Bishop Hill Inside mathematics
Jan 10, 2014 BBC Climate: Surface The BBC's Inside Science programme looked at the Ship of Fools expedition yesterday. We were treated to the unedifying sound of presenter wondering about whether what he called "deniers'" views on sea ice extent in the Southern Hemisphere carried any weight. We were led to beleive that they did not. Normal BBC fare I hear you say.
Some of what was said was bizarre though. Interviewee Professor John Turner of the British Antarctic Survey said the rise in Antarctic sea ice extent was less than 1% and was therefore well within the bounds of natural variability. I have no problem with the second part of that statement, but 1% seemed very low to me. Here is the relevant graph from the Cryosphere Today website.
The Antactic sea ice has an annual mean of 9 million km2 or so and the current anomaly is 1.395. So we are at 14 or 15% above normal, not less than 1%. Am I missing something, or does Prof Turner have a problem with his mathematics?
The audio is attached below.
Bishop Hill
Prof Turner emails:
Based on the Bootstrap version 2 passive microwave sea ice data the annual mean Southern Hemisphere sea ice extent increased at a rate of 147 × 10**3 km**2 per decade (1.2% per decade) over 1979-2011, which is significant at p<0.01.
I think that's him agreeing that the figure he gave on the BBC was wrong, but I've emailed back asking for clarification. Remember folks, there is nothing wrong with errors if they are corrected.
Bishop Hill
The British Antarctic Survey Twitter account just pointed me to this paper.
@etzpcm Here is a paper you might find useful with 1% per decade fig: http://t.co/q4E0lcUJh2. Thanks to you and @aDissentient for contact.
— BAS_News (@BAS_News) January 10, 2014
I'll see if I can get a copy, but the tweet makes it clear that when Prof Turner said there had only been a 1% increase in sea ice, he meant 1% per decade. Moreover, the paper was submitted to the journal in January 2009 and was published a few months later. It is therefore unlikely that it includes data beyond 2007.
Here's the longer term record of the Antarctic sea ice anomaly:
As you can see, things have moved on quite a bit since that time. That said, his original email said 1.2% up to 2011, so I'm still not entirely clear.
Bishop Hill
Another email from Prof Turner
- the trend was calculated from the annual mean total Southern Hemisphere sea ice extent.
- the percentage increase was based on the average annual SIE for 1979-2011 since there is no reliable whole Antarctic sea ice data before 1979.
If you include the 2012 data the trend increases to 157 × 10**3 km**2 per decade (1.27% per decade).
2012 didn't have the largest annual mean in the record, but it did have the largest daily ice extent, which occurred on 24 September. Our paper on this event is attached. In 2013 there was an even greater extent of Antarctic sea ice in September, so many people are trying to understand the mechanism or mechanisms responsible for the changes we are seeing.
Reader Comments (119)
No EM - you are expected to do your own legwork. If you want to make claims you need to be able to support them - otherwise you are just noise.
EM -
You miss understand the point I was making, which was wrt your comment (on page 1) on sea ice extent during the Holocene Optima. I was not making any reference to sea levels, and isostatic lift is irrelevant. I will spell it out for you. There are raised beaches and ridges on the north coast of Greenland, which have been caused by wave action (not drift-ice, the geologists know the difference). Thanks to driftwood, these ridges have been accurately carbon-dated, and were formed 7000-8000 years ago. This is evidence that during the Holocene Optima, there were open seas at least in the summer, along the north coast of Greenland, where now, permanent ice sheets are the norm.
Be careful when interpreting beaches.
EM
I'm not getting excited. If you notice I said I wasn't that concerned about a 3% increase and we weren't discussing the Arctic, were we?
I don't care what Turner is "used to working with". If he is going to put himself about on the media he had better learn to speak proper. If he says that since the late 70s the increase has been 1% then people are going to take that at face value.
I don't believe he is as naive as you are painting him. He knew quite well what he said and how it would be (mis)interpreted.
Not banned yet
To the contrary. I supply evidence and you supply nothing.
Lapogus
Thank you for your support. Modern temperatures are approaching those of the Holocene optimum. If the ocean north of Greenland was ice free then, there is good reason to expect it to become ice free again soon.
Mike Jackson
Why should he underestimate the increase?
Read his papers and you will see that Professor Turner is working on the effect of climate change and the ozone hole on wind patterns. He proposes a pattern of change which would lead to increased ice extent.
If anything, it would be in his professional interest to maximise the change, not minimise it.
Nope EM - wrong again. You supply the output of a scatter gun google search rather than answer the questions about the claims you make. Sure, I could spend all evening googling to find different bits of information which may or may not support your assertions but I'd still not know for sure what datasets you are basing your claims on, what time period they cover and what are the error bounds associated with them. Without that info, it is you who has supplied nothing - other than noise.
Here are your claims again:
//
Antarctic temperatures are static. Salinity is decreasing. Storms are intensifying and winds are becoming more meridional.Take your pick.
//
So once more: Which datasets, what periods in time and what are their error bounds? Three lines should cover it.
Not banned yet
I'm not clear what you are looking for.
Do you want quotes from some authority confirming what I've told you in a single sentence? The references I gave you did so at length, with confirming evidence. This included data, and, for the peer reviewed papers, confidence limits.
You did read them?
If you want the raw data you'll probably need to contact the authors directly.
nby - "So once more: Which datasets, what periods in time and what are their error bounds? Three lines should cover it."
EM - "I'm not clear what you are looking for."
Bye, bye.
EM -
Bollocks, not on the planet I live on.
http://jonova.s3.amazonaws.com/graphs/lappi/gisp-last-10000-new.png
Add one degree to bring the Lappi graph up to date for present, and we are at the most even with the MWP, so still at least 1.5C colder than the Holocene Optima.
Lapogus
Put the modern temperature record
http://data.giss.nasa.gov/gistemp/graphs_v3/Fig.A2.gif
together with the data from Marcott et al
http://m.sciencemag.org/content/339/6124/1198.abstract
and you'll find that 1880 temperature was 0.7C below the Holocene optimum and that since then the temperature has risen by a similar amount.
All that's keeping us from Holocene optimum sea levels and ice extents is the finite time it will take for the system to catch up and
come to equilibrium.
EM - so you are having to use Hansen's GIS and 'oh right up outta the elevator' Marcott and Shakun paper. I did not realise you were so gullible and beyond help.
Marcott method
Gissian statistics & data adjustments
I note that Marcott et al has had 31 citations in less than a year. It has rapidly become the standard Holocene temperature record. Since it shows the same temperature pattern as your own source, I fail to see your objection.
Pardon my cynicism, but it would be more accurate to say that many people are flumoxed by yet another model prediction failure and are trying to find a way to pin it on AGW
EM
That there are 31 citations for Marcott et al does not surprise, this is post-normal climate science after all. Here's what the authors belatedly posted on Real Climate, long after all the headlines and media coverage of their way up outta the elevator hockeystick':
.Fixing the Marcott Mess in Climate Science, Roger Pielke, Jr.
lapogus
We're not talking about the 20th century data, which Marcott et al themselves regarded as statistically less significant than the rest as fewer proxies were available.. We're talking about the Holocene optimum, which was much more soundly based.
I was fascinated by the way all the spin-sceptics were trying to discredit the minor part of the paper, the 20th century data which cross-validated the instrument record, and ignoring the rest.
It's a little hard to tell since your graph lacked a proper temperature scale, but Marcott et al's curve of Holocene temperatures was a reasonable match to the green regression curve. Why are you complaining about it?
EM - agree that there is a reasonable match between the Lappi and Marcott et al curves (pre-1900). But the inclusion of the high resolution post 1900 data to give the super hockey-stick blade, and all the spin and hype they generated from it tarnishes the whole paper. And you cited it without any qualification of the 20th Century uptick, and now accuse sceptics of being the spinners! If late 20th temperatures are almost as high as they were in the Holocene Optima, as you asserted, you will have to find better evidence for it than Marcott et al and a dodgdy GISS graph.
A brief postscript.
http://m.pnas.org/content/107/34/14987
This is a paper coauthored by Dr Judith Curry. It proposes a mechanism by which global warming produces increased Antarctic sea ice extent.