Geoscience Reference
In-Depth Information
The total volume of CO
2
that is stored in sediments blan-
keting the Pacific trench and active seamount systems is
unknown. At present, liquid CO
2
accumulations have
only been observed at sites in the western Paci
Today, submarine volcanoes that emit a separate liquid and
gas phase of CO
2
occur at shallow to intermediate water
depths. Lupton et al. [2008] point out that it is at these depths
and at the elevated temperatures that circulating
c. However,
CO
2
-rich fluids have been identified at other active vents
sites throughout the Paci
fluids boil
and degassing occurs. The CO
2
and CO
2
-rich fluids released
in this process vent directly into the water column or pass
through sediments blanketing the margin of volcanic centers.
Our hypothesis calls for increased storage and reduced net
c. The distribution of sites where
there is liquid CO
2
includes the Okinawa Trough [Inagaki et
al., 2006], the volcanoes along the Mariana and the Tonga-
Kermadec arcs [Lupton et al., 2008]. These and other occur-
rences of CO
2
-rich vent
fluids from sedi-
ments at volcanic centers located at intermediate depth as the
hydrate stability zone expanded vertically and horizontally
during glacial cooling (Plate 3). At the water depths where
hydrate is stable today (Plate 2), the
flux of CO
2
as liquid CO
2
and CO
2
-rich
fluids [Cheminée et al., 1991; Hilton
et al., 1998; McMurtry et al., 1993; Wheat et al., 2000] hint
at a potentially far greater distribution of sites that emit liquid
CO
2
and CO
2
-rich
fluids. We are drawn to these recent
observations as a way to explain the enigmatic nature of the
glacial/interglacial CO
2
changes summarized above, and
particularly the large
flux of CO
2
from active
volcanic centers is governed by the CO
2
production rate. But
the
flux of CO
2
from the sediments through which much of
the venting
14
C anomaly during the deglaciation,
while recognizing that hypothesis presented here will require
a more quantitative estimate of carbon
Δ
fluids pass also depends on the rate of diffusion
of CO
2
across the hydrate-water interface and the rate of
formation and subsequent dissolution of CO
2
at the bottom
and top of the hydrate layer (respectively) [Rehder et al.,
2004]. Whereas the rate of diffusion between seawater and
pure CO
2
hydrate has been determined empirically [Rehder
et al., 2004], the in situ diffusional
uxes and CO
2
stor-
age at active vent sites.
The CO
2
-rich fluids produced at volcanic arcs in the Pacific
have
13
C values similar to that of marine carbonate [Lupton
et al., 2006]. In fact, the elevated CO
2
:
3
He and carbon isotope
chemistry indicate a slab origin for the CO
2
. Lupton et al.
[2006] argue that the CO
2
is derived from decarbonation of
marine carbonate that is carried down the trench. As a result,
the CO
2
that vents into the water column at this vent site
would have virtually no in
δ
flux across a sediment
hydrate has not. Boundary layer constraints will influence the
diffusion
flux [Rehder et al., 2004]. Furthermore, because the
formation of hydrate leads to salt rejection, high salinity,
CO
2
-saturated pore waters would further reduce the rate of
exchange of CO
2
with the overlying seawater.
The critical depth for hydrate stability in the ocean today
occurs at 8.5°C and ~400 m (Plate 3, left). Plate 3 illustrates
the temperature (left) at ~400 m water depth [Levitus, 1994]
where in the modern ocean, the critical point of hydrate-
liquid-gaseous CO
2
occurs and how hydrate stability would
have changed in the glacial ocean (right) in response to
cooling. Orange shading highlights the area for which CO
2
hydrate becomes unstable at 400 m. Glacial subsurface tem-
peratures were estimated by adding the averaged glacial-
interglacial subsurface temperature anomalies simulated by
the CCSM3-coupled general circulation model and the
LOVECLIM Earth system model [Otto-Bliesner et al.,
2006; Timmermann et al., 2009] to the observed Levitus
temperatures at 400 m.
During the Last Glacial Maximum, most of the active
volcanic centers within the Paci
uence on the
13
C/
12
C of dissolved
carbon in the ocean. The alkalinity of vent
fluids also varies
significantly among various vent sites, and this appears to
re
ect the stage of magmatic development [McMurtry et al.,
1993; Resing et al., 2009]. This is a critical observation
because at the Kasuga vent, for example, McMurtry et al.
[1993] report titration alkalinities that are 13 times higher
than ambient seawater. The dissolved CO
2
content is estimated
to be 230 mmol kg
1
indicating fluids that are highly super-
saturated with CO
2
, principally as bicarbonate.
5. AN OCEAN CO
2
CAPACITOR?
The hypothesis we set forth here calls upon expansion of
the CO
2
hydrate stability horizon in the ocean during glacials
as temperatures cooled. Expansion and shoaling of the
hydrate stability horizon during glaciations would increase
the volume of sediment within which CO
2
accumulated and
lowered the net
c would have fallen within
the hydrate stability zone (
≤
8.5°C, 400 m) as the 8.5°C
isotherm shoaled by more than 100 m. As a result, the overall
flux of CO
2
to the water column. This
hypothesis makes two explicit but testable assumptions:
(1) sediments that blanket volcanic centers are capable of
storing large volumes of CO
2
as hydrate and CO
2
-rich
flux of CO
2
from active volcanic centers would have been
reduced due to a transition from buoyant liquid/gas CO
2
to
hydrate CO
2
. The Mariana and the Kermadec volcanic arcs
are located in the region where the hydrate stability zone
expanded during the last glacial. There are also extensive
regions throughout the Paci
uids
and (2) present-day estimates of the steady state
flux of CO
2
from volcanic centers is significantly underestimated. If either of
these assumptions proves incorrect, our hypothesis is nulli
ed.
c, particularly in the eastern
Search WWH ::
Custom Search