Do you want ice with that?

A popular pastime amongst those of us who have an interest in what's going on with the climate is to try to guess how much Arctic sea ice there will be each season when the annual minimum is reached.

For anyone reading who thinks this might interest them, there is an almost endless ongoing source of information, graphs and analysis at the excellent Arctic Sea Ice Blog/ Forum/Graphs.

There is also a science program which collects various predictions, from the technical (model-based) to the amateur (heuristic), called SIPN (used to be ARCUS).

Part of the fun of this is to try to work out the different factors which may or may not affect future sea ice level down the line. On a long time scale, this is easy enough, if you simply reference the climate data since 1979 and look at the history of decline in Summer sea ice levels. It is quite possible to do a 'rough fit' with historic data and come up with a prediction which is not far off the mark, though this is going to be more by chance than reason.

One of the most popular sources of data and information is the NSIDC dedicated sea ice and cryosphere pages. These focus on the measured data rather than prediction, but contain a wealth of useful graphics and links.

On the recent page you will notice a reference to a recent piece of research conducted by some of the members of the team, Serreze M. C. et. al., JGR (2016).

This paper demonstrates a connection which has an effect on the annual minimum:

"...They found that 68 percent of the variance in the date that ice retreats from the continental shelf break in the Chukchi Sea in spring can be explained by fluctuations in the April through June Bering Strait oceanic heat inflow..."


So, one way in which we can now make a more reliable prediction in relation to one part of the Arctic is in place.


Then, I noticed (not reported widely elsewhere yet) that others of the team have been involved in a different piece of research at the opposite side: Fram Strait sea ice export variability and September Arctic sea ice extent over the last 80 Years. (Smedrud et.al.). Here we find the useful:


"...Increased ice export during winter will generally result in new ice growth and contributes to thinning inside the Arctic Basin. Increased ice export during summer or spring will, in contrast, contribute directly to open water further north and a reduced summer sea ice extent through the ice–albedo feedback..."


and: "...We find a general moderate influence between export anomalies and the following September sea ice extent, explaining 18 % of the variance between 1935 and 2014, but with higher values since 2004...".


These two useful recent pieces of work add to our understanding of two of the mechanisms which contribute to Arctic Summer minima. They don't give the whole picture; the number and range of the variables are such that we don't have a full picture of all of the teleconnections or causes.


What we can do, though, is look at these two parameters and add some inferences to our other observations. If Fram export has been relatively high recently, then the September minimum is likely to be lower than the long term average. If Bering Sea heat input is higher than usual during the Spring, then the Septemberminimum is likely to be lower. 

If both cases occur simultaneously, it is reasonable to infer that there is a high likelihood of a reduced minimum, if the other other known variables (such as the Arctic oscillation - AO) do not contradict this, and if the historic patterns of weather and overall seasonal temperature anomalies are also in alignment.

This is an example of how ongoing scientific research and information collecting adds to our understanding of the ways in which the climate can vary, and the trends which show how the climate is changing relative to the past.

Sea ice, sea level and permafrost

Recent updates all over the place point out that Arctic sea ice levels are pretty low considering the time of year. Historically, the rate of decline of Winter Arctic sea ice cover (area, extent) is slower than that of the Summer (and the intermediate seasons). This year, the maximum so far (as pointed out at Neven's excellent blog, the finale is not yet concluded for certain, rather like a Tchaikovsky symphony) is below two standards deviation from the long term average. (here)

I don't subscribe to the notion that one season's extent is indicative of following seasons, for the simple reason that the range of inter-seasonal variability exceeds the range of climate variability. In other words, it isn't enough that there isn't much ice now to determine how much there will be at the end of this Summer (if only it was that easy!). But here we see the fortification (or continuation) of an existing and persistent trend, a reminder that we are apparently set into a pattern for the long term.

Meantime, Down South, recent observations (not yet officially published) are that the Larsen C ice shelf, thought to be stable when assessed in 2010, is now showing a big crack comparable to the one which preceded the loss of Larsen B some years back. Unlike some media, I won't publish more till the original is published, but you can check out the discussion paper at The Cryosphere.

Add to this the recent material illustrating the instability of the Totten Glacier in East Antarctica, the latest of a run of observations done in different parts of the EAIS showing a trend towards likely destabilisation (At least 4 papers recently on different glaciers).

Looking at a couple of other papers (I'll have to go find the sources at some point) I was caught by surprise when it was noted that the potential for sea level rise from the melting of permafrost is potentially very large (over time and relative to scale) and that permafrost is indeed melting reasonably rapidly across all Arctic landmasses (and, one supposes, potential under the ocean as well).

Which is all apropos of what?

One looks at a range of things and extrapolates a pattern. The apparent pattern is that a lot of the markers which might indicate an accelerated rate of sea level rise are all pointing in the same direction, and furthermore, that the observations of the changes are imminent, not down-the-line projections. For some time I've said, along with Rasmus Benestad and others, that the current 'mainstream' estimates of SLR are on the low side. 

I'd like to see a concerted effort to assess recent developments in observations and evidence as a whole and a review of the conclusions. IMO the odds of a 1 metre rise by the end of the century are now much shorter than they would have been twelve months ago. if a reader can point me to any material on the projected contribution of permafrost melting to SLR, that would be helpful, too.

Antarctic ice break-up update

http://www.iup.uni-bremen.de:8084/amsr2/antarctic_AMSR2_nic.png 

Update on my earlier suggestions about big lumps of ice in the Antarctic. Take a look on the right hand side of yesterday's AMSR2 image. Around 80E, 62S: I can't find the satellite imagery yet, but if we are to believe the picture, there's a HUGE chunk of ice (I don't know how thick, continuous or potentially persistent it might be) free-floating there. If this was a single entity iceberg it would be hundreds of miles across. I don't think it is likely to be so. Looking closely around the coastline nearby, more big chunks are likely to be liberated shortly.

I don't think there's any risk associated with these at the moment, I just find them interesting. There's probably precedent, too, but I'm too lazy to go check.

Call me paranoid but look anyway

It's probably just a function of an overactive imagination, but looking at the AMSR2 map of Antarctic sea ice (as one does), I get a 'feeling':

http://www.iup.uni-bremen.de:8084/amsr2/antarctic_AMSR2_nic.png

I sort of know that it isn't unusual for the sea ice to decline around the coasts a little quicker than it does in the middle of the 'pack', but it seems early in the melt season to see open water or low concentrations indicated at the level implied by the image.

Looking a little closer, a neurotic might note the correlation between the areas of lower coastal concentration and the glacier zones.

It would be useful for a Cryosphere/Antarctic specialist to call in and provide some reassurance, because right at this moment, I gotta feeling, and it's mainly anxiety.

Reasons to be cheerful, one, two, three

Back to the AR5 on the subject of the Antarctic, and in particular the Marine Ice Sheet Instability Hypothesis. The AR5 confidently asserts that 'it is likely that abrupt and irreversible ice loss from WAIS is possible'. Why does this matter? Because if it occurred, that additional sea level rise which the AR5 does not expect, of 'several tens of centimetres', would become a reality.

For non-scientists, (usual warnings apply), a 'sudden and irreversible ice loss' refers to the possible 'collapse' of the mechanism which currently constrains the export of ice into the oceans and therefore the contribution to sea level rise. This would be signalled by rapid, non-linear changes to the Pine Island Glacier (PIG) and Thwaites Glacier (TG). These two represent a substantial proportion of the potential sea-level contribution of the WAIS, which is why they are important. One of the keys to recognising that a sudden process  is under way is the retreat of the Grounding Line (GL) of the glaciers. Recent research suggests that Pine Island is more risky than Thwaites because of the relative basal geography of the two.

The process follows a theoretical sequence: The glaciers at the sea end thin, meltwater seeps down from above, the glacier fractures deep into the outlet section, then collapses; the glaciers behind speed up, pushing more ice seawards, and the grounding line (where the sea can no longer get 'under' the glacier, retreats inland. 

In Chapter 13, The AR5 discusses the condition under which such an event might occur and the consequences:

"Two processes that could trigger GL retreat are particularly relevant to contemporary polar climate change. The first is the presence of warmer ocean water under ice shelves, which leads to enhanced submarine iceshelf melt (Jacobs et al., 2011). The second is the presence of melt water ponds on the surface of the ice shelf, which can cause stress concentrations allowing fractures to penetrate the full ice-shelf thickness. This process appears to have been a primary factor in the collapse of the Larsen B Ice Shelf (LBIS) over the course of two months in 2002 (MacAyeal et al., 2003). The collapse of the LBIS provided a natural demonstration of the linkage between the structural integrity of an ice shelf and the flow of grounded ice draining into it. Following the breakup of LBIS, the speeds of the glaciers feeding the collapsed portion of the shelf increased two-to-eightfold, while the flow of glaciers draining into a surviving sector was unaltered (Rignot et al., 2004; Scambos et al., 2004; Rott et al., 2011). This indicates that a mechanical link does indeed exist between shelf and sheet, and has important implications for the future evolution of the far more significant PIG and TG systems of the WAIS."

In addition to any other, earlier material, a new paper in The Cryosphere describes Condition One:

Dutrieux, P., Vaughan, D. G., Corr, H. F. J., Jenkins, A., Holland, P. R., Joughin, I., and Fleming, A. H.: Pine Island glacier ice shelf melt distributed at kilometre scales, The Cryosphere, 7, 1543-1555, doi:10.5194/tc-7-1543-2013, 2013. There are other observations and calculations substantiating that there is water under the PIG, that it is relatively warm, that it is progressively undermining the glacier. There seems little reason to doubt that this condition is currently being met. 

Condition two requires the presence of meltwater at the top of the glacier. I can't find the source (maybe someone can help), but I am certain that I have very recently seen a record of observation of this phenomenon - I believe it was at the Thwaites Glacier, but there is no reason to suppose that the phenomenon would exist at one glacier and not the other.Forthcoming papers (not yet published) look at the 'condition' of PIG and seem to offer further evidence that the conditions exist for a full-depth fracture of the PIG.
So; it would appear that the required conditions for the fracture and subsequent collapse of the Pine Island Glacier are being met. If this is correct, the next 'critical' stage would be the appearance of a glacier-wide fracture, probably during the Austral Summer. On the surface, it very much feels like this is no longer a matter of 'if', but of 'when'.

Prediction a non-linear shift is, of course impossible. It has to be a sophisticated form of educated guesswork as to when this could occur. To me, given that other indicators suggest that the transformational process in the cryosphere are systematically understated in the 'official' lines, it could as easily be this Summer, or ten years, or twenty, thirty years away.

If any of my readers can give guidance as to the plausible timeline in which we are working here, it would be much appreciated. Likewise, if someone thinks that I am misrepresenting or overstating the case. But from the viewpoint of this blogger, that 'likely that it is possible' statement of the AR5 goes nowhere near capturing the reality of the situation or the risk. As a consequence, the decision to eliminate sea level rise estimates which incorporate a collapse in the 21st century look at best misguided, at worst, much too conservative.

I'll look out for other connections on the subject: for example, Aslak Grinsted's excellent summary of the estimates, and RealClimate's ongoing discussions.

I think this is a real, substantive risk which has a significant timeline for the current generation, and that serious and urgent investment in Antarctic research is now essential.

Sea Oddity - it's a godawful small affair

Well, RC went and stole a march by posting about chapter 13, and deservedly their material will get more attention because the contributors are specialists in the area. But whilst they discuss some technical matters, they don't seem to have covered some of the 'oddities' which, to my eye, point towards an innate conservatism in the Sea level rise assessment. Since RC has done it I will also, after all, use some of the draft, but will try to do so sparingly. Interested readers can read it for themselves.

As usual this is the 'outside' view, so I expect (and hope) for feedback.

First up is the paragraph about the two model intercomparison papers, Bindschadler et. al. (2013) and Nowicki et, al (2013), which concludes:

"The resultant projection included contributions from lubrication, marine melt and SMB-coupling and generated a mean SLR at 2100 of 162 mm over five models, or 53 mm if an outlier with anomalously high response is removed (including SMB results in SLR at 2100 of 223 and 114 mm for five- and four-model means, respectively). This comparison lends further weight to our confidence."

I don't have a problem with the concept of removing outliers as such, and it is clear from the very large differences in results that the outlier must be well out, but if they are included in the model intercomparisons, should they not be included in the assessment of the range? My beef with this is that they seem to be saying that once they have fiddled a bit with the original papers (and I'll admit, I might also be inclined to discount an outrageous outlier if there was no sensible reason to consider its' plausibility), they end up with a number which adds confidence to their 'chosen' assessment range. This doesn't seem, on the surface, to be entirely transparent.

Next up is the 'introduction' of the Marine Ice Sheet Instability material:

"There is an underlying concern that observations presage the onset of large-scale grounding line retreat in what is termed the Marine Ice Sheet Instability (MISI; Box 13.2), and much of the research assessed here attempts to understand the applicability of this theoretical concept to projected SLR from Antarctica."

Now call me a pedant (it's a reasonable charge), but the language chosen: 'this theoretical concept' at this point, whilst it may be technically correct, might predispose a reader to imagine that MISI is an implausible scenario. Later on, the Chapter includes the memorable:

"In summary, based on ice-dynamics theory, numerical simulations, and paleo records, it is likely that abrupt and irreversible ice loss from WAIS is possible. However, theoretical considerations, current observations, numerical models, and paleo records currently do not allow a quantification of the perturbation that is necessary to destabilize the ice sheet."

Can someone tell me what that first sentence means? It's a cracker.

Last up for this post is what appears to my eyes to be mildly remarkable:

"More detailed regional modelling using scenario A1B illustrates the potential for warm water to invade the ocean cavity underlying the Filchner-Ronne ice shelf in the second half of the 21st century, with an associated 20-fold increase in melt (Hellmer et al., 2012). Based on the limited literature, there is medium confidence that oceanic processes may potentially trigger further dynamical change particularly in the latter part of the 21st century, while there is also medium confidence that atmospheric change will not

affect dynamics outside of the Antarctic Peninsula."

Wow. So there is another hypothetical process that indicates a possibility of a larger potential sea level rise, which includes local changes in orders of magnitude. Does this not indicate that the reasonable range of SLR might be upped a bit? After all, we are talking about a range, aren't we?

I'm going to finish this off in another post because I am tired.

Comments, please.

Little by Little

After the last IPCC (AR4) I spent some time on the available literature as well as conversing with several qualified people (scientists) and concluded (purely on the basis of how it all seemed to add up), that the AR4 estimates of sea level rise (SLR) probably were conservative and that a rise of around a metre by 2100 was a reasonable estimate, that 1.5 metres was plausible, but that much more seemed unlikely in the timescale, given the range of evidence.

Having gone through chapter 13 of the new AR5, I am steadily moving towards the view that a metre by 2100 is possibly the lower bound of SLR, that 1.5 metres is now the reasonable estimate, and that a larger rise is not so much unlikely as uncertain, which is a very different thing.

Of course I'm not allowed to print the stuff (well..) in the draft itself, but I extracted some parts, particularly in relation to the contribution to Sea Level Rise of the Antarctic Ice Sheet and the Greenland Ice Sheet, thinking to comment as a whole. But there is too much - even reduced to the bits which seemed most relevant to the dialogues here and on other websites.

In addition, there are plenty of better qualified commentators already working on this in the public domain. Thanks to Steve Bloom, who provided these sample links (you will need to expand the comments). This is also thematically connected to the discussion of the cryosphere (Chapter 4) here.

What I wanted to draw out of the AR5 were the potential reasons why others might think that their assessment is conservative. I have come across 9 slight oddities which will be dealt with a bit at a time (otherwise it's just too much). The first one concerns the basis for estimating the SLR contribution from the Antarctic.

It looks to me as if the core statement of estimated likely SLR from this source is based on Little et. al. (2013). [NOTE - the Little paper in question is behind a paywall at Nature Climate Change, but available as a pdf via Google Scholar] Actually, that's what the report says. But then it goes on to say, apparently, that the reason for choosing this as the reference point is that it's somewhere in the middle of the available ( process model-based) assessments, not including outliers (!) , and not taking into considering semi-empirical model estimates. But this is undermined in the selfsame explanatory paragraph by stating that it falls at the lower end of the estimates from the other process model-based assessments which were available. [NOTE - the Little paper in question is behind a paywall at Nature Climate Change, but available as a pdf via Google Scholar]

It's important to remember that the people who put this stuff together for the IPCC are experienced professional scientists and they work very hard to present what is the representative view in a balanced way. Yet, throughout the chapter, there are side-comments and minor details which, put together, strongly suggest that their chosen rubric is a consciously conservative one. 

The other point which is frequently being made is that this is an example of over-dependence on the process models. There is enough evidence in the chapter to support this view, some of it overt. As things stand, I am inclined to think that not enough weight has been given to either the semi-empirical models, which consistently produce SLR rates higher than the process models, or to the people on the ground, who also seem consistently to question the AR5 estimates and suggest they are understating things.

As was pointed out to me, the most important source of difference between estimates of sea level rise is the Antarctic. Why? In broad terms, the contribution from other sources, such as the Greenland Ice Sheet, can be imagined as adding to SLR in a way comparable to the way a tap adds water to a bath, steadily. With Antarctica, this is also the case, but there also is an identifiable (but uncertain) risk that someone is planning to jump into the bath - via a destabilisation of the West Antarctic Ice Sheet (definitely more on this in a later post). The analogy also works when we think about the proposed timescales for change. There's a huge difference between adapting to a slightly overfilled bath by being careful, and clearing up after the idiot went and jumped in.

There is a lot more to be said on this subject, suffice it to say for know that I am now more concerned, after reading the report, than I was before, about the ability to adapt to some of the possible changes in the next forty-odd years, and even more worried about the following half century.

On the rocks, no ice

A swift trawl through some of the Open Access Journals has brought up this gem from The Cryosphere:

Global glacier changes: a revised assessment of committed mass losses and sampling uncertainties

S. H. Mernild^1,2, W. H. Lipscomb³, D. B. Bahr^4,5, V. Radić⁶, and M. Zemp⁷

Abstract. Most glaciers and ice caps (GIC) are out of balance with the current climate. To return to equilibrium, GIC must thin and retreat, losing additional mass and raising sea level. Because glacier observations are sparse and geographically biased, there is an undersampling problem common to all global assessments. Here, we further develop an assessment approach based on accumulation-area ratios (AAR) to estimate committed mass losses and analyze the undersampling problem. We compiled all available AAR observations for 144 GIC from 1971 to 2010, and found that most glaciers and ice caps are farther from balance than previously believed. Accounting for regional and global undersampling errors, our model suggests that GIC are committed to additional losses of 32 ± 12% of their area and 38 ± 16% of their volume if the future climate resembles the climate of the past decade. These losses imply global mean sea-level rise of 163 ± 69 mm, assuming total glacier volume of 430 mm sea-level equivalent. To reduce the large uncertainties in these projections, more long-term glacier measurements are needed in poorly sampled regions.

In layman's terms, the authors are suggesting that (accepting high uncertainty) the World's glaciers and ice caps are shrinking faster more out of balance than previously estimated, with a consequential increase in their contribution to global sea level rise. This conclusion is premised on a future climate resembling that of the last ten years (so no projection included for increasing mean annual temperatures).

I'd be interested to get Stoat's response - it's his kind of thing, as well as those of similar inclination. My first thoughts are that here is more suggestive evidence that the synthesis estimates of future change represented by the AR5 tend to underestimate likely future climate-related trends. 

The niggling worry is that there may be a cumulative imbalance - for each metric a conservative estimate, each conclusion/estimate slightly understated - and, if it proves otherwise, the gross cumulative change is far more substantial than the projections. In this case, the paper draws our attention to something I have previously researched and have an interest in - global seal level rise. There are plenty of voices in the scientific community already saying that the AR5 estimates are conservative - here is another potential string to their bow. At it, friends...

(Indeed, Stoat responds here: http://scienceblogs.com/stoat/2013/09/30/ar5-cursory-review-of-chapter-4-cryosphere-mass-balance-of-antarctica/ see comments 8&9.) Thanks, William.

Further comment: since it is a well established and credible blog, the discussion continues at WC's site - no point in duplicating effort. Please go there and read on...