Woods Hole, MA 02543 USA
Gas and Oil Seepage and Hydrothermal Venting
in the Ocean Bottom
Detection by Fluorescence
Jean K. Whelan, Nathan Mah, and Greg Eischeid, Departement of Marine
Chemistry and Geochemistry, Woods Hole Oceanographic Institution, Woods
Hole MA 02543
Robert Chen and Xuchen Wang, Department of Environmental Sciences,
University of Massachusetts, Boston, MA 02125
Harry Roberts and Paul Aharon, Center for Coastal Studies and
Department of Geology and Geophysics, Louisiana State University, Baton
Rouge, LA 70803
An abundance of evidence suggests continuous or episodic upward movement
of fluids from deeper sediments into surface sediments and ocean bottom
waters (Whelan, 1997).
These seepages may have been volumetrically
underestimated in the past, both in oil and gas productive areas and in
hydrothermal vent areas
because they often occur through small
seemingly unimportant localized cracks in the seafloor. However, even
if these venting features are small, the volume of expelled fluid can be
important: even a small fracture can deliver orders of magnitude more
fluid than ordinary compaction and diffusion processes
(e.g. Hunt, 1996)
Here we present initial results on successful deployment of CTD
fluorescence for continuous detection of bottom venting fluids both
in a known oil and gas seep area (Green Canyon in the offshore Louisiana
Gulf of Mexico, Figures
and in a hydrothermal vent area
(Guaymas Basin, Gulf of California, Figures
results confirm the very localized and possibly episodic nature of these
venting features which would not have been detected without a continuous
localized detection technique
Louisiana Gulf Coast upper continental shelf Green Canyon area
The offshore Louisiana Gulf Coast area studied from
the manned submersible, the Johnson Sealink in August of 1997 is shown in
Widespread natural oil and gas seepage has been documented over the entire
study area as reviewed in Roberts and Carney
The oil and gas venting
commonly occur through observable faults and fractures in the seafloor.
Here we show data from three dives:
was in 1400 feet of water located in Green Canyon Block 184
where oil seepage was visible from the
submersible as small oil bubbles discharging from bottom sediments. Because
of calm sea conditions, upward moving oil drops were also visible from the
ship, appearing episodically as small slicks dispersing on the sea surface
over the entire area. Large communities of tubeworms and other organisms
were pervasive over the bottom at this site.
shows a small but distinct
signal near the bottom which we attribute to oil seepage accompanied by a
slight increase in DOC
GCMS of a methylene chloride extract
shows a homologous series of n-alkanes typical of oil.
Dive 2894 Bush Hill (Figure 2B) had free
gas bubbles, some 3 feet or more in diameter, disrupted bottom sediments
around the submersible and stirred large chunks of bottom sediments and
thioplocca mat throughout the water column. Previous work showed this
gas to be biogenic methane delivered to the seafloor through a fault from
Pleistocene gas reservoir. The near-bottom CTD fluorescence signal for this
site was very strong throughout the gassy region. This signal disappears
within about 50m of the bottom, probably due to bubble
dissolution during upward ascent. The fluorescence signal is strongest
during the last half of the dive, where the largest bubbles were also
visible. Bubbles were not visually evident in the early part of the dive,
although visibility was poor because of the particles in the water column.
No water was collected so we don't know whether the fluorescence signal is
entirely due to light reflection/refraction of the bubbles or whether there
is also some contribution from bottom pore-water humic substances, as
was in an area of no visual gas or oil bubbling and minimal bottom biology.
No CTD fluorescence was evident in bottom waters
Dive 2900 was in Green Canyon Block 232 in
1867 feet of water. The initial landing site was on a gas hydrate mound
covered with tubeworms. Numerous patches of white and a few of red
Beggiatoa were present. Gas bubbles and expelled yellowish chunks of
hydrate were visible. The small gas bubbles may be responsible for the
somewhat noisy baseline observed in the CTD fluorescence profile near the
bottom (not shown) and for the increases in DOC within 200 feet of the bottom
Trace amount of n-alkanes are present at this site
was at Green Canyon Block 152 in 1500
feet of water was not visually very interesting. It does show a slightly
noisy CTD baseline near the seafloor, similar to that observed for Site
2893, suggesting oil seepage. DOC increases near the bottom
However, GCMS shows little or no n-alkane signal at this site.
Guaymas Basin is a sediment covered hydrothermal
ridge in the Gulf of California
Magma drives waters contained in
sediments upward into the overlying water column. The hydrothermal heating
also produces in situ oil and gas generation in the organic-rich
sediments. Waters from the vents are somewhat enriched in DOC (62 to 69
µM in comparison to normal open ocean mid-water sea water values of
about 45 µM).
The venting fluids also show ubiquitous fluorescence, as
shown by the relative strength of fluorescence on continuous tracks for
on the first day of dives when the whole southern rift area
was surveyed. Comparison of the fluorescence maxima with the cruise event
log clearly shows the association between CTD fluorescence and the
hydrothermal plume. This result was unexpected, but turned out to be quite
useful in locating the plume when it was not visually present.
Generally, the DOC (Seatech) and aromatic hydrocarbon
(Chelsea) fluorometers showed maxima at the same location, with the Seatech
DOC signal being generally stronger
(Figures 5 and 6).
However, in a few
locations, the Chelsea gave a stronger signal possibly indicative of the
presence of hydrothermal petroleum.
Maxima in fluorescence, reflecting maxima in venting
fluids, are very localized, often occurring on less than a one meter scale
These features could have been easily missed without the
continuous CTD fluorescence profiles.
In Green Canyon, all CTD fluorescence was produced by gas and
oil seepage. In Dive 2893 to an oil seep site, the small fluorescence
anomolies are associated with seeping oil. The presence of oil dispersed
in the water column is also shown by a gas chromatograph of an organic
extract of bottom water from the same site which shows the homologous
n-alkanes typical of unbiodegraded oil. These peaks are much weaker for
water collected at Dive Site 2901
The very high CTD fluorescent spikes observed for
Green Canyon Dive site 2894 are most likely caused largely by light
refraction from the large gas bubbles which were visually observed to be
present. This hypothesis is supported by the decrease and then
disappearance of CTD fluorescence during the ascent through the water
Methane has a solubility in water of about 3%, so that
methane bubbles should disappear relatively quickly as they ascend through
the water column. Part of the Dive 2894 fluorescence could be derived from
bottom sediments and the DOM associated with pore waters being flung
upward into the water column by the ejecting gas. No gas bubbles were seen
during the first half of the dive, although the waters were so cloudy that
small bubbles would not have been distinguishable from particles.
Previously, particles alone have not been found to be a source of
fluorescence; however, we have not previously encountered the abundance of
particles observed here. Unfortunately, no water samples were collected
from this site so that we were unable to carry out chemical analyses to
distinguish these possibilities.
CTD fluorescence from the
is clearly tracking the hydrothermal plume
(Figures 5 and
The substance responsible for the fluorescence
could be DOM or petroleum pushed out of the sediment pores and up into the
water column by the venting fluid. Alternatively, fluorescing biota
associated with the chemoautotrophic ecosystem in the plumes could also be
The most important information provided by the Guaymas
CTD fluorescence data is that plume fluorescence is often intense, but is
also very localized, even on distance scales of less than a meter
Any discrete sampling program, even on a very fine grid, would have missed
the fluids corresponding with the fluorescent maxima. Also, even through
Jason tracks are crossing each other on a scale of a meter or less, the CTD
signal is sometimes stronger on one crossing than another. This could be
due to depth variations of the cruise track. Alternatively, fluorescence
and the plume could also be changing temporally, even within the short time
span of a dive.
In the offshore Gulf Coast oil and gas seep area:
1) CTD fluorescence was ubiquitous at all oil and gas seep sites;
2) No fluorescence was observed in areas of no seepage;
3) Free gas bubbles produce a very strong CTD fluorescence signal.
In the Guaymas Basin hydrothermal area:
4) Strong CTD fluorescent signals were found to be diagnostic of the plume.
5) Different types of fluorescent organic matter were
detected using two fluorometers operating at different wavelengths. In
most cases, the signal from the longer wavelength Seatech DOM fluorometer
was stronger. In a few cases, the signal from the shorter wavelength
Chelsea "aromatic" fluorometer was stronger, suggesting a
stronger influence from hydrothermally produced petroleum.
6) Large fluorescence variations occur on very short
distance scales (less than 1m) and also possibly on short time scales (less
than an hour). Fluorescence maxima appear to correspond to plume maxima and
probably would not have been detected by a descrete sampling program.
Figures and Tables
Figure 1: Locations of initial bottom
CTD fluorescence measurements:
overview of location of Gulf Coast Green Canyon area;
close-up: location of Green Canyon manned submersible dive sites;
overview of location of Gulf of California Guaymas Basin hydrothermal
close-up: Guaymas Basin, location of Jason (unmanned submersible)
Green Canyon: CTD fluorescence data, Seatech
DOC fluorometer. Fluorescence is plotted versus time on the left and versus
depth on the right. A) Oil seep site - Dive 2893; B) Gas seep site - Dive
2894; C) Site of no gas or oil leakage, Dive Site 2896
Green Canyon: Dissolved organic carbon
Green Canyon water, GCMS total ion
chromatogram (TIC) of methylene chloride extract in comparison to
tetracosane recovery standard.
Guaymas Basin: CTD fluorescence data, Seatech
(DOC) and Chelsea (aromatic hydrocarbon) fluorometers - overall.
Guaymas Basin comparative fluorometer measurements (dive of 4/20/98)
Summary of global oil seepage data suggesting
seepage to bottom sediments may have been underestimated in past. Note that
total estimated oil seepage volumes from the two single events shown are
approximately equal to the entire estimated world seepage for a single
year. Also note the large petroleum sources potentially available to
produce additional undetected seepage.
Hunt, J.M. 1996. Petroleum Geochemistry and Geology.
Freeman, San Francisco, 2nd Ed., Chapter 9
Peltzer, E.T. and P.G. Brewer (1993) Some practical
aspects of measuring DOC -- sampling artifacts and analytical problems with
marine samples. Marine Chemistry, v.41, pp 243-252.
Roberts, H.H. and Carney, R. (1997)
Evidence of episodic fluid, gas and sedimnet venting on the northern Gulf
of Mexico continental slope. Economic Geology, v. 92, pp 863-879.
Whelan, J.K. (1997)
The dynamic migration hypothesis. How fast are oil and gas leaking to the
ocean floor and replenishing themselves in some reservoirs? Sea Technology,
v.38, No. 9, pp 10-18.
Captain and crew of
Edwin Link and Johnson Sea Link; Dana Yoerger and the Deep Submergence
Group for "piggybacking" us on their Guaymas Jason dive; Al
Bradley and Ellen Druffel for helpful discussions which got us to thinking
about this work. Financial support from the Vetlesen Foundation through a
grant to the Woods Hole Oceanographic Foundation and from the Department of
Energy (grant no. DE-FG02-86ER13466 to Jean Whelan) is gratefully
Initial measurements were carried out using a DOM fluorometer (Sea
Tech) mounted about eight feet from the base of the manned submersible
Johnson Sealink. Continuous measurements were made during 12 dives carried
out over a six day period in August, 1997
(Figure 2). Fluorometer readings in
Figure 2 are relative and were not calibrated to an external standard in this
preliminary work. In the Sea Tech fluorometer, the emission and absorbance
windows are mounted at 90° to each other and fluorescence is measured on in
situ seawater present within a light path of about 1 cm2 outside
of the pressure casing. Maximum excitation and emission wavelengths were
300-360nm and 410-480 nm, respectively.
For later measurements in the Guaymas Basin
hydrothermal vent field from the unmanned submersible, Jason, a second
lower wavelength fluorometer (Chelsea Instruments) was added, maximum
excitation and emission wavelengths 226-252 nm and 347-373 nm,
The signal outputs of the Sea Tech and Chelsea
fluorometer intensities were linear and logarithmic signals,
The water for DOC measurements was collected in 100ml precombusted
bottles following precautions described by Peltzer
and summarized in Peltzer and Brewer
Prior to sample collection, the
sample lines and collection devices were completely flushed with indigenous
seawater to eliminate any "memory" of previous samples. As soon
as possible after each collection, water samples with no visible particles
were treated with phosphoric acid and either frozen or stored at 4°C .
Samples suspected of containing particles were filtered on precombusted
glass fiber filters. Dissolved organic carbon measurements were carried out
using the stripping-combustion method of Peltzer and Brewer
Whenever possible, a larger 5-liter sample was also collected near the
bottom and treated and stored in the same way as the smaller samples.
Water extraction; analysis of methylene chloride extract.
Water samples (500 ml) were spiked with a deuterated tetracosane
recovery standard, extracted with methylene chloride, and the extract
dried with anhydrous sodium sulfate. After concentration by rotary
evaporation, the residual extract was analyzed on a Hewlett Packard 6890
Gas Chromatograph equipped with a model no. 5973 mass selective detector,
a 50m DB1 capillary column, an auto sampler and a HP Chem Station data