Arguments
The Antarctic ice sheet is a sleeping giant, beginning to stir
Posted on 14 January 2015 by John Abraham
In a paper I just published with colleague Dr. Ted Scambos from the National Snow and Ice Data Center, we highlight the impact of southern ice sheet loss, particularly the West Antarctic Ice Sheet on sea-level rise around the world.
We know that human emissions of greenhouse gases are causing the Earth’s temperature to rise and are creating other changes across the Earth’s climate system. One change that gets a great deal of attention is the current and future rates of sea-level rise. A rising sea level affects coastal communities around the world; approximately 150 million people live within 1 meter of current sea level.
The waters are rising because of a number of factors. First, water expands as it warms. In the past, this “thermal expansion” was the largest source of sea-level rise. But as the Earth’s temperatures continued to increase, another factor (melting ice, particularly from large ice sheets in Greenland and Antarctica) has played an ever increasing role.
In the Southern Hemisphere, the largest player is the Western Antarctic Ice Sheet (WAIS). It is less stable than Eastern Antarctica and is particularly vulnerable to melting from below by warmed ocean waters. Scientists are closely watching the ice near the edges of the WAIS because they buttress large volumes of ice that are more inland. When these buttressing ice shelves melt, the ice upstream will slide more rapidly toward the ocean waters.
As reported in our paper, according to some studies, “no further acceleration of climate change and only modest extrapolations of the current increasing mass loss rate are necessary for the system to eventually collapse ... resulting in 1-3 m of sea-level rise.” And this is from just one component of the great southern sheets.
What we also discuss is that sea-level rise will not be uniform. Antarctica (and Greenland) are currently losing gigatons of ice each year. That ice is heavy, and we know from first-year physics courses that mass (particularly heavy items) expresses a gravitational attraction. So, all that ice sitting atop Antarctica is pulling ocean waters toward it.
As the ice melts, the gravitational force will lessen, and the waters will “slosh” away from Antarctica. In our paper, we report that sea level rise in the Northern Hemisphere will be greater than the world-wide average whereas sea levels in the region next to Antarctica may actually fall. This means that infrastructure planning on the east and west coasts of North America as well as in Europe must be prepared for a greater than average sea-level rise.
"no further acceleration of climate change and only modest extrapolations of the current increasing mass loss rate are necessary for the system to eventually collapse ... resulting in 1-3 m of sea-level rise"
Wow. Is that eventually, or in this century. I note that the original paper's abstract includes the point: "sea-level rise above the ∼1 m expected by 2100 is possible if ice sheet response begins to exceed present rates"
And is there ANY chance that the ice sheet response will NOT exceed present rates? Isn't it CERTAIN to accelerate?
wili@1: I don't envy the problem glaciologists have before them. As I point out to people, the reason communities set explosives to cause avalanches in the Spring in the Alps and Rockies is not because the resulting avalanche will do less damage, but simply because it gives them the chance to predict when the avalanche will occur, so they can warn people. Otherwise they pretty much have no idea. A similar problem awaits those trying to predict the kinetics of ice flow at the WAIS and in Greenland.
Dr. Abraham,
Thank you for your update on this interesting topic. This is the first I have heard that the uneven distribution of melt water may be significant to planning.
If there was say 1 meter of sea level rise globally from melting in the great ice sheets, approximately how much extra might there be in the Northern Hemisphere? 5 cm? 10 cm? 25cm? Obviously it depends on a lot of factors but can you suggest a ballpark figure. Can you suggest a paper I could read that reviews this topic?
Wili, from the paper (free download):
"The time to reach this collapsed state is 200 to 500 years. This may be further accelerated if calving-face instability is a factor in the retreat. "
However, it is my understanding that once the calving face recedes back behind the grounding line and over the retrograde bedrock slope beyond, thus exposing the underside of the ice sheet to intrusion by the sea, then complete collapse will be inevitable, as it would take a reduction in global mean temperature well below pre-industrial level to halt it.
After reading the paper I read this reference which estimates the sea level rise in the USA as about 1.3 times the global average from melting in the WAIS. Affects from Greenland are also not uniform and might be lower in the USA since Greenland is close to the USA.
michael sweet,
Your question about gravitational rebound (+centrifugal forces of Earth rotation) on regional SLR from Greenland IS was answered by Jerry Mitrovica here. In the embeded video, from 16:00 on, Jerry explains the results of his simulation of 1m SLRe instantanous GIS melt.
As you can see on the picture, SL would fall (effect negative) in Scotland, Scandinavia and Labrador while SL would be up to 1.2m (120%) around SAtlantic (SAmerica) and Equatorial-Northern Pacific.
In US (your particular interest) the effect ranges widely on E coast (from ~50% in NY to some 80% in Miami. On the W coast the effect seems to be uniform 100%.
This link from Univ Colorado shows how variable the SLR is at the regional level. Higher than average increases are east of Japan and the Philipines and north of Australia whereas much of the US seems to be lower than average. As the site says -"Please note that these trends have been determined for a finite period (1993 - present), and reflect the impact of decadal scale climate variability on the regional distribution of sea level rise."
sealevel.colorado.edu/content/map-sea-level-trends
I'm guessing this variability is not our friend as most areas will get more than their share of SLR at some time as the decadal climate varies one way and then the other.
Wait a minute: The gravitatational attraction of the ice is NOT going to be a big effect. As the ice melts it still has the same mass. It's just now in a form that it can slosh around.
If the WAIS slide off into the ocean next Tuesday (doesn't need to melt, just float) and there is enough volume to raise the ocean level by 1-3 meters, that rise will be world wide. There won't be any 6 meters in the Antarctic ocean, and half a meter in Oslo.
That said: Surface waves travel at a few miles per hour. I would (naively, perhaps) expect water level changes to spread at similar speeds. If it spreads at 1.5 miles an hour, then it would take roughly a year to spread to the north pole. I suspect this is an underestimate.
Re: Gravitational rebound. If the reference is to tectonic plates rising after the ice load is removed, this happens on a MUCH slower time scale. Hudson Bay is still rising after the loss of ice 9000 years ago.
"Wait a minute: The gravitatational attraction of the ice is NOT going to be a big effect. As the ice melts it still has the same mass. It's just now in a form that it can slosh around."
The above statement sounds confused. Currently the Greenland ice sheet exerts gravitational attraction that makes local sea level surrounding Greenland higher than it would otherwise be. With part or all of that ice gone, along with its gravitational attraction, local sea level would fall, offsetting some or all of the global rise due to the mass of the Greenland ice being distributed in the ocean. It's not that there would be a new gravitational attraction, but rather an existing local attraction would disappear. The gravitational attraction of continental land masses will remain, however, and since there is more land mass in the northern hemisphere, sea level rise will be greater in the northern hemisphere than in the southern h.
sgbotsford @8, not only can the water slosh around, but it can also be drawn to the equator by centripetal forces. Coupled with the loss of mass at Greenland itself, that means the water is not sloshing around the Greenland coast and there is a large loss of local mass resulting in even more sea water been drawn away. Further, Greenland will start rising as it seeks neutral bouyancy the magma, resulting in the ocean floor near Greenland also rising with a consequent reduction of sea level relative to the coast. Finally, the loss of all that mass in one location will sligthly alter the axial tilt of the Earth, resulting in further changes in depth around the globe. The overall result for a 1 meter sea level from melting of the Greenland Ice Sheet would be something like this:
Of course, the West Antarctic Ice Sheet will also be melting, causing something like this:
Obviously the two effects will partially cancel in some areas, and reinforce in others. The net effect should be similar to the observed rate of sea level change due to ice melt:
To that must be added the sea level rise due to thermal expansion, which will also vary by region.
Y'all are a bit confused about the gravitational and crustal rebound effects from ice sheet melt. When a land ice mass melts and drains into the ocean, the water is redistributed with the speed of gravity waves. For all intents and purposes, we can assume here this is pretty much instantaneous (takes a few weeks - there's some additional effects on ocean currents due to changes in the water's density field, but let's neglect that for now). But now the relative water level around the world's coastlines has not changed uniformly - how's that? That is because of SLA: self-attraction and loading. Self-attraction is about the gravity field: mass has spread from one concentrated place (the ice) over a much larger area (the global ocean). Since water aligns along equipotential surfaces to first order (let's neglect the dynamic sea surface height here ...), and the equipotential surface has just been changed, the relative sea level change is not uniform. So far so good. Now: it also turn's out 'rock-solid' is not so solid afterall, the Earth's crust and upper mantle are actually pretty elastic! Think of memory foam: when you press it down and take off the weight it quickly rebounds. That's the elastic part (elastic also means: instantaneous!). So the solid Earth rises where it was depressed by the ice, and (relative!) sea level there falls. But wait: memory foam also rebounds a while after the weight that depressed it is long gone! That's because it is also visco-elastic, just like the Earth's crust and mantle. This viscoelastic part is called glacial isostatic adjustment (GIA) and takes place for centuries and millenia even after an ice sheet has disappeard (this is why Fennoscandia is rising today even though all ice there disappeared some tenthousand years ago). The gravity part and elastic as well as viscoelsatic part are actually also coupled to each other a little bit, and thus it's pretty complex to solve this so-called 'sea-level-equation' with all feedbacks (like the rotational feedback) and elastic as well as visco-elastic processes taken into account.
Correct FWL, in fact we already have an example of instantaneous effect of gravity changing sea-level in the daily tidal changes. So if the moon can change the sea level down at my local beach by a few metres in half-a-day, I expect that the GIS can do the same.
I personally don't believe that the ice covering Antartica is exerting a significant gravitational force on the water below it, or that this has any impact on the current sea levels at the moment. I agree with the previous comment by sgbotsford - the ice has the same mass regardless of whether it is ice or water - because water has a consistent density. We only really worry about density changes when we consider gases, because of the PV=nRT equation, where pressure and temperature are involved with volume and mols/grams of a molecule.Therefore, the 'reduction' in the gravitational force exerted by the ice on the ocean as it melts isn't very plausible when we think about ocean levels rising. I think there's a confusion between the displacement of fluids when discussing sea levels rising. Also, the paper cited by Mouginot, Rignot and Scheuchl, 2014 in the one mentioned here brings up an interesting point regarding the dynamic movements of ice sheets and how the intial, rapid change in the retreat may have been caused by longitudinal stresses exerted on the ice as basal melting proceeded (which is a natural process when we discuss ice sheets). As the basal melting proceeded, the stress continued to be amplified and a domino-like effect was the end result, where the ice sheet was forced to undergo rapid retreat in a relaively short amount of time. This could explain the period of stabilization that resulted after the rapid change in retreat. They do not really place an emphasis on anthropogenic factors and their closing remarks state that the causes are still largely uknown for the pattern.
KD @13, they way you phrase your comment suggests there is some confusion about what is being claimed. In addition to the total gravitational attraction from the Earth, regions of local mass density exert a gravitational force on nearby bodies. This was first measured in relationship to mountains by Pierre Bouguer and Charles Marie de La Condamine in 1738. Equivalent techniques are the basis of the GRACE mission, which measures mass anomalies.
So, it does not make any difference to the Earth's total gravitational field whether ice is piled up 2 km deep in Greenland or spread millimeters deep across the entire surface of the Ocean. However, it makes a very large difference to the lateral gravitational attraction of the ocean towards Greenland. Enough, of all of the ice currently on Greenland is involved, to make a difference of several meters in the local sea level around Greenland (but not appreciably to global sea level). Further, if the ice is melted from Greenland it does not spread evenly over the Ocean surface, even ignoring gravitational effects. Rather, it tends to preferentially pile up near the equator due to the centripetal force. That in turn creates a smaller mass anomally which results in an increased sea level at the equator due to lateral gravitational attraction.
Overall, the effects are complex to calculate, but they are real. They have actually been observed, as shown in the third figure in my post @10. And the theory behind the effect is as old as Newton, who was the first to propose that mountains would divert plumb bobs from the vertical by a slight amount.
KD - all that water piled onto antarctica, depresses the crust below it. Remove and it the crust rebounds (slowly). It also acts (like a land mass) on the water around it (not below it). Melt it and that mass is then spread thinly over ocean instead of concentrated in one place. The effects of that are mentioned in the paper and the calculations described in Bamber 2009. Are you challenging the Bamber maths?