U.S. Dept. of Commerce / NOAA / OAR / PMEL / Publications

Submarine venting of liquid carbon dioxide on a Mariana Arc volcano

J. Lupton1, D. Butterfield2, M. Lilley3, L. Evans4, K. Nakamura5, W. Chadwick Jr.4, J. Resing2, R. Embley1, E. Olson3, G. Proskurowski3,6, E. Baker7, C. de Ronde8, K. Roe3, R. Greene4, G. Lebon2, and C. Young9

1NOAA/Pacific Marine Environmental Laboratory, Newport, Oregon
2JISAO/University of Washington, Seattle, Washington
3School of Oceanography, University of Washington, Seattle, Washington
4CIMRS/Oregon State University, Newport, Oregon
5National Institute of Advanced Industrial Science and Technology, Tsukuba, Japan
6Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts
7NOAA/Pacific Marine Environmental Laboratory, Seattle, Washington
8Institute of Geological and Nuclear Sciences, Lower Hutt, New Zealand
96450 Eagles Crest Road, Ramona, California

Geochem. Geophys. Geosyst., 7, Q08007, doi: 10.1029/2005GC001152, 2006 .
Copyright ©2006 by the American Geophysical Union. Further electronic distribution is not allowed.

Figure 1

Figure 1. (a) Location map for NW Eifuku in the Mariana Arc. (b) Oblique 3-D representation of NW Eifuku viewed from the southwest, generated from EM300 bathymetry. No vertical exaggeration. Depths range from 1550 to 3000 m.

Figure 2

Figure 2. (a) Bathymetric map showing locations of Daikoku, Eifuku, and NW Eifuku submarine volcanoes. Box shows location of Figure 2b. (b) Detailed bathymetric map of NW Eifuku. Box shows location of Figure 2c. (c) High-resolution bathymetry of the summit of NW Eifuku, showing location of the Champagne site and Sulfur Dendrite (SD) and Cliff House (CH) vent sites. This high-resolution bathymetry was collected using the Imagenex sonar system mounted on the ROPOS ROV [Chadwick et al., 2001, 2004].

Figure 3

Figure 3. Photographs of the Champagne hydrothermal site taken with the ROPOS ROV. (a, b, and c) Small chimneys venting 103°C vent fluid. Liquid CO droplets are also visible. (d) Close-up of liquid CO droplets rising in a stream from the seafloor. (e) Liquid CO droplets collecting on the underside of the ROV bumper-bar and camera. (f) Mussel bed only tens of meters from the Champagne vent site. See also Movies 1 and 2. Additional photographs and video clips from the 2004 Submarine Ring of Fire expedition and from NW Eifuku in particular are available at the Ocean Exploration Web site: http://oceanexplorer.noaa.gov/explorations/04fire/logs/april10/april10.html and http://oceanexplorer.noaa.gov/explorations/04fire/logs/photolog/photolog.html.

Figure 4

Figure 4. Diagram of near surface conditions at the Champagne vent field. The "?" indicates possible regions where liquid CO and/or CO hydrate are being entrained into the vent fluid flow.

Figure 5

Figure 5. Schematic of the extraction line used for sample processing at sea.

Figure 6

Figure 6. Photographs of sampling at the Champagne site in 2004 and 2005. (a) Fluid sampler being inserted into 103°C hydrothermal vent. (b) Droplets of liquid CO being collected in an inverted plastic cylinder held in the ROV arm. (c) Photograph of the plastic cylinder taken at about 400 m depth during the ROV's ascent to the surface. Most of the liquid droplets have converted to hydrate, and the hydrate is beginning to sublime into gaseous CO. (d) Close-up of the "droplet catcher" used during the 2005 expedition attached to the small volume gas-tight bottle. (e and f) The ROV Hyper-Dolphin sampling liquid CO with the droplet catcher and gas-tight bottle at the Champagne site.

Figure 7

Figure 7. Phase diagram for CO showing regions where solid, liquid, gas, and super-critical fluid (SCF) exist. P and T denote the critical pressure and temperature. The dashed line denotes the boundary of hydrate stability [Sloan, 1990]. The P, T conditions for the Champagne site liquid droplets and for the 103°C vent fluid are shown.

Figure 8

Figure 8. Vent fluid concentrations for CO, HS, and He versus Mg for NW Eifuku vent fluids. Vent designations are CH (Champagne), SD (Sulfur Dendrite), CL (Cliff House), and DS (Diffuse Site). The 2004 collections are shown in red; the 2005 collections are shown in blue. Fluid discharge temperatures are indicated in parentheses. All of these sites are near the Champagne vent field (see Figure 1b). Possible mixing lines are shown indicating end-member compositions for 2004 samples (red lines) and 2005 samples (blue lines). The solubility of CO in water at these conditions (160 bars, 100°C) is shown for comparison.

Figure 9

Figure 9. Histogram comparing estimated end-member CO concentrations for vent fluids from mid-ocean ridges [Kelley et al., 2004], the Okinawa Trough [Sakai et al., 1990a, 1990b], and NW Eifuku (this work). For the MOR and Okinawa Trough samples, the end-member concentrations were derived in the usual way by extrapolating to zero Mg. For the NW Eifuku samples a value of 43 mmol/kg was used for this end-member extrapolation (see text for explanation).

Table 1

Table 1. Gas Compositions for Vent Fluid and Liquid Droplet Samples From NW Eifuku

Table 2

Table 2. Isotope Ratios for Vent Fluid and Liquid Droplet Samples From NW Eifuku

Table 3

Table 3. Estimated End-Member Compositions, Based on Extrapolating to a Mg Concentration of 43 mmol/kg

Figure 10

Figure 10. Plots showing He, He, and CO data collected in the water column over NW Eifuku in 2003 (filled triangles) and 2004 (filled circles). (a) [He] and CO versus depth. The depth of the Champagne vent field is indicated for comparison. (b) [He] versus [He] showing an estimate of the end-member He/He ratio based on a linear regression fit. Here R = He/He and R = R = 1.39 × 10. (c) CO versus [He] showing a similar estimate of the end-member CO/He ratio.

Figure 11

Figure 11. Diagram showing a model proposed for the gas flux from NW Eifuku, in which a CO-rich gas is directly degassing from the magma chamber. As this hot gas rises through the system, it cools, and CO condenses as a separate liquid phase on the periphery of the main conduit. Seawater circulates through the system, but the penetration of water into the core of the system is limited at temperatures below 250°C due to CO-HO immiscibility. At the volcano summit the liquid CO collects beneath a hydrate "cap" layer that forms where the liquid CO comes in contact with seawater. Because the penetration of water is limited and the enthalpy is carried by the CO gas, there is little high-temperature water-rock interaction.

Figure 12

Figure 12. (a) Chart comparing C (‰) for CO from various MOR sites [Kelley et al., 2004], the Okinawa Trough [Sakai et al., 1990a, 1990b], NW Eifuku (this work), marine carbonates [Hoefs, 1980], and typical volcanic arcs [Sano and Williams, 1996; van Soest et al., 1998]. (b) Similar chart comparing CO/He ratios for MOR vents [Kelley et al., 2004], the Okinawa Trough [Sakai et al., 1990a, 1990b], and typical volcanic arcs [Sano and Williams, 1996; van Soest et al., 1998].

Return to previous section or return to Abstract

PMEL Outstanding Papers

PMEL Publications Search

PMEL Homepage