Arguments
The Past and Future of the Greenland Ice Sheet
Posted on 31 July 2010 by Ned
Guest post by Ned
The Greenland ice sheet has a negative mass balance, meaning that it is losing ice (Velicogna 2009, Jiang 2010). This loss occurs because the gain of new ice (in the form of snowfall within the ice sheet's interior zone of accumulation) cannot keep up with the rapid loss of ice through melting and the discharge of ice by marine terminating outlet glaciers (van den Broeke 2009). Figure 1 shows the overall downward trend in the ice sheet's mass:
Figure 1: Greenland ice mass anomaly (black). Orange line is quadratic fit (John Wahr).
As recently as the 1980s Greenland's ice was probably more or less in equilibrium (Rignot 2008, Jiang 2010). But by the late 1990s and early 2000s this ice began to retreat, and the rate of retreat may be accelerating (Velicogna 2009, van den Broeke 2009). Some claim that there's no need to worry about this, since only a tiny fraction of the total ice mass is being lost each year. Others point out that the accelerating rate of loss means that there may be trouble ahead.
Who is right, and how can we tell? There are two ways of answering this question - we can look at the history of the Greenland ice sheet to see whether past episodes of warmth have led to extensive loss of ice, and we can use physical ice sheet models to predict the future behavior of Greenland's ice in a warming climate.
Pleistocene history of the Greenland ice sheet
A new paper published this month in Quaternary Science Reviews (Alley et al. 2010) provides a very comprehensive overview of the history of the Greenland ice sheet over the past half-million years. In general, the further back we look in time, the less certainty there is about the condition and extent of Greenland's ice.
Paleoclimate scientists divide the Quaternary cycles of glacial cooling and interglacial warmth into a series of Marine Isotope Stages (MIS). The Holocene, the current warm stage in which our civilization has arisen and flourished, is MIS-1. The most recent glacial episode (from approximately 110,000 to 14,000 years ago) is divided into MIS-2, -3, and -4. The last previous interglacial, from 130,000 to 110,000 years ago, is referred to as MIS-5e. Looking back further in time, glacial and interglacial episodes alternate, with MIS-6, -8, and -10 being glacial and MIS-7, -9, and -11 being interglacial.
Note that the earliest of these interglacial episodes (MIS-11, around 400,000 years before present) is believed to be the best analog to our current MIS-1 interglacial climate, based on the geometry of the Earth's orbit (Berger and Loutre, 1991).
History of the Greenland ice sheet prior to the last interglacial: Greenland's ice is believed to have shrunk during warm interglacial episodes MIS-11, -9, and -7. During MIS-11, the most similar to our current interglacial, sea levels were much higher, probably enough to necessitate the near-total loss of Greenland's ice (Alley 2010), and ancient DNA found at the bottom of ice cores likewise suggests an ice-free Greenland at this time (Willerslev 2010). Likewise, during glacial advances MIS-10, -8, and -6, Greenland's ice expanded and global sea levels dropped. The greatest expansion of ice may have occurred during MIS-6, just prior to the last interglacial (Alley 2010).
Greenland during the last interglacial (120,000 years before present): The Greenland ice sheet did not entirely disappear during MIS-5e, though it was smaller and steeper. There is no ice predating MIS-5e at the bottom of cores from south, northwest, and east Greenland, but older ice is present in central and north-central Greenland ice cores, where (based on the gas content of bubbles in the ice) it appears to have been close to or slightly thinner than today's ice sheet (Alley 2010). Figure 2 shows a model comparison of today's ice sheet with the ice sheet during MIS-5e.
Figure 2: Modeled configuration of the Greenland Ice Sheet today (left) and in MIS 5e (right), from Otto-Bliesner (2006).
Alley (2010) conclude that it is probable that the loss of ice from Greenland during this time period contributed approximately 3–4 m to global sea levels, in response to a local warming of around 3°–4 °C in Greenland. This nicely fits with our understanding that sea levels were at least 6 m higher that today; the remainder of that rise would have come from the loss of ice in West Antarctica, mountain glaciers, and thermal expansion of seawater.
The last 100,000 years: The Greenland ice sheet expanded during the final Pleistocene glacial advance (MIS-4, -3, and -2), until around 24,000 years before present, when it covered an area 40% larger than its current extent. With the end of this last glacial episode the world warmed, the other large continental ice sheets in North America and Eurasia retreated, sea levels rose by tens of meters, and Greenland's ice sheet shrunk significantly. However, superimposed on this broad pattern of expansion and contraction were a large number of shorter-duration (millennial-scale) increases and decreases in ice mass, generally associated with changes in North Atlantic circulation and other regional climate transitions.
The future of the Greenland ice sheet
Stone (2010) have used models to simulate the likely future loss of ice from Greenland in response to anticipated future warming. The loss of most of the ice sheet would likely occur at atmospheric CO2 concentrations somewhere between 400 and 560 ppm, a rather disturbing finding given that we are currently at 392 ppm and will probably exceed 560 ppm later this century, as shown in Figure 3.
Figure 3: Atmospheric CO2 concentrations as observed at Mauna Loa from 1958 to 2008 (black dashed line) and projected under the 6 SRES marker and illustrative scenarios. From IPCC.
This loss of ice from Greenland alone would be enough to raise sea levels by roughly six meters. This process would probably take centuries or millennia. The fact that Greenland was largely free of ice during the previous interglacial episode MIS-11 (Alley 2010) confirms that this is not an unrealistic scenario.
Some people find the subject of post-2100 climate changes rather abstract. In contrast, 2100 itself is not that distant -- based on life expectancy data, the average girl born in Japan in 2015 will still be living in 2100. So what are the likely outcomes for Greenland (and global sea level) over the remainder of this century?
An accurate answer to this question requires consideration of the physical constraints on the discharge of ice from Greenland into the surrounding ocean. Figure 4 shows some of these constraints, including the locations of marine terminating glaciers and the fraction of Greenland's bed that is below sea level. Realistic modeling of the kinetics of glaciers suggests that a total increase in sea level of 0.8 m from all sources is likely by 2100, with increases of up to 2 m possible but increasingly unlikely (Pfeffer 2008). This matches closely the results of another semi-empirical study (Vermeer and Rahmstorf 2009) of the relationship between temperature and sea level.
The uncertainty in this range mostly relates to the rate at which ice is lost through calving by Greenland's marine-terminating outlet glaciers; the surface mass balance (between precipitation and melting/runoff) is much more predictable. The 0.8 to 2 m range of global sea level rise by 2100 would imply 7.1 cm of sea level rise from Greenland's surface mass balance, plus 9.3 to 46.7 cm from ice discharged into the ocean by Greenland's outlet glaciers (Pfeffer 2008).
Figure 4: Map showing Greenland and outlet glacier gates; marine-based gates are shown as dark green and nonmarine as black. Regions below sea level are colored blue. Ice velocities at 
2000 m elevation shown by red dots (Pfeffer 2008).
In conclusion, a pessimistic but reasonable scenario would produce the melting of somewhere around 5% of the Greenland ice sheet by 2100, contributing 16 to 54 cm to global sea level rise (which in turn would then total 80 cm to 2 m from all sources). However, at that point the collapse of Greenland's ice sheet would just be getting started - failure to constrain CO2 concentrations below 400-560 ppm would almost certainly lead to the near-total loss of the ice sheet, as we have seen from both model results and comparison to the MIS-11 interglacial climate of 400,000 years ago.
0 
0
2004 Nature, 431, 56-59
DOI: 10.1038/nature02903
Decline of surface temperature and salinity in the western tropical Pacific Ocean in the Holocene epoch
Stott, L., Cannariato, K., Thunell, R., Haug, G. H., Koutavas, A., Lund, S.
There is also evidence of the Northern coast of Greenland being free of sea ice between 8.5 to 6 kyr BP. See here beach ridges, striated boulders & marine sediment
The Bednarski 1989 paper you cited, and many other more recent ones also provide convincing circumstantial evidence (such as driftwood deposits) that at least some of these coastal Arctic areas which are now ice bound were bounded by areas of open water for at least some periods since the last glacial. This is generally accepted and is consistent with the proxy temperature data.
The current global decadal temperature trend is upwards, and appears “amplified” in the high latitude NH. If the current or centennial trend is sustained as modeling suggests, Greenland temperatures will exceed post glacial Holocene temperatures in a relatively short timescale. The word “feeble” is perhaps inappropriate given the likely persistent nature of GHG forcing. The recent Arctic temperature rise is not unprecedented over geological timescales, but is highly significant compared to the past 2000 year proxy records (see Kaufman 2010 update and the zoomed in ice core record from Agassiz (Vinther 2009) and GISP2 (from Alley 2004 update).
It should be emphasized, that the orbital parameter forcing which is thought to have driven the gradual fall in NH temperatures since the post glacial temperature maximum is not global in effect, it is not clearly apparent in the deep sea sediment alkenone temperature proxy records for much of the rest of the globe (as is clear in Rimbu 2003) and is not apparent in the Antarctic ice core records.
The Greenland Ice Sheet mass is currently diminishing at an accelerating rate due to localized warm waters and warm currents transported from lower latitudes Hannah 2009, Di Iorio 2009, Straneo 2010, Rignot 2010 as well as recent rapidly warming Arctic air temperatures. The recent rapid change in positive forcing from increasing anthropogenic GHGs is a new factor not present in previous glaciation/deglaciation cycles. It is effectively a global effect (rather than a high latitude effect such as NH insolation) – which combines with the effects of other forcings. The oversimplistic point you make about higher relative NH insolation forcing levels in the past does not strictly hold.
Perhaps more importantly, the massive ice sheets that covered much of the NH in the last glacial period did melt (in most places completely) over the “deglaciation” time span, and we know rapid changes in temperature have been triggered by a combination of the slow change in solar forcing combined with positive feedbacks possibly from changes in the Atlantic Meridional Overturning Circulation and/or changes in GHG. This shows how sensitive our climate can be to proportionally small changes in forcing. I also do not expect the massive central Greenland ice sheet to vanish overnight, but the relatively rapid recent increased rate of loss is of legitimate concern and should not be belittled.
As you can see depositional elevation over the Greenland ice sheet has decreased by 200 m during 8 millennia when both local air and sea surface temperatures were 2-3°C higher than today. That translates to a rate of -2.5 m/century. The area of the ice sheet is about 1.78×106 km2. Therefore the century scale loss of ice volume was 4400 km3 what makes 4000 km3 meltwater. Surface area of all the seas is 3.6×108 km2, therefore it implies an eustatic sea level rise at a rate of about 11 mm/century. Hardly alarming, even if we take into account that holocene melt rate was highly variable, for short periods it could have been 3-4 times the average.
Eight thousand years of ice sheet response is a bit better sample from a climatic point of view than a decade. You guys are making a fuss about weather noise.
and I am not “you guys”
Sorry for that, sir. It will never happen again.
The difference between elevation histories at core sites in the centre of the Greenland ice sheet and the margins gives evidence that the loss at the margins of the ice sheet is very sensitive to temperature changes and these margins have responded rapidly in the past. This is analogous to what we are currently measuring and seeing. The recent reversal in the long term downwards Arctic temperature trend (Kaufman 2009 etc), appears to require more than is possible from natural variations or forcings, and Vinther argues in his conclusion that relatively small temperature changes could lead to greater mass losses than previously thought.
Vinther 2009 states: “Greenland ice sheet (GIS) is an important concern, especially in the light of new evidence of rapidly changing flow and melt conditions at the GIS margins”.
I agree. These recent mass losses are not “weather noise”.
As present ice loss rate is supposed to be much faster than anything seen during the early holocene when arctic temperatures were higher than now, it is either measurement error or a transient which can not last for long (a "weather event").
Greenland is very interesting, but rest assured, not dangerous for coastal cities.
BTW, I have just found something curious. It is not immediately related to Greenland, but to the wider context of glacial - interglacial transition.
As the great continental ice sheets of America and Europe melted, sea level rose by about 120 m. This added water displaced an equal volume of atmosphere, which increased air pressure over land everywhere on the globe by about 1%, except at former ice margins where uplift of land was larger than that of the atmosphere due to postglacial rebound.
The pressure increase translates to a 0.5°C global land surface warming via lapse rate. This additional heating have not occurred over the oceans, so it must have had a pretty large effect on global circulation patterns.
Anyway, it would deserve a closer look and somewhat more accurate calculations.
In the meantime tell the Chinese to stop emitting untold amounts of soot in the atmosphere that's carried off right into the Arctic by prevalent winds. Soot pollution is not a necessary consequence of burning stuff. To make combustion more efficient and filtering smoke properly is not that expensive, the technology is ready, it is already done in any sane country.
Best Hope for Saving Arctic Sea Ice Is Cutting Soot Emissions, Say Researchers
While we are at it. Going for a ban on small diesel engines (a great source of tiny black carbon particles) and installing proper filters on large ones (including ship engines) is also a possible line of action, far more promising than the futile war on CO2.
Figure 1: Ice volume estimates from 3-D ice sheet models Huy 2 (black) and previous Huy 1 (grey), from Simpson 2009.
The results from Simpson 2009 in terms of volume changes are of course not directly comparable with elevation changes, but the important result from Vinther is the likelihood of greater elevation (and again probably Ice volume) in the early Holocene, and thus again even greater sensitivity of ice loss to temperature changes.
Figure 2: revised elevation from Vinther 2009 (black) vs various 3-D thermomechanical model outputs including Huy 1 (orange) from Vinther 2009.
The revisions based on ice core proxy temperatures from the work of Vinther ties in to more recent modeling work ( Stone 2010) which uses much more up to date estimates of ice thickness and bedrock topography (compared with previous work such as Simpson 2009 or Lunt 2009 etc). The results from Stone et al suggest again a higher sensitivity to temperature and therefore greater potential changes in future for given temperature variations and CO2 concentrations than previously predicted (for example in IPCC 2007)
As for sea level budget contribution due to melting of the Greenland Ice Sheet, you are correct that Simpson does suggest a relatively low averaged value for the early Holocene, but as suggested above, this is likely to be an underestimate. The most recent IPCC (AR4, 2007) value for Greenlands contribution (based on 1993 to 2003 range) were between between +0.14 to +0.28mm/yr. Updated estimates suggest an increase to +0.46mm/yr for the period 2000 to 2008, van den Broeke 2009, Velicogna, 2009, with estimates for the most recent period as high as +0.75mm/yr . In a recent review Cazanave 2010 gives an average of several recent estimates of Greenland Ice Sheet contribution between 2003 and 2007 of 0.5mm/yr. The Alley 2010 paper Ned cites above also has a pertinent chart of various recent estimates of sea level contributions which show apparent acceleration of loss and recent values well above 0.2mm/yr. There is also corroborating evidence of resultant accelerating North Atlantic (precision GPS measured) uplift in coastal bedrock Jiang 2010. Uncertainties in this uplift, and uncertainties due to the relatively short observation period and also different mass loss derivations from GRACE, are still apparent from some of the differing values (by potential factors of two) of ice mass loss in the recent work to date, as noted by Bromwich 2010 and Sorenson 2010. Nevertheless very recent estimates (published August 2010) trying to account for GPS measured uplift estimates still show large and recently accelerating Greenland ice mass loss, even if the absolute magnitude of these losses is yet to be quantified with precision, see: Khan 2010 and Wu 2010. It should be noted that these estimates are placed in the context of general consistency with data from other sources (with longer measurement periods of the variables used to estimate mass variation), and with updated models. There is little argument that the recent Greenland ice mass loss is highly significant and currently increasing. In summary, this does seem to be rather more than “noise”.
(a) Monthly (black) and annual (red) measured BC levels at ice core site D4 in Greenland (71.4 N, 44 W) and (b) relative Winter/Summer levels from site D5, which is approximately 350km further south, from McConnell 2007
Given this evidence it is unlikely that BC contributes to the most recent increases in Arctic surface air temperature and Greenland ice margin melt rates, and is even more unlikely to contribute to warmer waters and warmer Ocean currents causing basal glacier outlet melt and overall accelerated mass loss in Greenland.
Comments