Bishop Hill More parliamentary statistics
Jun 11, 2013 Climate: Parliament Climate: Statistics Climate: Surface Parliamentarians still seem to be showing an admirable interest in the nitty gritty of statistics as applied in the climate change field. Here's a question and answer exchange between Peter Lilley and Greg Barker:
Lilley: To ask the Secretary of State for Energy and Climate Change what assessment his Department has made of whether the decrease in the extent of Arctic sea ice since reliable records began is statistically significant; and what statistical model his Department has used to conduct that assessment.
Barker: The Department has not commissioned any assessment of the statistical significance of long-term trends in Arctic sea-ice extent. Work undertaken under the Climate Programme at the Met Office Hadley Centre has assessed the physical reasons for the decrease in ice extent and used physically-based climate models to assess its future course ('Assessment of possibility and impact of rapid climate change in the Arctic':
We note that the downward trend in Arctic sea-ice extent, taking account of the seasonal cycle, is now well-established from satellite observations since 1979 and has been reported as being statistically significant in the peer-reviewed scientific literature.
Given what we know about the amount of checking that doesn't go on in academic studies these days, particularly climate change, for policymakers to rely on the scientific literature is foolish in the extreme. In fact one could go so far as to characterise it as negligence.
Reader Comments (57)
I fail to understand why a supposedly well informed MP such as the right honourable Mr Lilley excludes Antarctica from his question. It would be interesting to get an answer from Barker/MET about the statistical significance of the increase in sea ice down there since 1979.
C'mon JP, one step at a time. Cover too much and you give the 'experts' even more wriggle room than they will take anyway. The Arctic is important because it is so often cited as a reason for ongoing alarm - since thermometers became some of the worst deniers on the planet. It will be interesting to see how this plays out. And Lilley's background in Nat Sci has to be help.
The DMI has sea ice doing rather well this year. Makes the reason for Lilley's question intriguing.
If climate science is anything like medical science, then relying on scientific literature could be considered to be criminal negligence!
I think the most important parliamentary statistic is that 267 MPs voted to decarbonise electricity production by 2030. Fortunately they were narrowly defeated but 267 is a powerful bunch of lunatics.
The Met Office has given a reasonably straight answer here. That is major progress, and we should be very glad for that.
The statement that “the downward trend in Arctic sea-ice extent, taking account of the seasonal cycle, is now well-established from satellite observations since 1979 and has been reported as being statistically significant in the peer-reviewed scientific literature” is true. Unfortunately, almost all of that literature uses a straight line with iid [independent identically-distributed] Gaussian noise.
That implies that the analyses ignore that what happens in one year will have some effect on what happens the next year. Simply put, those peer-reviewed analyses are garbage. More generally, there is no competent statistical analysis that would conclude that the decrease in sea-ice extent is significant. Even so, it seems fair for the answer to mention the peer-reviewed literature.
'Barking mad' hasn't a clue what his response means.
So if the arctic ice decline is linear, when does the ice area go negative? And what does that even mean?
"Climate models do not predict that rapid changes will occur in the near horizon but they
are capable of producing periods of rapid decline in ice extent, as well as slowdowns in ice loss." (MO)
Climate models can do that..?
jamesp, sure they can, if you beleive that the model is more valid than reality, as appears to be the norm for Climate Scientists.
From the Alarmist Handbook -- "How to win a debate in Climate Science". When your opponent argues statistics, you counter with physics and when your opponent argues physics you counter with statistics.
mpaul: Very good. You've rumbled 'em.
As far as I know, nothing extraordinary has happened to any of the usual meteorological and sea state observations in recent decades. There is much that is poorly measured and poorly understood in the climate system, and the pursuit of dealing with that may well have been hampered, not helped, by the shameful fuss generated over a trace gas with levels that were if anything uncomfortably close to those below which plantlife starts failing.
In other words, absent the scaremongering, we would have been pleased to see the levels rising, and perhaps modified global observation systems a little in case one day sufficient good data might be accumulated to help detect the small effects that might be expected from that trace gas. But the main effort of observation and analysis would have been on the airborne water cycle, the behaviour of the oceans, and perhaps increasingly on the impact of cosmic ray variations on cloud formation at various heights. Computer models of the entire system would have remained where they still belong, and that is as an esoteric sideshow in academia where the serious modelling efforts would have been on modest subsets of the system - modest enough for tractability and improved scope for testable theories.
"When your opponent argues statistics, you counter with physics "
Okay so how much has the sea level gone up ?
Their answer is either none or dont know.
Perhaps the question should be how much is their uncertainty significant ?
"So if the arctic ice decline is linear, when does the ice area go negative?"
Immediately after it hits zero. Around 2015 according to the Grauniad.
"And what does that even mean?"
area = width × width
If area is -ve, it means that width is an imaginary number.
Jun 11, 2013 at 2:18 PM | Richard Drake
'...since thermometers became some of the worst deniers on the planet...'
Sir, request permission to use that marvellous turn of phrase in polite conversation.
But the Arctic is of no great concern. A complete ice free Arctic would not in itself contribute anything to sea level rise.
From a climatic point of view, what is there not to like about an ice free Arctic since the heat loss to space (the ice lid having been removed thereby enabling oceanic heat to escape) would more than make up for any warming due to albedo change.
From a commercial view, an ice free Arctic has numerous benefits, and would reduce pollution by reducing ocean steaming on a number of routes.
We should be more concerned about the Antarctic since that could contribute towards sea level rise 9as well changing ocean currents), however, there the ice is growing not declining.
Just imagine how much more vociferous the warmists would be if the situation was reversed and it was Antarctic ice in decline and Arctic ice slightly growing. The shreiks of we are all doomed because of unprecedented sea level rise would be deafening.
Richard Verney: Very helpful. Though would it shock you to hear that I'm a sceptic about our knowledge of the full impact on all of humanity of an ice-free Arctic (and I take it that means for some months during the summer in any case - how many weeks or months is surely an additional piece of precision worth having). And that's not to say we shouldn't do our best.
In any event, and partly for this reason, I think Peter Lilley's much more limited question is a good one. As the Bish intimates, this is becoming a useful habit. Perhaps the worldwide denialist conspiracy we've heard so much about has finally achieved an belated modicum of organisation and expertise behind the scenes :)
Arctic sea ice, bejeepers that old canard...... ha ha - there really is nothing to see here. Though, Lilley is making mischief and in quite a nuanced way - with submarine torpedoes speeding towards and aimed right at the Met Office's exposed starboard side.
I feel a slam dunk from Lilley coming on. You cannot force the Met Office out from under their rock, they simply ignore you; but, you can force a reply from a Minister and if that Minister has to consult the Met Office for the reply there is no escape for Slingo et Al. The Dept has to exercise due diligence and I don't think they are going to get away with a waffle.
@ Jun 11, 2013 at 4:50 PM | John Shade ...
+1 !
'Given what we know about the amount of checking that doesn't go on in academic studies these days, particularly climate change'
It was Phil Jones that blew the gaff on this one.
'The most startling observation came when he was asked how often scientists reviewing his papers for probity before publication asked to see details of his raw data, methodology and computer codes. "They've never asked," he said.'
http://www.guardian.co.uk/environment/cif-green/2010/mar/01/phil-jones-commons-emails-inquiry
Phil Jones has published over 200 'scientific' papers. Nobody has ever checked his work. Not once.
And he is not noted as a man for whom organisation and method are watchwords.......the exact opposite seem to be his guiding principles. How likely is it that he has been perfectly infallible for over 30 years?
From what I can see recently there has been a trend for a reduction in Arctic ice extent in the summer in recent years. However this doesn't seem to make any significant difference to ice extent in the winter which has stayed much the same for at least a decade if not longer. Does this perhaps indicate a trend to warmer summer weather locally or is some other phenomenon responsible?
Does Jones actually do anything "scientific"? All his work I have seen mentioned in the climate context seems to have been analysis of other people's data.
Martin, the ice extend is a surface measure. It does not tell much about the thickness.
The Arctic ice thickness and volume are decreasing over the years.
Unfortunately even Arctic ice is not now following the AGW agenda
http://arctic-roos.org/observations/satellite-data/sea-ice/observation_images/ssmi1_ice_area.png
How to explain that Arctic sea ice formation in the winter (measured by extent) has substantially increased while thickness, volume and summer minimum have all substantially decreased. And while Antarctic sea ice continues its steady increase in all seasons?
Answer: Embedded black carbon, probably coupled with a small decrease in clouds, likely due to reduced aerosols (particularly from Russia). While embedded BC disproportionately melts older thicker ice, it doesn't affect ice formation and is almost completely absent from the Antarctic.
Apologies if doubled posted my wifi is a bit flakey.
I suggest you all read Climate Dialogue's recent discussion on the significance of long-term trends.
http://www.climatedialogue.org/long-term-persistence-and-trend-significance/
Entropic Man,
Your reference seems to be but a repackaging of IPCC reports.
Philip, can Antarctic sea ice really be increasing in "all seasons". The figures at WUWT seem to show it is nearly all gone in summer and no apparent trend at that time of the year. Also do you have a reference for the disproportionate melting of older ice - it is very interesting. Thanks.
An oddity of the Arctic, of earth really, one pole is land the other a hole, a basin surrounded by land.
Ice extent bumps it's head against land, it can only grow selectively.
Abdumassatov has pointed out that on the basis of solar and Russian experimental knowledge, the Arctic will be as cold in 2020 as it was in 1900.
Russian climatologists are healthily sceptical about Western Climate Alchemy, which is based ultimately on Sagan's incorrect aerosol optical physics so that part of global dimming is really global brightening! My view is that CO2-AGW is zero because CO2 used as the working fluid in a radiative heat engine, but only engineers can work this sort of stuff out.....:o)
Helen,
Unfortunately the link from WUWT's is broken so I'll have to link to the whole page.
http://wattsupwiththat.com/reference-pages/sea-ice-page/
See SH sea ice area.
While the increasing trend for the summer minimum is not as clear as the trend for the maximum, you can see there is an increasing frequency of minimums above 2 million sq km.
Nasa uses ice area and presents summer data here. Post 2000 summer minimum is on average roughly 350,000 sq km higher than pre 2000 average.
This is Nasa on faster melt of older thicker ice.
http://www.nasa.gov/topics/earth/features/thick-melt.html
Oops, forgot the 2nd link
http://earthobservatory.nasa.gov/Features/WorldOfChange/sea_ice_south.php
@entropic
Give us the highlights first, then we can decide whether it is worth reading in its entirety.
Getting back to Peter Lilley, see this exchange with Bryony Worthington. It's depressing that she seems unable to even acknowledge the possibility of a potential capacity shortfall and considers it helpful to trot out discredited green clichés: fossil fuel subsidies, "97% of scientists" ...
"Give us the highlights first, then we can decide whether it is worth reading in its entirety."
If I may say so LA that is a question from a man whose technique have been honed to razor sharpness by many years at the bar. You know he hasn't read it, you just want to show the jury! Excellent.
As for Arctic Sea Ice extent, it has always varied, it's only since 1979 we've been observing it, and as it happens during that period the Great Global Warming Scam emerged, giving the scammers the opportunity to point to melting sea ice, and loss of sea ice density as "proof" of global warming.
There is a problem though. The actual Arctic temperature hasn't gone up, at least since the DMI have been measuring it, that's around 55 years ago. The summer temperature haven't exceeded 275K in the entire period.
<link>http://ocean.dmi.dk/arctic/meant80n.uk.php</link>
@ Andre van Delft - there seems to be no shortage of thick ice in Feb/March 2012 according to the US military - http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn_nowcast_anim365d.gif
Despite the 20 years of alarmist bed-wetting we have had from many scientists who should know better, and the idiots in the media, the late 20th century decline of the Arctic sea ice is nothing unusual or anything to worry about, once given some historical context: For example:
Historic Variation in Arctic Ice – Tony Brown, 2009.
Greenland was 2–3°C warmer than today between 6000 and 4000 years ago, and http://www.sciencedirect.com/science/article/pii/S0277379112004209.
This is confirmed by good archaeological, driftwood and geological evidence:
North Greenland coast, open seas, archaeological evidence for much warmer climate Arctic Ocean, Independence I Culture, as needed seals and driftwood to survive 4000-4500 years ago. But then climate cooled. http://www.ngu.no/en-gb/Aktuelt/2008/Less-ice-in-the-Arctic-Ocean-6000-7000-years-ago/ See also: http://www.bbc.co.uk/news/science-environment-14408930 also.
Here are some more ice and climate history links for anyone interested:
North West Passage ice area data (current) - http://www.ec.gc.ca/glaces-ice/default.asp?Lang=En&n=2D414464-1&wsdoc=61BCCE40-2248-11DF-9BC2-B9D43796A02E
climate history - Viking farming settlements on Greenland - http://www.archaeology.org/online/features/greenland/
climate history - Greenland - old Danish aerial photos - glaciers melted faster in early 20th century - http://motls.blogspot.co.uk/2012/05/knud-rasmussen-pictures-greenland-is.html
climate - Greenland, the long slow thaw from the LIA Jakobshaven Glacier late 20th Century retreat in context - http://farm5.static.flickr.com/4139/4753025383_5c34f3dd93_b.jpg and https://stevengoddard.wordpress.com/.../most-of-the-jakobshavn-glaciers-retreat-occurred-when-co2-was-less-than-310-ppm/
climate history - Arctic temperature last 1500 years - Rørvik et al. state “Bottom water reconstructions in the Malangen fjord clearly reflect a cold period during the LIA and the dinocyst assemblages indicate that the time period from AD1500 to 1940 represents a cool period.”
http://www.worldclimatereport.com/index.php/2011/03/10/amazing-arctic-reconstructions/
climate - list and abstracts of 14 papers which suggest Arctic was warmer in early Holocene http://climatesanity.wordpress.com/2008/10/15/dont-panic-the-arctic-has-survived-warmer-temperatures-in-the-past/.
climate - John Daly: What the (Arctic) stations say.
Nice elision by the Minister. No, we didn't use any statistics or models, so can't help you there. What we used was (refer to the vast plains, hills and swamps aka peer reviewed climate science) - and your question didn't cover what they used.
OK, they've both got their pawns out. What next?
So submarines still surface at the pole? It seemed to be regular photo-opportunity for Nat.Geog. photographers in the 50's...
Bish,
I am surprised and a little disappointed at your concluding paragraph. Sure, we know there are distortions in the peer review process, and it is true that there are a lot of bad papers published. However, at the end of the day we need a clear road back from the present scientific over-enthusiasm for man-made catastrophic climate change. There are an increasing number of extremely well-written and unrefuted cases made in the mainstream literature now which argue either implicitly or explicitly for a lower climate sensitivity on the basis of observational data from multiple sources. There are an increasing number of well-argued papers accepting the reality of (large amplitude) unforced natural variation. There are an increasing number of papers which are re-establishing the erstwhile accepted dogma of multicentennial variation of climate and temperature. Suppose this mini-flood becomes a roaring torrent. If you tarnish all peer-reviewed literature now, then what are you going to point to in 5 years time to support the changing story? (Ha, ha, you said that all peer reviewed literature was junk, but now you are saying...")
The MO should be praised for avoiding the trap that they dug for themselves when answering Lord Donahue's questions. I look forward to Peter Lilley's follow-up question, which should be: Which particular papers in the peer-reviewed literature are you relying to support the view that the downward trend is statistically significant, and does the government have any view on the appropriate statistical model on which to base the null hypothesis for this test? </>
I would draw readers' attention to a very important Climate Science Meeting in Taiwan ( the recent 2012 National Taiwan University International Science Conference on Climate Change). The focus was on understanding multidecadal climate variability and in particular distinguishing between forced response and natural variability.
http://journals.ametsoc.org/do.....13-00015.1
It is interesting that the summary document suggests widespread acceptance of the reality of unforced multidecadal and multicentennial oscillations. One element in particular is relevant to the ice-loss question:-
[AMOC = Atlantic Meridonial Overturning Circulation]
Entropic Man,
Your reference seems to be but a repackaging of IPCC reports.
Jun 12, 2013 at 2:52 AM | Old Mike
Not surprising. The IPCC is tasked with producing a review of the published literature on climate change and using it to advise governments on possible outcomes.
Any other discussion relating to climate change will draw on the same published literature.
Philip, those trends look tenuous at best. And I saw no reference to black carbon affecting thick/old ice disproportionately in the NASA links. Really, the CAGW story has enough of its own problems without making new ones up ;-)
I look at the Sea Ice Page on Wattsupwiththat at least a couplr of times a week. From my observations of the (unbiased) satellite-derived charts I have one thing to say:
'Arctic sea ice trending in 2013 looks to be incredibly average...'
For those of you who may be interested I did try to highlight the problems with the Met Office answer to Lord Donahue by presenting a counter-example which produces the opposite results i.e. it proves that the 20th century temperature gain is not statistically significant. For reasons which Doug Keenan at least will understand, the method is scientifically baseless - but not as flawed as the Met Office approach.
The article is here:-
http://rankexploits.com/musings/2013/the-occams-razor-oscillatory-model/
Paul_K
I had missed that - very interesting, and roughly tallies with my own view.
Sorry, here is the URL for the Taiwan climate conference from which the Arctic Ice reference came:-
http://journals.ametsoc.org/doi/pdf/10.1175/BAMS-D-13-00015.1
I have two problems with the oscillatory models.
Both a 60 year oscillation ( AMO?) and an a decadal oscillation (solar cycle?) show in the temperature record, but produce too small a change to account for the overall temperature increase.
If the oscillations were sufficient to explain the change, they would show clearly as sufficient changes in energy flow. Neither cycle is processing enough energy to produce the overall changes observed.This applies to all the known "natural variation" drivers and any of the unknowns. Their effect would show in the energy flows.
Temperature changes are driven by energy changes,. If you wish to explain the temperature change, you need to explain where the energy flows come from. This is why CAGW is so widely accepted outside your circles. Nothing else matches as well.
If you wish to convert the scientists give them a better explaination than the one they are currently using, with proper numbers.
"Given what we know about the amount of checking that doesn't go on in academic studies these days, particularly climate change, for policymakers to rely on the scientific literature is foolish in the extreme."
If you want to blacken the peer reviewed science for political purposes, that's your privilage as a citizen. Unfortunately, unless you can replace that science with something better you are only fooling those credulous to listen, and possibly yourself.