-hour AUV mission along 77 kilometers of trackline mapped the NELSC. The NELSC spreading segment is about 10 kilometers long, and the depth of the axis ranges from about 1,500 meters to 1,800 meters. The segment consists of five distinct subsegments delineated by major changes in morphology and subtle changes in trend.
The morphology varied from a narrow ridge in the north, to a wide flat plateau, to a line of discrete pillow mounds on a wide dome-shaped ridge, and finally to a nearly conical volcano in the south that shoals to 1,520 meters at the summit. This volcano had a pair of well-developed rift zones extending from the summit along the axis of spreading. Across the axis, the slope and texture was consistent with being a steep apron of loose sand and gravel of volcanic glass. A number of inward-facing fault scarps cut the flanks of the volcano parallel to the axis of spreading, suggesting recent extension or the early stages of caldera formation.
Two more 17-hour AUV missions surveyed the actively erupting West Mata Volcano, providing almost complete coverage of the summit region and the primary rift zones.
West Mata is a nearly conical volcano that rises to a depth of 1,200 meters. The summit is a 60-meter-long, five-meter-wide ridge bounded by curved steep walls on both sides. The two active volcanic vents were on the upper rim of the northern wall, about five meters below the summit crest. The volcano’s three rift zones were covered by thick lobate lava flows, a few of which had collapse features. The flanks of the volcano were mostly smooth slopes of loose volcanic sand and gravel debris, in a few places capped by lava flows cascading down from the rift zones.
The AUV maps provided the spatial framework for understanding the ROV observations of the volcanoes. Two active vents were found near the summit of West Mata. Clague’s group named them Hades and Prometheus, and visited them several times with the ROV (Figure 14). The Prometheus vent erupted with fire fountains during each visit. A dense rain of glassy particles fell from the smoke-like plume of gas. Rock fragments (clasts) that fell near the vent were pulled back into the vent by water flowing in to replace that in the rising plume; these clasts were rounded like beach cobbles from being recycled numerous times. The Hades vent activity varied among passive degassing of fumes, low fire fountains, and Strombolian activity (vigorous expulsions of incandescent gas that blew large lava bubbles). When the bubbles burst, abundant spatter and thin, glassy, delicate, and intricately folded volcanic glass threads called “limu o Pele” and “Pele’s hairs” were produced. This is the same type of material the MBARI team had found at other deep-sea eruption sites around the world, and confirmed their interpretation that the particles form from mildly explosive eruptions.
The eruptive behavior at these vents appears to be primarily related to magma supply or rise rates, with vigor probably varying due to depth in the conduit where fragmentation takes place. Passive degassing occurs when the eruption stagnates and gases escape from the magma deep in the conduit. Fire fountains result at higher lava rise rates, when degassing and magmatic gas bubble formation occur in the shallow part of the conduit, fragmenting the lava into pieces that are ejected during the eruption. Strombolian bubble-burst activity (Figure 15), generating spatter and accompanied by extrusion of pillow lavas, occurs at intermediate lava rise rates that allow some of the magmatic gas bubbles to coalesce in the shallow conduit.
Geochemical analysis of the volcanic glass fragments from the two active vents at West Mata indicates that the eruption consists of boninite lava, an unusual primitive type of lava that is high in silicon dioxide and erupts in the early stages of volcanic arc formation. Analysis of fragments more distant from the active vents indicates that much of the volcaniclastic debris mapped by the AUV on the surface is also boninite, although quite variable in composition. A sand sample from partway down the southwest rift contains a range of basalt compositions, derived from several eruptions. At the NELSC, the eruption in November 2008 produced a lava flow of fluid, bubble-rich basalt from the crest of the ridge axis that extends for about 1.9 kilometers. Much of the ridge and flanks are buried beneath deposits of glassy sands. Clague suspects these volcaniclastic deposits were produced, at least in part, during the November eruption, as the range of composition of the sands is small and close to that of the lava flow.