20 September 2012

How British government's climate forecasting MET Office gets the Arctic wrong

Introductory note: It is now very likely that the Arctic will be sea-ice-free within a decade, with enormous consequences. Not only has the 2007 record minimum low been smashed, but the new record low 2012 sea-ice extent (3.4 million square kilometres) is less than half the figure three decades ago. And the volume of ice is now down to just one-fifth of the figure three decades ago. In 1979 the summer volume was 16,855 cubic kilometres, on 3 September this year it was just 3,407 cubic kilometres. 80 per cent of the ice has already been lost.
     The IPCC 2007 got the Arctic sea-ice wrong in projecting an Arctic still containing summer sea-ice by 2100. Now, Arctic specialists relying on new, regional climate models such as Prof. Peter Wadhams of Cambridge are making the big call of a summer ice-free Arctic as early as 2015.  By contrast, those relying for Arctic forecasts on global general circulation models (GCMs) used for the 2007 IPCC report are sticking to a 2030-2040 projection, but lament that “We just don’t know exactly why this (sea-ice loss) is moving so fast”. The predictions of those who rely too much on the GCM models seem strangely removed from the reality on the ground and even a common sense view of the evidence, as this analysis from Arctic News shows. Underestimating the speed and likely future rate of change in the climate system has deadly consequences. – David

UK MET Office keeps downplaying significance of events in the Arctic

by Sam Carana, Arctic News

One of the most respected datasets on Arctic sea-ice volume is  PIOMAS, the  Pan-Arctic Ice Ocean Modeling and Assimilation System developed at the Polar Science Center, Applied Physics Laboratory, University of Washington. The graph below shows PIOMAS data for annual minimum Arctic sea ice volume (black dots) with an exponential trend added (in red).

Note by Climate Code Red: the trend will not necessarily follow the exponential path to an ice-free state, but so far it is the "best fit".
PIOMAS Arctic sea-ice volume 1979-2011 data (black line) with exponential trend (red line) and 95% confidence interval (pink lines)
The Arctic Methane Emergency Group (AMEG), in a 12 February 2012, written submission to the UK Environmental Audit Committee, pointed out that:

 . . summer volume [is] less than 30% of its value 20 years ago. The trend in volume is such that if one extrapolates the observed rate forward in time, by following an exponential trend line, one obtains a September near-disappearance of the ice by 2015.
Note by Climate Code Red: With the 2012 melt, the volume is now just on 20 per cent of the value 30 year ago.
The MET Office, in an 8 March 2012 written submission stated:

Climate models project the Arctic will become ice-free during summer at some point this century – though likely not before 2040… In September 2007, sea ice extent reached an all-time low, raising the question of whether the sea ice is likely to melt more quickly than has been projected. There is, however, no evidence to support claims that this represents an exponential acceleration in the decline. Indeed, modelling evidence suggests that Arctic sea ice loss would be broadly reversible if the underlying warming were reversed.

Professor Slingo, Chief Scientist, MET Office, elaborated on this in her 14 March 2012 oral submission:

Q114 Chair: …when the Arctic will be ice free in summer?
Professor Slingo: Our own model would say between 2040 and 2060…

Q115 Chair: You would rule out an ice-free summer by as early as 2015, for example?
Professor Slingo: Yes we would…

Q117 Chair: In terms of the modelling that you are using, does that cover… volume of ice?
Professor Slingo: We run quite a sophisticated sea ice model… and we are looking forward now to the new measurements from CryoSat-2.

Q118 Chair: …evidence that we had suggested that the volume of ice had already declined by 75%, and that further decreases may cause an immediate collapse of ice cover.
Professor Slingo: I wouldn’t [give credence to that]. We don’t know what the thickness of ice is across the whole Arctic with any confidence… I probably would [rule it out altogether] … to say we have lost 75% of the volume is inconsistent with our assessments.

Professor Laxon, director of Centre for Polar Observation and Modelling, where CryoSat-2 data is being analysed, said in a 24 August 2012, written submission:

…[analysis of] CryoSat-2 and ICESat data… suggest a decrease in ice volume over the period 2003–12 at least as large as that simulated by PIOMAS, and possibly higher.

The Met Office, in a 31 August 2012, supplementary written submission:

The changes in observed sea-ice volume only extends [sic] over a few years and cannot in isolation be interpreted as representative of a long term trend… The extrapolation of short-term trends in ice volume is not a reliable way to predict when the Arctic will be seasonally ice free as negative feedbacks and changing weather patterns may slow the rate of ice loss… it is worth noting that climate models can show a period of recovery in ice volume following periods of large ice volume loss.

For some curious reason, some people seek to downplay the significance of the events taking place in the Arctic, as well as the risk of methane releases. AMEG added, in its above February 12, 2012 written submission:

The catastrophic risk of global warming leading to very large emissions of methane from large Arctic carbon pools, especially from sub-sea methane hydrate, is documented in the 2007 IPCC assessment.
     By collaborating with others to protect the Arctic, a climate of cooperation can be engendered to protect the whole planet for the benefit of ourselves and future generations.

Professor Lenton, in a 21 February 2012, oral submission said:

…the Hadley Centre [has] permafrost in the latest state-of-the-art model… their best estimate is we may get 0.1°C of extra warming at the end of the century from the loss of methane from the northern high latitudes.

Professor Slingo, in the above 14 March 2012, oral submission:

Q126 Dr Whitehead:what sort of modelling factors may be accounted for by the possibility of tipping points or feedback attached to these? For example, the argument that follows very substantially from the extent of continental shelf that there is within the Arctic Basin and, therefore, the particular relationship that warming on that relatively shallow sea has on trapped methane-for example, the emergence of methane plumes in that continental shelf, apparently in quite an anomalous way-leading possibly to the idea that there may be either tipping points there or catastrophic feedback mechanisms there, which could then have other effects on things, such as more stabilised caps like the Greenland ice cap and so on. I rapidly collated all the possible catastrophe theories, but I mean how are those factored into the modelling process?

Professor Slingo: …we are not looking at catastrophic releases of methane… We don’t see catastrophic change in the Arctic that would lead to catastrophic releases of methane, or very large changes in the thermohaline circulation, within the next century. Our understanding of the various feedbacks — and it is a very complex system — both through observations and modelling, suggests that we won’t see those catastrophic changes, in terms of the physical system.
Note by Climate Code Red: Sligo's position is in direct contradiction with a new research paper by MacDougall et al. which shows a “significant contribution to climate warming from the permafrost carbon feedback” of 0.4 to 0.8°C of additional warming to 2100.
Note that the above are excerpts, to make things easier to read. For the full text, click on the respective links.

Below an update of the image, produced earlier this month, with recent volume data for 2012 added. Note that the value for 2012 may still have a substantial way to go down further before reaching its 2012 minimum.

PIOMAS Arctic sea-ice volume 1979-2011 data (black line) with exponential trend (red line), 95% confidence interval (pink lines) and volume at 3 September 2012 (blue dot)
Concluding note: It's hard to read the MET Office evidence as other than scientists blindly following a poorly-performing model and willfully denying carefully accrued physical observations. If there is an issue in climate science, it's not drawing too pessimistic projections, but underplaying the problem. In this instance, the MET Office view is similar in character to the IPCC 2007 report with regard to the Arctic. Peter Wadhams told the Guardian on Monday that: "I have been predicting [the collapse of sea ice in summer months] for many years. The main cause is simply global warming: as the climate has warmed there has been less ice growth during the winter and more ice melt during the summer. At first this didn't [get] noticed; the summer ice limits slowly shrank back, at a rate which suggested that the ice would last another 50 years or so. But in the end the summer melt overtook the winter growth such that the entire ice sheet melts or breaks up during the summer months.This collapse, I predicted would occur in 2015-16 at which time the summer Arctic (August to September) would become ice-free. The final collapse towards that state is now happening and will probably be complete by those dates."  It is hard to disagree. – David