NASA Analysis Shows Greenland Ice Sheet Losing Mass Faster than Gaining
20 October 2006
For the first time, NASA scientists have analyzed data from direct, detailed satellite measurements to show that ice losses now far surpass ice gains in the shrinking Greenland ice sheet. The study appeared 19 Oct in Science Express, the advance edition of Science.
Using a novel technique that reveals regional changes in the weight of the massive ice sheet across the entire continent, scientists at NASA’s Goddard Space Flight Center reported that Greenland’s low coastal regions lost 155 gigatons (41 cubic miles) of ice per year between 2003 and 2005 from excess melting and icebergs, while the high-elevation interior gained 54 gigatons (14 cubic miles) annually from excess snowfall.
With this new analysis we observe dramatic ice mass losses concentrated in the low-elevation coastal regions, with nearly half of the loss coming from southeast Greenland. In the 1990’s the ice was very close to balance with gains at about the same level as losses. That situation has now changed significantly, with an annual net loss of ice equal to nearly six years of average water flow from the Colorado River.
—Scott Luthcke, NASA Goddard Planetary Geodynamics Laboratory
The study is based on an innovative use of data from the Gravity Recovery and Climate Experiment (GRACE) satellite that reveals detailed information about where and when the Greenland ice mass has changed. Other recent studies using GRACE observations have reported continent-wide ice mass declines, but none has shown these changes in enough detail for scientists to investigate how the change is distributed among the different areas of the ice sheet.
The authors divided Greenland into six drainage systems according to the directions that the ice sheet flows from the interior toward the coasts. The drainage systems were further divided into high- and low-elevation areas. The amount of ice in the two northernmost systems remained essentially unchanged from the start to the end of the 24-month study period (July 2003-July 2005). The largest ice mass loss occurred along the southeastern coast. Continued monitoring in the future is needed to determine whether this ice loss is a long-term trend, the authors point out.
To achieve this more-detailed view of the ice sheet’s behavior, Luthcke and his colleagues used a technique that brings GRACE’s global view of the Earth down to a more local and frequent view. The pair of GRACE satellites orbiting in close formation detect changes in the Earth’s mass directly below them by measuring changes in the distance between the two satellites as the gravitational force of the mass causes each to speed up or slow down.
Standard GRACE data products infer local mass changes from a global data set of these satellite measurements. The new study used only data from over the Greenland region, which produced smaller-scale mass change readings over Greenland every 10 days. This allowed the Goddard scientists to monitor the ice sheet region by region and season by season and compare what they found to other known features of the ice sheet and its dynamic behavior.
The new results capture more precisely where changes are taking place, showing that the losses of ice mass are occurring in the same three drainage systems where other studies have reported increased glacier flow and ice-quakes in outlet glaciers. The new results show a dramatic speed up in the rate of ice mass loss since the late 1990s that is nearly identical to reports earlier this year based on radar measurements of glacier acceleration.
Although the ice mass loss observed in the new study is less than half of what other recent research has reported, the results show that Greenland is now losing 20% more mass than it receives from new snowfall each year.
This is a very large change in a very short time. In the 1990s, the ice sheet was growing inland and shrinking significantly at the edges, which is what climate models predicted as a result of global warming. Now the processes of mass loss are clearly beginning to dominate the inland growth, and we are only in the early stages of the climate warming predicted for this century.
—Jay Zwally, ICESat project scientist
NASA’s ICESat laser satellite is currently acquiring even higher-resolution data on where the ice changes are taking place, which is providing more detail into the mechanisms that are causing the ice changes. The authors are also using ICESat data and the new GRACE technique to study individual drainage systems in Antarctica. GRACE is a joint partnership between NASA and the German Aerospace Center, Deutsches Zentrum für Luft und Raumfahrt. The satellites, launched in 2002, are managed by the Jet Propulsion Laboratory.
Resources:
“Recent Greenland Ice Mass Loss by Drainage System from Satellite Gravity Observations”; S. B. Luthcke, H. J. Zwally, W. Abdalati, D. D. Rowlands, R. D. Ray, R. S. Nerem, F. G. Lemoine, J. J. McCarthy, D. S. Chinn; Science DOI: 10.1126/science.1130776
OK, so net loss = 41 - 14 = 27 cubic miles. But what is this as a percentage of the ice mass? 1%? 10%? 0.01%? Anybody know?
Posted by: JN2 | 20 October 2006 at 06:26 AM
OK JN2, Google is your friend. Total ice mass of Greenland is 2.5 million km^3. Mulitiply by 0.2399 to get cubic miles = approx 600,000 cubic miles. So net loss of 27 cubic miles = 0.000045 or about one 200th of one percent of total ice mass. Is this significant? Are my calcs correct?
Posted by: JN2 | 20 October 2006 at 06:35 AM
I think they issue this press release every 6 weeks or so.
Posted by: Robert Schwartz | 20 October 2006 at 10:10 AM
Yeah 0.0045% loss doesn't seem like much to me, to be honest. You'd think with a 2*C rise that it would have more of an effect than that.
Posted by: Sid Hoffman | 20 October 2006 at 10:31 AM
There is concern that the volume of fresh water entering the ocean on Greenland's eastern seaboard could dilute the Gulf Stream by enough to inhibit thermohaline convection to the sea floor. That in turn could disrupt this critical ocean circulation and expose Europe to far colder winters, though the process operates on very long timescales. The jury is also still out on the extent to which Atlantic and Pacific ocean circulations are coupled via the atmosphere:
http://www.asp.ucar.edu/colloquium/2000/Lectures/sarachik.doc
Perhaps the relevant measure is not the fraction of total ice mass that melts annually but the ratio of meltwater to Gulf Stream mass flow.
What we do know is that the planet experienced at least one "little ice age" recently, during the Middle Ages. It was preceded by the medieval climate optimum, during which Leif Ericsson's men discovered wild grape vines in Labrador (ca. 1000 AD).
http://en.wikipedia.org/wiki/Little_Ice_Age
http://en.wikipedia.org/wiki/Leif_Ericsson
Obviously, those economically significant climate variations were due to natural dynamics. The largely anthropomorphic warming we are seeing today is leading to a much more rapid retreat of arctic ice than is believed to have taken place back then. The effects are still unknown. We're in the middle of a huge, uncontrolled experiment, with our globalized economy as the guinea pig.
Posted by: Rafael Seidl | 20 October 2006 at 01:11 PM
1/200th...~400years for the Greenland Icecap to melt. Sounds a little slow, but one started, it is hard to stop, much less reverse. It might just speed up, and we end up with ~200yrs. All conjecture here. The main variables are:
a) how fast are the glaciers slipping into the seas.
b) how much precip is added each year.
c) how much melts and flows to the sea, thus lubricating the ice above.
d) climate patterns could make it more or less conducive for the ice to melt.
e) the Sun powers the Earth's weather, making shifts in output affect the climate.
There are more, but you get the idea.
Posted by: allen_Z | 20 October 2006 at 01:20 PM
Allen, it will be a cold day in hell when I think I can teach you anything about engineering, or math, for that matter, but I think you slipped a few digits. If JN2 is right, and the independent figures suggest he lowballed the volume by about 5%, then if Greenland is losing 27 cubic miles of ice annually then 630,000/27 gives us about 23,000 years to melt it down. Even if the older figures on ice melt were right it would take 10,000 years to melt the ice up there.
Regarding the Gulf Stream being endangered by an increase in fresh water running off Greenland, the Gulf Stream pushes about 80,000,000 cubic meters a second past Cape Hatteras... I make the total annual melt on Greenland to be about 23 minutes worth of the Gulf Stream at Cape Hatteras. So the ice melt would be about 0.00887% of the Gulf Stream flow in the area.
( I make 27 cubic miles to be about 112,540,000,000 cubic meters, so divided that by the 80,000,000 cubic meters the Gulf Stream moves which gave me 1400 for the seconds it would take for the GS to push the amount of the ice melt. 1400/60 = 23 minutes, which is a fairly miniscule fraction of 180 days when the ice is melting... 259,200 minutes/23 minutes... Obviously numbers are not my day job.)
http://hypertextbook.com/facts/2000/HannaBerenblit.shtml
http://en.wikipedia.org/wiki/Gulf_Stream
http://www.sadruddin.com/convert.htm
Posted by: Ziv | 20 October 2006 at 04:53 PM
Conjecture.
Posted by: allen_Z | 20 October 2006 at 05:44 PM
Allen, are you saying that the melt of the Greenland ice is conjecture, or the idea that if we look at the facts presented by NASA, that the amount of melt is rather small and the Gulf Stream is rather large, is conjecture? But NASA's figures do indicate it would take 20,000 years to melt Greenland don't they? The article seems to indicate that your questions a) b) & c) lead to a conclusion that Greenland is losing 27 cubic miles of ice a year. Your questions d) and e) are more difficult to evaluate, but the numbers seem to indicate that we are seeing 0.00434% of Greenlands ice cap melt each year. Serious, but hardly 'the sky is falling' sort of stuff, or maybe it is 'chicken little' material, come to think of it.
Posted by: ziv | 20 October 2006 at 07:30 PM
Suppose the rate of net melting doubles every 5 years, i.e. 27 cu.mi. in 2005, 54 cu.mi. in 2010 etc. In that scenario, assuming 600,000 cu.mi. now, all the ice could be gone by about 2060.
Year rate loss Ice Left % ice left
2000 13.5 67.5 599932.5 99.99%
2005 27 135 599797.5 99.97%
2010 54 270 599527.5 99.92%
2015 108 540 598987.5 99.83%
2020 216 1080 597907.5 99.65%
2025 432 2160 595747.5 99.29%
2030 864 4320 591427.5 98.57%
2035 1728 8640 582787.5 97.13%
2040 3456 17280 565507.5 94.25%
2045 6912 34560 530947.5 88.49%
2050 13824 69120 461827.5 76.97%
2055 27648 138240 323587.5 53.93%
2060 55296 276480 47107.5 7.85%
I hope it doesn't go like this. Sorry for the format, not much I can do with it. More measurements required.
Posted by: mark | 20 October 2006 at 09:25 PM
Ziv,
I was speculating, and commenting on the future of climate change. More likely than not, it is going to keep melting and at faster rates. All those variables make predictions of the future for the Greenland Icecap a little uncertain. It may go fast (like mark's back of envelope calc; yikes), and may take the Global Ocean Conveyor with it, or it can take thousands of years. Along the way, we may reverse the effects of human pollution, only to see supervolcanos dump up into the next ice age.
Conjecture.
Posted by: allen_Z | 21 October 2006 at 06:35 AM
Of course, theorems originate from conjecture which makes conjecture important to discuss and definately not dismiss.
Also there are degrees of conjecture; I would consider the risks to future generations from Icecap melt far greater than the risk of a supervolcano.
Unfortunately climate change denialists never consider degrees of conjecture, and would use your last sentence, allen Z, as reason to say "why bother".
Posted by: Shaun Williams | 21 October 2006 at 08:36 PM