Image above:
A NASA Goddard Institute for Space Studies (GISS) computer
model simulated the 8,200-year climate response to freshwater
entering the Hudson Bay. The left and right panels show simulated
changes in surface air temperature and precipitation, respectively (Image
credit: NASA GISS).
This work is the first to consistently recreate the event
by computer modeling, and the first time that the model results
have been confirmed by comparison to the climate record,
which includes such things as ice core and ocean sediment
data.
“We only have one example of how the climate reacts
to changes, the past,” said Gavin Schmidt, a NASA Goddard
Institute for Space Studies (GISS) researcher and co-author
on the study. “If we’re going to accurately simulate
the Earth’s future, we need to be able to replicate
past events. This was a real test of the model’s skill.”
The team used the NCCS’s 1,392-processor HP/Compaq
AlphaServer SC45 for the ensemble simulations. Each simulation
used one shared-memory node (four processors), which calculated
approximately three model-years per day.
“Over the 12 simulations and controls that we used,
we simulated over 3,000 model-years,” Schmidt said. “With
a number of false starts, adjustments, and the evolution
of our research, it took about 18 months to do all the simulations.”
The study was led by Allegra LeGrande, a graduate student
in the department of Earth and Environmental Sciences at
Columbia University. The results appeared in the journal Proceedings
of the National Academy of Sciences (PNAS) in January
2006.
The group used the coupled atmosphere-ocean model known
as GISS Model E-R to simulate the climate impact of a massive
freshwater flood into the North Atlantic that happened about
8,200 years ago after the end of the last Ice Age. Retreating
glaciers opened a route for two ancient meltwater lakes,
known as Agassiz and Ojibway, to suddenly and catastrophically
drain from the middle of the North American continent.
At approximately the same time, climate records show that
the Earth experienced its last abrupt climate shift. Scientists
believe that the massive freshwater pulse interfered with
the ocean’s overturning circulation, which distributes
heat around the globe. According to the record of what are
known as “climate proxies,” average air temperatures
apparently fell as much as several degrees in some areas
of the Northern Hemisphere.
Climate researchers use these proxies, chemical signals
locked in minerals and ice bubbles as well as pollen and
other biological indicators, as indirect measures of temperature
and precipitation patterns in the distant past. Because the
GISS Model E-R simulates several of these past climate proxies,
the authors of the PNAS study were able to compare their
results directly to the historical record.
The researchers prodded their model with a freshwater flow
equal to between 25 and 50 times the flow of the Amazon River
in the 12 model runs. Although the simulations largely agreed
with records from North Atlantic sediment cores and Greenland
ice cores, the team’s results showed that the flood
had much milder effects around the globe than many people
thought.
According to the model, temperatures in the North Atlantic
and Greenland showed the largest decrease, with slightly
less cooling over parts of North America and Europe. The
rest of the Northern Hemisphere, however, showed very little
effect, and temperatures in the Southern Hemisphere remained
largely unchanged. Moreover, ocean circulation, which initially
dropped by half after simulated flood, appeared to rebound
within 50 to 150 years.
“The flood we looked at was even larger than anything
that could happen today,” said LeGrande. “Still,
it’s important for us to study because the real thing
occurred during a period when conditions were not that much
different from the present day.”
The GISS Model E-R is also being used for the latest simulations
by the Intergovernmental Panel on Climate Change to simulate
the Earth’s present and future climate. “Hopefully,
successful simulations of the past such as this will increase
confidence in the validity of model projections,” said
Schmidt.
The study was funded by NASA, the National Defense Science
and Engineering Graduate Fellowship program, and the National
Science Foundation.
http://www.giss.nasa.gov |