Arguments
Why is Antarctic sea ice growing?
Posted on 12 December 2013 by dana1981
This is a re-post of an article at The Conversation by Guy Williams from the University of Tasmania
Recently NASA reported that this year’s maximum wintertime extent of Antarctic sea ice was the largest on record, even greater than the previous year’s record.
This is understandably at odds with the public’s perception of how polar ice should respond to a warming climate, given the dramatic headlines of severe decline in Arctic summertime extent. But the “paradox of Antarctic sea ice” has been on climate scientists' minds for some time.
Continental v. sea ice
First off, sea ice is different to the “continental ice” associated with polar ice caps, glaciers, ice shelves and icebergs. Continental ice is formed by the gradual deposition, build up and compaction of snow, resulting in ice that is hundreds to thousands of metres thick, storing and releasing freshwater that influences global sea-level over thousands of years.
Sea ice, though equally important to the climate system, is completely different. It is the thin layer (typically 1-2m) of ice that forms on the surface of the ocean when the latter is sufficiently cooled enough by the atmosphere.
From there sea ice can move with the winds and currents, continuing to grow both by freezing and through collisions (between the floes that make up the ice cover). When the atmosphere, and/or ocean is suitably warm again, such as in spring or if the sea ice has moved sufficiently towards the equator, then the sea ice melts again.
Antarctic v. Arctic
Secondly, we need to understand that the Arctic and Antarctic climate systems are very different, particularly in sea ice.
In the Arctic, sea ice forms in an ocean roughly centred on the North Pole that is surrounded by continents. A relatively large (though diminishing) proportion of the ice persists over multiple years before ultimately departing for warmer latitudes through exit points such as Fram Strait between Greenland and Svalbard.
In the south, on the other hand, sea ice forms outwards from the continental Antarctic Ice Sheet, where it is exposed to and strongly influenced by the winds and waters of the Southern Ocean. Here, there is a much stronger seasonal ebb and flow to sea ice coverage as over 80% of the sea ice area grows each autumn-winter and decays each spring-summer. This annual expansion-contraction from about 4 to 19 million square kms is one of the greatest seasonal changes on the Earth’s surface.
Area v. volume
Finally we need to remember that “extent” or “areal coverage” is only one way with which we monitor and study sea ice.
Sea ice turns out to be a very complex and variable medium that is very difficult to observe over large-scales. It is also constantly moving and restructuring. Until we achieve the “holy grail” of monitoring total sea ice volume from space and how it changes over time (and there are great steps towards this with European Space Agency’s environmental research satellite CryoSat-II), we are limited to interpreting its global behaviour through area.
What happened this winter?
This winter, the maximum total Antarctic sea ice extent was reported to be 19.47 million square kilometres, which is 3.6% above the winter average calculated from 1981 to 2010. This continues a trend that is weakly positive and remains in stark contrast to the decline in Arctic summer sea ice extent (2013 was 18% below the mean from 1981-2010).
To further complicate this picture, we find this net increase actually masks strong declines in particular regions around Antarctica, such as in the Bellingshausen Sea, which are on par or greater than those in the Arctic.
So while there is much greater attention given to the Arctic decline and the prediction of “ice-free summers” at the North Pole this century, Antarctic climate scientists still have their work cut out to understand the regional declines amidst the mild “net” expansion occurring in the southern hemisphere.
Here are some of the leading hypotheses currently being explored through a combination of satellite remote sensing, fieldwork in Antarctica and numerical model simulations – to help explain the increasing trend in overall Antarctic sea ice coverage:
-
Increased westerly winds around the Southern Ocean, linked to changes in the large-scale atmospheric circulation related to ozone depletion, will see greater northward movement of sea ice, and hence extent, of Antarctic sea ice.
-
Increased precipitation, in the form of either rain or snow, will increase the density stratification between the upper and middle layers of the Southern Ocean. This might reduce the oceanic heat transfer from relatively warm waters at below the surface layer, and therefore enhancing conditions at the surface for sea ice.
-
Similarly, a freshening of the surface layers from this precipitation would also increase the local freezing point of sea ice formation.
-
Another potential source of cooling and freshening in the upper ocean around Antarctica is increased melting of Antarctic continental ice, through ocean/ice shelf interaction and iceberg decay.
-
The observed changes in sea ice extent could be influenced by a combination of all these factors and still fall within the bounds of natural variability.
The take home messages is that while the increase in total Antarctic sea ice area is relatively minor compared to the Arctic, it masks the fact that some regions are in strong decline. Given the complex interactions of winds and currents driving patterns of sea ice variability and change in the Southern Ocean climate system, this is not unexpected.
But it is still fascinating to study.
Here, there is a much stronger seasonal ebb and flow to sea ice coverage as over 80% of the sea ice area grows each autumn-winter and decays each spring-summer.
With the recent drastic reduction in summer Arctic sea ice extent isn't this contrast now outdated? Last year the summer ice in the Arctic was below 25% of the winter extent.
Is it acceptable to claim that Antarctic sea ice is 'increasing' from zero? Prior to 1979, when NASA begins its accounting, reconstructed sea ice was much higher than after 1979. As Arctic sea ice extent has been collapsing in extent in the last 30 years, Antarctic sea ice extent mostly collapsed in the 30 years prior to 1980. In any case, that's what I'm pulling away from this article by Tamino, in which he reconstructs both Polar sea ice extents all the way back to 1880::
http://tamino.wordpress.com/2010/10/16/history-of-arctic-and-antarctic-sea-ice-part-1/
[RH] Hotlinked URL.
The other big difference between the Arctic and Antarctic has to do with albedo effect. The change in planetary albedo that comes with Arctic sea ice loss is significant, and it's something to be very concerned about. But there is almost no change in the Antarctic sea ice minimum, and thus there is little to no change in planetary albedo at the south pole, positive or negative.
Deniers keep focusing on the Antarctic sea ice maximum as if it actually means anything at all. The maximum occurs during the late southern winter when there is very little sunlight hitting the pole, and thus any increase has a very small albedo effect related to it.
Something I've never seen is a chart of planetary albedo over time.
Rob Honeycutt @3, I agree. The antarctic sea ice is essentially climate neutral, but the arctic certainly isnt. People with open minds can see the antarctic sea ice issue is a very weak sceptical argument, so people who persist with it must have ulterior motives.
On your comment about the arctic, this is a big sort of regional heating effect. I believe the northern hemisphere is also warming more than the southern. Im no climate expert and this may be a naieve comment, but as far as Im aware these temperature differentials can alter pressure systems, winds and currents.
Arent we altering virtually everything? I dont beleive you can actually fully model something so complex. I think its madness to invite changing these patterns.
I would add too, that the current behaviour of the Antarctic sea ice is influenced by the present state of the ocean-atmosphere circulation. The current negative (cool) phase of the Interdecadal Pacific Oscillation (IPO) is likely assisting the growth of sea ice, but may reverse when the IPO moves to its positive (warm) phase. This can be seen in the modelled trends from Meehl (2013):
.......and the models vs observations from the UK Met Office:
...from a non-scientist...as I understand it, water vapor is a more powerful green house gas than CO2 or methane? If so, does fresh water or highly diluted sea water evaporate faster than 'typical' sea water? The rapid artic ice melt is alarming, but doesn't some of the melt water feed back as water vapor at some point in the hydrlogic cycle?
meb58... Yes, WV is a stronger greenhouse gas but it's presense is temperature dependent. I don't believe the ice melting would be what would increase WV, but rather the rising temperature. Freshwater evaporates faster (or requires less energy to evaporate) than saltwater, but I would imagine the difference is probably minimal enough to not have a significant impact on the process.
You mention Cryosat-ll. As far as I can find, they still haven't reported the ice volume for Sept 2013 so that we can compare it with Sept 2012 and see if ice volume has also increased 60% as did ice extent. I understand there is some problem with September associated with ponding water on top of the ice but even the figures for August and October for 2012 and 2013 would give us an approximate figure to compare. Why does the ESA (European Space Agency) which operates Cryosat-ll not produce daily updates for volume as does NSIDC for extent. If the satellite is in the normal 90 min orbit, that means that it passes over the Arctic 16 times a day, each time over a different path. Surly that is enough to give a pretty good estimate of ice volume in real time.
William
Cryosphere II may not be available for 2013 but PIOMAS is:
From here:
The second largest rebound after 1996. But nowhere near 60%
Thank you for this post. Over the years, I've seen many mechanisms mentioned as possibly explaining the different trends observed between Arctic and Antarctic sea ice, but I rarely hear people mention average tropospheric temperature. According to the UAH data set, the tropospheric temperatures in the "North Pole" region from 1978-2013 increased at a rate more than 3 times the global average; while in the "South Pole" region, the tropospheric temperatures over the same time frame have a net trend of *zero*. Might this explain some of the very different longer-term behavior observed across the two poles? I'm not suggesting that the other hypotheses listed in the post aren't important. I am just wondering why tropospheric temperature never seems to get mentioned when evaluating the differences between the two poles.
#9 Glenn Tamblyn:
"The second largest rebound after 1996. But nowhere near 60%"
That's not entirely correct, Glenn.
According to PIOMAS the September sea ice volume in Arctic was 3300 km3 in 2012 and 4900 km3 in 2013, pretty close to a 50% increase, and by far the largest relative increase from one year to the next. This also tells us that the average ice thickness was more or less the same these two years.
The problem of course (for the deniers and the Arctic sea ice itself) is that the 1600 km3 of absolute increase is tiny compared to the 13,600 km3 of ice loss from 1979 to 2012. The downward trend is still crystal clear, and there is no reason to believe this "recovery" will be more than a temporary one.
Enginerd - why tropospheric temperature trends are different is interesting and discussed elsewhere. However, sea ice in Antarctica only grows a long way from pole (though still within the circumpolar circulation).
Thanks for the response, Scaddenp. I wasn't wondering why the temperature trend is different at the poles. I was wondering why the notable difference in the temperature trends might explain part of the notable difference in the sea ice trends (i.e., why Arctic sea ice is decreasing so rapidly and why Antarctic sea ice isn't). For what it's worth, UAH temperature for the "North Pole" and "South Pole" represent 60-85 degrees latitude, which is why I referred to them in quotes. I think the trend for that region is relevant.
Well some of the reasons for increasing Antarctica seaice such as ozone depletion, circum polar circulation, do operate by reducing temperature. However, whatever an average satellite tropospheric temperature shows, Antarctica is also warming around the edges (note the decrease in land ice as well the temperature trends) which does imply the increased sea ices is not because it is getting colder.
@scaddenp #14:
I'd challenge the contention that Antarctica is warming "around the edges". Maybe the peninsula is, but that's the exception to the rest of the land data from Antarctica. The manned weather stations on the coast don't show any overall warming since 1980 or so. If you have data to demonstrate warming around the edges, please direct me to it.
Klapper @15... You could try this one:
And I think scaddenp was probably referring to this chart from Zhang 2007.
And this extends Rob's GISS map back to the Zhang period.
Besides papers pointed to above, I would also add the indirect evidence of ice-loss from both GRACE and altimetry. Steig et al 2009 and the O'Donnell et al 2010 (co-author one S McIntyre) show positive warming from weather stations on the coast, so I wonder about your source for coastal weather stations showing no warming? I would also note the tropospheric trends from Screen and Symonds 2012.
However, for the matter of sea ice, it is the SST data that provides the interesting question.
#15 Klapper:
It’s definitely warming in most of western Antarctica, not only the peninsula. The trend has been about 0.5-1.0oC since 1980 (dark yellow). Eastern Antarctica has some warming and some cooling, with the overall trend there close to zero.
@Scaddenp #19 et al:
I used to keep up a database of all the manned weather stations around Antarctica (including Vostok, and Amundsen Scott), all the Russian, Australian, US, Japanese etc., and back a few years ago a majority would have shown either no warming or cooling trends over the last 30 years.
However, now with KNMI Explorer, I can extract the gridded data. Here's the header on a GISTEMP source file with grid cells between -60 and -75 latitude which I calculated a linear regression on between 1980 and 2013.
The regression slope is -0.05C/decade. In the comments above from others it appears the discussion has shifted to both sea ice and the land stations, so I took both but I think I have an option to check land only, although I'm not sure if your "around the edges" is just land or both.
UAH TLT (version 5.3) also show a slight cooling trend in the same band (-60 to -75 latitude), about 0.04C/decade.
My image in #20 disappeared. Let’s try again:
As you see, 0.5–1.0oC of warming in most of western Antarctica, while the overall trend in eastern Antarctica is close to zero.
Well I am not exactly convinced that latitudinal band is best way to way to assess the margin. As maps from same data presented here show, most of coast is warming. You get that even more so from the more sophisticated methods used in Steig 2009 and ODonnell 2010 (see their maps for 1982 - 2006).
However, this is quite far from the topic of sea ice. The expansion of the sea ice is far from land where the SSTs on this margin are warming. Increasing ice despite warming marginal seas is what this topic covers.
There was a lot of data within the article that describes possible explanations for the expansion of surface ice around Antarctica all of which are valid, however, the expansion of surface ice is not a reflection of a cooling condition around the continent. Gravity data collected from space using NASA's Grace satellite show that Antarctica has been losing more than a hundred cubic kilometers (24 cubic miles) of ice each year since 2002. Remarkably, Eastern Antarctica is showing some ice loss. I would think that the surface air temps and Ice loss are not largely related.
"I would think that the surface air temps and Ice loss are not largely related." yeah, me too! However you could argue that some ice loss (west Antarctica) is due loss of loss of buttressing ice-shelves which have be eroded by warmer waters rather than warmer air. However I note the East Antarctic ice loss corresponds with warmer air temperature. Indirectly this is a possible factor in sea ice increase due to more surface fresh water from the melt.
Disappointing article. Sorry. I expected more of the "paradox of Antarctic sea ice” being explained than a less than exhaustive list of hypotheses (plausible though some may be) and "take aways" of: "while the increase in total Antarctic sea ice area is relatively minor compared to the Arctic ...." (that's irrelevant) and ".... this is not unexpected" (when is a paradox expected?!).
Hokeith... The paradox would be merely the fact that we have increasing sea ice in the Antarctic (for now) in response to warming. Guy Williams has listed what are believed to be the primary influences causing this effect.
Not sure what you have to be disappointed about.
Concerning the trends in Antarctic sea-ice extent, it is worthwile to mention that the wind is the key driver behind its seasonal and annual variability and matter more than anomalies in temperature. Local trends in Antarctic sea-ice extent are thus primarily driven by trends in near-surface wind that either act to accelerate or hamper the seasonal growth/retreat of the sea ice. Consequently, trends in the frequencies and numbers of low and high pressure system, which steer the near-surface winds, or trends in the storm-track location and intensity are of primary importance to understand why Antarctic sea-ice appears to grow/extend in some regions.
(1) Holland, P. R., & Kwok, R. (2012). Wind-driven trends in Antarctic sea ice drift. Nature Geoscience, 5, 872–875. https://doi.org/10.1038/ngeo1627
(2) Holland, M. M., Landrum, L., Raphael, M., & Stammerjohn, S. (2017). Springtime winds drive Ross sea ice variability and change in the following autumn. Nature Communications, 8, 731. https://doi.org/10.1038/s41467-017-00820-0
(3) Holland, P. R. (2014). The seasonality of Antarctic sea ice trends. Geophysical Research Letters, 41, 4230–4237. https://doi.org/10.1002/2014GL060172
(4) Schemm, S. (2018). Regional trends in weather systems help explain Antarctic sea ice trends. Geophysical Research Letters, 45. https://doi.org/10.1029/2018GL079109