Arguments
Is Antarctica losing or gaining ice?
What the science says...
| Select a level... |
Basic
|
Intermediate
| |||
|
Antarctic sea ice extent has expanded at times but is currently (2023) low. In contrast, Antarctica is losing land ice at an accelerating rate and that has serious implications for sea level rise. |
|||||
Climate Myth...
Antarctica is gaining ice
"[Ice] is expanding in much of Antarctica, contrary to the widespread public belief that global warming is melting the continental ice cap." (Greg Roberts, The Australian)
At a glance
Who discovered the great, South Pole-straddling continent of Antarctica? According to the National Geographic, Captain Cook came within an estimated 80 miles of it in the late 1700s, but the three first 'official' discoveries all took place in 1820 by Russian, British and American teams of seafarers respectively.
Since that initial discovery, Antarctica has attracted and inspired researchers and explorers alike. It's a challenging place, fringed by sea-ice that, unlike the Arctic, has not steadily declined but whose extent fluctuates on a seasonal basis: it's currently (February 2023) at a very low coverage, but it can and does recover from such dips. Antarctic sea-ice is no great problem, with the exception of albedo-loss in low extent years: if it all melted, it would have no effect on global sea-levels. It's the stuff on land we need to focus upon.
The land of Antarctica is a continent in two parts, divided by the 2,000 m high Transantarctic Mountains. The two parts differ in so many respects that they need to be considered separately. East Antarctica, that includes the South Pole, has the far greater landmass out of the two, some 4,000 by 2,500 kilometres in size. Although its massive ice-sheet, mostly grounded above sea level, would cause 52 metres of sea level rise if it completely melted, so far it has remained relatively stable. Snow accumulation seems to be keeping in step with any peripheral melting.
In contrast, in the absence of ice, West Antarctica would consist of islands of various sizes plus the West Antarctic Peninsula, a long mountainous arm pointing northwards towards the tip of South America. The ice sheet overlying this mixed topography is therefore grounded below sea level in many places and that's what makes it far more prone to melting as the oceans warm up. Currently, the ice-sheet is buttressed by the huge ice-shelves that surround it, extending out to sea. These slow down the glaciers that drain the ice-sheet seawards.
The risk in West Antarctica is that these shelves will break up and then there will be nothing to hold back those glaciers. This has already happened along the West Antarctic Peninsula: in 1998-2002 much of the Larsen B ice-shelf collapsed. On Western Antarctica's west coast, the ice-sheet buttressing the Thwaites Glacier – a huge body of ice with a similar surface area to the UK - is a major cause for concern. The glacier, grounded 1,000 metres below sea level, is retreating quickly. If it all melted, that would raise global sea levels by 65 centimetres.
Such processes are happening right now and may not be stoppable - they certainly will not be if our CO2 emissions continue apace. But there’s another number to consider: 615 ppm. That is the CO2 level beneath which East Antarctica’s main ice sheet behaves in a mostly stable fashion. Go above that figure and the opposite occurs - major instability. And through our emissions, we’ve gone more than a third of the way there (320 to 420 ppm) since 1965. If we don’t curb those emissions, we’ll cross that line in well under a century.
Please use this form to provide feedback about this new "At a glance" section. Read a more technical version below or dig deeper via the tabs above!
Further details
Arguments that we needn't worry about loss of ice in the Antarctic because sea ice is growing or even that sea ice in the Antarctic disproves that global warming is a real concern hinge on confusion about differences between sea and land ice, and what our best information about Antarctic ice tells us.
As well, the trend in Antarctic sea ice is not a permanent feature, as we'll see. But let's look at the main issues first.
- Sea ice doesn't play a role in sea level rise or fall.
- Melting land ice contributes to sea level rise.
- The net, total behavior of all ice in the Antarctic is causing a significant and accelerating rise in sea level.
Antarctic sea ice is ice which forms in salt water mostly during winter months. When sea ice melts, sea level does not change.
Antarctic land ice is the ice which has accumulated over thousands of years in Antarctica by snowfall. This land ice is stored ocean water that once fell as precipitation. When this ice melts, the resulting water returns to the ocean, raising sea level.
What's up with Antarctic sea ice?
At both poles, sea ice grows and shrinks on an annual basis. While the maximum amount of cover varies from year to year, there is no effect on sea level due to this cyclic process.
Figure 1: Coverage of sea ice in both the Arctic (Top) and Antarctica (Bottom) for both summer minimums and winter maximums. Source: National Snow and Ice Data Center
Trends in Antarctic sea ice are easily deceptive. For many years, Antarctic sea was increasing overall, but that shows signs of changing as ice extent has sharply declined more recently. Meanwhile, what's the relationship of sea ice to our activities? Ironically, plausible reasons for change may be of our own making:
- Ozone levels over Antarctica have dropped causing stratospheric cooling and increasing winds which lead to more areas of open water that can be frozen (Gillett & Thompson 2003, Thompson & Solomon 2002, Turner et al. 2009).
- The Southern Ocean is freshening because of increased rain and snowfall as well as an increase in meltwater coming from the edges of Antarctica's land ice (Zhang 2007, Bintanja et al. 2013). Together, these change the composition of the different layers in the ocean there causing less mixing between warm and cold layers and thus less melted sea and coastal land ice.
Against those factors, we continue to search for final answers to why certain areas of Antarctic sea ice grew over the past few decades (Turner et al. 2015).
More lately, sea ice in southern latitudes has shown a precipitous year-on-year decline (Parkinson 2019). While there's a remaining net increase in annual high point sea ice, the total increase has been sharply reduced and continues to decline.
How is Antarctic land ice doing?
We've seen that Antarctic sea ice is irrelevant to the main problem we're facing with overall loss of ice in the Antarctic: rising sea level. That leaves land ice to consider.
Figure 2: Total Antarctic land ice changes and approximate sea level contributions using a combination of different measurement techniques (IMBIE, 2017). Shaded areas represent measurement uncertainty.
Estimates of recent changes in Antarctic land ice (Figure 2) show an increasing contribution to sea level. Between 1992 and 2017, the Antarctic Ice Sheets overall lost 2,720 giga-tonnes (Gt) or 2,720,000,000,000 tonnes into the oceans, at an average rate of 108 Gt per year (Gt/yr). Because a reduction in mass of 360 Gt/year represents an annual global-average sea level rise of 1 mm, these estimates equate to an increase in global-average sea levels by 0.3 mm/yr.
There is variation between regions within Antarctica as can be seen in Figure 2. The West Antarctic Ice Sheet and the Antarctic Peninsula Ice Sheet are losing a lot of ice mass, at an overall increasing rate. The East Antarctic Ice Sheet has grown slightly over the period shown. The net result is a massive loss of ice. However, under a high-emissions scenario, ice-loss from the East Antarctic ice-sheet is expected to be a much greater in the decades after 2100, as reported recently by Stokes et al. (2022). That’s a scenario we must avoid at all costs.
Takeaway
Independent data from multiple measurement techniques (explained here) show the same thing: Antarctica is losing land ice as a whole and these losses are accelerating. Meanwhile, Antarctic sea ice is irrelevant to what's important about Antarctic ice in general.
Last updated on 14 February 2023 by John Mason. View Archives
Eclectic @525,
I don't think it is correct to say that Swally's Antarctic ice loss hypothesis is "now obsolete." It is glaceologists conundrum which implies a level of ice loss from Antarctic Ice Sheets that are certainly outside the levels calculated by more conventional methods, thus an "outlier." But surely the conundrum has yet to be resolved, thus Zwally et al (2021).
My apologies, MA Rodger, for using the word "obsolete" , in describing the Zwally 2015 paper. I am guilty of achieving more brevity than wit.
And thank you for the Zwally (et al.) 2021 paper , illustrating the academic bun-fight re mass balance (too many buns for me to wish to digest).
However, it was amusing to see his wording: "... disparate estimates [of about 100 Gigatonnes per annum] of the [East Antarctic] mass balance have not been properly resolved."
Overall, my suspicious mind detected a whiff of rear-guard justification, with Zwally attempting to juggle isostatic changes / altimetric measurements / gravimetric measurements.
For me, the take-home message from all the Antarctic studies, is that the totality of the ice-sheet is in approximate stasis (plus or minus a few decades). And most importantly, Antarctica supplies no support for science-denialists who wish to claim that the south polar situation disproves the mainstream climate science.
[ Note to Moderator ]
Strangely, the post #526 by MA Rodger has been jumped from page 21 (which would normally host 50 comments) onto a new page [22].
Is the SkS long-suffering volunteer coder able to correct this? And if I may also point out ~ clicking on the latest comment for this thread (in the standard heading comments section) produces a jump to page 11, rather than to the latest page.
[ Note to Moderator ]
I must take back part of my above comment ~ I see on other threads that each page of comments holds only 25 comments. Must have been dozing at my keyboard all this time . . . I had remembered 50 comments as the standard size page.
Nevertheless, can the coding be altered to deliver the latest page, when clicking on the latest comment on the general Comments section? Thanks.
[BL] What's the difference between 25 and 50, between friends?
I, too, have noticed the odd behaviour of following links from the Recent Comments page to the long comments sections of some blog posts. In this case, clicking on your most recent post gets me to page 11 (claiming to be out of 20), when your post is actually on page 22. I need to go to page 13 or later before page 22 is listed as an option.
I'll bring this to the attention of our more technical volunteers.
Thank you , Bob.
A coffee + diligent searching, has shown me that not all SkS pages rollover at just 25 comments. But your calm philosophical attitude points me toward taking a more relaxed view of this "existential" threat posed by page irregularities. (When did existential become the must-use word for all politicians' discussions? )
Oh, and please consider deleting all these off-topic comments here.
Thanks, BL.
Please note: the basic version of this rebuttal has been updated on Feb 14, 2023 and now includes an "at a glance“ section at the top. To learn more about these updates and how you can help with evaluating their effectiveness, please check out the accompanying blog post @
https://sks.to/at-a-glance
Thanks - the Skeptical Science Team.
It looks like the Antartican Ice Sheet had a very good year, as far as we can see. At least, the mass balance over the period november 2021 - november 2022 was far positive. This can be due to the very low extend of the sea ice. The Surface Mass Balance over the melting period of last year turned out very positive. I don't read much about this, the focus in de media is on the low extent of the sea ice. Any thoughts about this?
I did expect a new update of de GRACE data of December 2022, but it comes late again.
Bart Vreeken @533:
What exactly is your point? The links between sea ice area and land ice mass are not simple, and have been discussed in the detailed sections of the blog post and earlier comments.
In your graphic, it is obvious that the two major years of land ice gain (2016, 2022) follow several years of strong mass ice loss. This is easily explained as a rebound effect.
This web page on Grace data has an embedded video with data to 2020. Rather than portraying the annual changes it shows the overall trend in the absolute value from year to year. Clearly, Antarctic land ice is losing mass in the long term - with short terms ups and downs.
Are you perhaps over-analyzing the significance of short-term changes, as is often done with temperature changes? (As seen in The Escalator).
Bart Vreeken @533,
You appear to be plotting out the GRACE/GRACE-FO data as per this NASA web page (which shows data to Oct 2022). This gravity data does not measure Sea Ice which is floating. And for clarity, it is not Surface Mass Balance which you correctly say had an exceptional year last year (as per this NSIDC post of January 2023, snowfall being high enough to "completely offset recent net ice losses from faster ice flow off the ice sheet for this assessment period. Most of the past decade has seen annual net losses of 50 to 150 billion tons."
So a record year for the 2023 Antarctic Sea Ice Extent minimum as well as a record year for the 2022 Antarctic Surface Mass Balance.
Antarctica doesn't get a lot of attention, compared to the Arctic cryosphere. Certainly for Antarctic Sea Ice, the mechanisms driving the variations is a lot less straightforward in the Antarctic.
One other thing to note in Bart Vreeken's graph in comment 533.
Bart's graph has a dotted line in it, and an equation in the upper right corner, which I presume is the regression equation for the line.
Note that the indicated slope is -0.0672 (so, almost zero), and the intercept is -113.83. This indicates an average mass loss of 113.83 Gt per year. A bit lower than the 149 Gt per year in the web page I linked to in comment 534 - but the page I linked to does not include the 2021 and 2022 data, which pull the average up.
Hi Bob @534
I don't see a clear rebound effect in my figure.
And of course the mass gain of last year shall be exceptional. But at least it's an interesting thing to notice. And maybe the increasing precipitation can offset the increasing discharge in the coming years as we can read in the article below. As you say, the average mass loss is now something like 114 Gt per year. That's much less then the 151 Gt we read about on the website of NASA (Vital Signs).
tc.copernicus.org/articles/16/4053/2022/tc-16-4053-2022.pdf
The difference between the 114Gt/y and the 151Gt/y isvery likely that the former is an average and the latter a linear rate of loss calculated using OLS. That NASA (Vital Signs) graph showing the 151Gt/y also gives a 20.5-year mass loss Apr2002-Nov2022 of 2,352Gt which averages to 115Gt/y.
Bart Vreeken @ 537:
Ahh, so you are focussing on the noise, not the signal. The year-to-year variation is large, and one year of adding ice does not a trend make. So, there has been a year where close 400 Gt was added - and there have been four years in the record you show where 300-400 Gt per year was lost (2007, 2010, 2015, 2018).
The "rebound" of 2022 is simply a strong positive departure from the long-term trend after several years where individual years were below the long-term trend. You might benefit from reading about Regression toward the Mean.
In a noisy, short data set, adding or removing one exceptional year will have a strong effect on the average. With only 20 years of record, +400Gt in one year will shift the average by 20 Gt. This is not exactly ground-breaking analysis.
And in continental-scale ice sheets, it takes decades to centuries for mass added in the central areas to reach the perimeter. You are familiar with the phrase "glacially slow", aren't you?
So again, exactly what is your point, other than "this is interesting"? And what, exactly, do you want me to read about in the paper you link to? Refer to a diagram, section of text, or something concrete.
MAR @ 538:
Yes, the linear regression vs. "average" is a factor. The page I linked to shows a graph that starts at 0 ice loss in 2002 and ends with the current accumulated ice loss. It points out two specific values:
...for a total change of 2873 Gt in 19.3 years. That is a change of 149 Gt per year. It is not a regression of the time series. It also takes the maximum of 65 Gt at a time slightly after the record begins, and chooses the minimum in February 2020, not the final value at the latest point on the series.
As I said - noise in the system. Short record, year-to-year variability. Add close to a year at the start; add a couple of years at the end; results change. Not a surprise.
Bob Loblaw @ 539
I don't think focussing on the noise is a good idea ;-)
When we just look at the GRACE data then the year 2022 is one in a row of 20. After last year the average has changed, which is interesting by itself.
But during this 20 years things have changed. The extent of the sea ice has declined. So more water vapour comes to the continent, which gives more precipitation. The question is how this works out. The last year gave us a hint that it can add a lot to the Surface Mass Balance. The next question is: how much will the SMB increase, and how much will the discharge increase. Of course, that's very uncertain. In the paper I called it says in the abstract:
The surface mass balance in SSP5–8.5 simulations shows a pattern of strong decrease on ice shelves, caused by increased melting, and strong increase on grounded ice, caused by increased snowfall. Despite strong surface and basal melting of the ice shelves, increased snowfall dominates the mass budget of the grounded ice, leading to an ensemble mean Antarctic contribution to global mean sea level of a fall of 22 mm by 2100 in the SSP5–8.5 scenario. We hypothesise that this signal would revert to sea-level rise on longer timescales, caused by the ice sheet dynamic response to ice shelf thinning. These results demonstrate the need for fully coupled ice–climate models in reducing the substantial uncertainty in sea-level rise from the Antarctic Ice Sheet.
Bart @541... "After last year the average has changed, which is interesting by itself."
Again here, you're focusing on one data point, where the abstract you post is focused on the long term uncertainties. The 2022 datapoint may technically alter the 20 average, but that's pretty darned meaningless since the following years may likely revert to the long term mean.
What researchers are trying to do (to my understanding) is reduce their uncertainties for long term ice mass loss as it pertains to sea level rise contributions. Having a higher degree of confidence on whether we're going to see 0.3m or 1.3m of sea level rise by 2100 is very important information for the broad purposes of governments and societies to inform them how to prepare.
Bart @ 541:
What Rob said to you.
You say the change in the average is "interesting". So what? The average went up a lot when 2022 gets added to the series.
This is simple arithmetic. Short time series see big jumps in the average when a single large value is added. It really has very little meaning.
You then go on to postulate "how much will the SMB increase"? You are speculating that this one-year large positive value is the key to the future trend. That is highly speculative. Not just uncertain - highly speculative. You are taking one value from a noisy signal, and treating it as if it represents a long-term trend.
You are focussing on the noise represented in a single value, and it really is not a good idea.
Hi Bob, the increase of the SMB of Antarctica is not so very speculative. It started last century, this NASA study says:
climate.nasa.gov/news/2836/antarcticas-contribution-to-sea-level-rise-was-mitigated-by-snowfall/
When we look at the last seven years then five of them were above average; four of them were far above average and none of them were far below average. So, it's not only last year.
Most of the uncertainty is in the expected discharge, I think.
Bart @ 544:
You are really presenting a scrambled set of statements that lack clarity and consistency.
The link you provide does not mention any increase in SMB - it discusses small increases in snowfall, and how this has made the decreases in the SMB less than they would have been otherwise. The opening of the second paragraph is [emphasis added]:
The second-last paragraph says [emphasis added]:
So, your reference provides no support for your claim that the 2022 increase in SMB "started last century". Snowfall is not SMB - it is only part of it. Stop jumping from one measure to another, as if they are equivalent.
When you refer to "the last seven years then five of them were above average; four of them were far above average and none of them were far below average" you completely fail to tell us what "them" are. The article you link to provides no annual numbers for anything. This description does not appear to be consistent with the diagram you presented originally in comment 533, and I have no idea what data set you are talking about.
You appear to be taking small bits from articles that you read, misunderstanding what they say, and interpreting them (incorrectly) as evidence that supports your position.
Bob @545
I expected a more serious discussion here.
Of course, snowfall is not SMB. There is also surface melting, runoff, wind blow, evaporation. In the figure i posted above you can see the difference between the SMB and the snowfall (dashed line). But of course, there is a big corralation between SMB and snowfall.
And SMB is not the same as the total Mass Balance. I never said the the mass loss has stopped. (OK, last year incidentely).
The SMB of the last seven years is showed in the figure I posted above. Source:
www.climato.uliege.be/cms/c_5652669/fr/climato-antarctica
As you can see, the SMB of season 2022-2023 ended ca 310 Gt above average. And so on.
And yes, its not completly consistent with the diagram in comment 533. The diagram shows the mass change between 2021/11/14 and 14 2022/11/14, based on gravimetry. The SMB is calculated over 2022/03/01 until 2023/03/01 based on weather models.
With a close look to the SMB figure you can also derive a SMB over the same period as the GRACE data.
As you can see above, I also looked for a correlation between the sea ice extent and the SMB in the last decennia. With some cut and paste I made a combination of two figures. The SMB is calculated over March to February in the next year. So the peak in the SMB in 1992 comes together with a low minimum in 1993, and the low SMB in 1994 comes with a high minimum sea ice extent in 1995. Interesting, isn't it?
Bart @547... "Interesting, isn't it?"
Perhaps you can explain why you think this is interesting.
Rob @548
"Perhaps you can explain why you think this is interesting"
The interesting thing is that there seems to be a correlation between years with low sea ice and years with a high SMB. So, when the amount if sea ice stays low, we can expect more years with a high SMB. As long as most of the precipitation keeps falling as snow, not as rain.
Of course, the low amount of sea ice can have other effects too. The calving and the melting of the ice shelfs shall also increase, and with that the speed of the glaciers.
Bart @ 546, 547, and 549:
When I am having a serious discussion with someone in the comments section here, I expect certain things:
You've jumped from Grace data showing total mass, to links to papers discussing snowfall changes, to Surface Mass Balance, and back again - and it is all jumbled together in an incoherent mess.
In 546, you state, "I never said the the [sic] mass loss has stopped. (OK, last year incidentely) [sic]."
And now the primary evidence from your original comment (the 2022 Grace data) is dropped as if you just mentioned it "incidentely" [sic] and never meant it to be a claim that the mass loss had stopped?
I expected an honest discussion here, not a game of "Look, squirrel". And in 549 your response to Rob's request to explain what is "interesting" is basically a hand-waving speculation of maybes. If you are posting maybes so that you can backtrack and say things like "I never said the the [sic]mass loss has stopped", when that was the obvious implication of what you said, then it is impossible to have a serious discussion with you.