Methane escaping from the sea floor to the atmosphere has been a popular
suspect for causing rapid climate changes during and at the end of the last ice
age. But new data derived from a Greenland ice core have delivered a killer blow
to the idea.
Methane (CH4) is a much stronger greenhouse gas than carbon dioxide. It is
usually released from swamps or through biomass burning. But it is also trapped
in huge amounts in some ocean-floor sediments, where it lies buried in a strange
kind of ice known as 'methane clathrate'. These clathrates are stable only
within a certain range of temperatures and pressures; when brought to the
surface, they melt rapidly and release burnable gas to the air.
A catastrophic release of trillions of tonnes of methane is thought to have
triggered a temperature jump some 55 million years ago in an already warm
climate at the Palaeocene/Eocene boundary (see 'Gas leak!'). But some scientists
suspect that similar methane bursts, triggered perhaps by submarine landslides,
sea-level drops or changes in water temperature, may also have caused a number
of rapid warming episodes during and at the end of the last glacial period.
The theory has been popularized as the 'clathrate gun hypothesis'1. But now an
isotope analysis of methane trapped in bubbles of a Greenland ice core seems to
disprove the idea.
No sign of a burp
Todd Sowers, a palaeoceanographer at Pennsylvania State University in
Philadelphia, measured hydrogen isotopes of atmospheric methane from three
distinct warming episodes, 38,000, 14,500 and 11,500 years ago. Methane from
clathrates contains more deuterium (the heavy form of hydrogen) than methane
from land-based sources, thanks in part to the bacteria that create the gas on
the sea floor, and the material they consume.
He found no evidence whatsoever in the data for increased amounts of methane
from marine clathrates. "This means that seafloor methane reservoirs must have
been stable at these times, or at least that no significant amounts of methane
escaped the ocean," says Sowers, whose study is published in Science this week2.
"The data are convincing," says Kai-Uwe Hinrichs, a geochemist at the University
of Bremen in Germany. "They won't exactly increase the attractiveness of the
clathrate gun hypothesis." At least for the three periods Sowell has looked at
in high resolution, they may even be a "killer argument", adds Jerome Chappellaz,
a geochemist at the CNRS Laboratory of Glaciology and Geophysics of the
Environment in Grenoble, France.
Controversial killer
The clathrate gun hypothesis has been controversial from the onset. All
Quaternary warming episodes seem to have been accompanied by increased abundance
of atmospheric methane. But many climate scientists think this is an effect,
rather than the trigger, of warming climates. In most cases, there is evidence
that the methane values only began to rise several decades after the temperature
started to climb.
None of this means that marine methane hydrates don't occasionally erupt,
however. Hinrichs has used fossil remnants of bacteria that flourish only under
high methane concentrations to show that large quantities of the gas must have
been released in the Santa Barbara Basin off California during an event some
44,000 years ago3. This gas didn't necessarily escape to the atmosphere, he
says, but it did come from underwater ice.
Researchers are now exploring the isotopic values of gas bubbles trapped in ice
cores going back some 900,000 years, to find out where methane came from in the
past.
How the world's methane hydrates will respond to future global warming and other
disturbances is uncertain. Seafloor reservoirs currently contain twice as much
methane as all known conventional fossil-fuel reserves. This makes them a target
for the energy industry, but mining the gas could cause a runaway greenhouse
effect, says Sowers.
References
Kennett J. P., Cannariato K. G., Hendy I. L. & Behl R. J. American Geophysical
Union, Special Publication, Methane Hydrates in Quaternary Climate Change: The
Clathrate Gun Hypothesis. 54, (2003).
Sowers T. Science, 311. 838 - 840 (2006).
Hinrichs K.U., Hmelo L. & Sylva S. Science, 299 . 1214 - 1217 (2003).