Buy

Books
Click images for more details

Twitter
Support

 

Recent comments
Recent posts
Currently discussing
Links

A few sites I've stumbled across recently....

Powered by Squarespace
« Josh meets xkcd, well, kind of - Josh 255 | Main | The big EAsy »
Sunday
Feb022014

What to do with a hot model

One of the things I've noticed about climatologists is that once they get each generation of models out into the open they spend the following few years producing papers that analyse some aspect of the model output. This is no doubt an easy way of making an impact on the research evaluation exercises to which all academics are subjected. And if the papers are accompanied by bloodcurdling headlines about future disaster are no doubt good for promotion, salary increases and invitations to speak to the United Nations.

This paper (via Leo Hickman's Twitter feed) looks to be from the same drawer. Here's the abstract:

Trends of Arctic September sea ice area (SSIA) are investigated through analysis of Coupled Model Intercomparison Project phase 5 (CMIP5) data. The large range across models is reduced by weighting them according to how they match nine observed parameters. Calibration of this refined SSIA projection to observations of different 5 year averages suggests that nearly ice-free conditions, where ice area is less than 1 × 106 km2, will likely occur between 2039 and 2045, not accounting for internal variability. When adding internal variability, we demonstrate that ice-free conditions could occur as early as 2032. The 2013 rebound in ice extent has little effect on these projections. We also identify that our refined projection displays a change in the variability of SSIA, indicating a possible change in regime.

So far so bloodcurdling. However, it seems to me that the authors, and indeed the climatological community as a whole, have a problem. We know that the aerosol forcing figures in the CMIP5 models are far greater than the best observational evidence would suggest. This being the case the models will necessarily run too hot. This presumably makes the claim that the Arctic ice will be gone by 2032 just a weeny bit shaky.

Aren't they going to have to sort the aerosols out before they can start to make predictions?

PrintView Printer Friendly Version

Reader Comments (29)

"the claim that the Arctic ice will be gone by 2032 just a weeny bit shaky.

Aren't they going to have to sort the aerosols out before they can start to make predictions?"

Gosh no, they'll have retired by the time that prediction fails.

Feb 2, 2014 at 9:02 PM | Unregistered CommenterTinyCO2

What to do with a hot model?

Why, take her our for a drink, of course!

Feb 2, 2014 at 9:06 PM | Unregistered Commentertheduke

TinyCO2 said:

... they'll have retired by the time that prediction fails.

That was exactly what I was going to say, but you beat me to it!

Feb 2, 2014 at 9:15 PM | Unregistered CommenterRoy

I can't keep up. I've only just got over the fact that there is still arctic ice in 2014 after we were promised faithfully that it would all be gone in 2013. Now I'm told that I've got to wait until 2032, by which time I'll be too gaga to notice. Has anyone ever explained millennialism to the climastrologists?

Feb 2, 2014 at 9:25 PM | Unregistered CommenterKevin Lohse

predictions, predictions. Here's one from Cambridgeshire, hot off the press...

It’s an Early Spring!
Goodbye winter says Abington Guinea Pig

Abington, Cambridgeshire
Sunday, February 2nd, 2014

It's Guinea Pig Day, and the long-range forecast is in from Abington's rodent "weatherman", Gnipper. According to legend, if the guinea pig can see the shadow of his tail on Candlemass morning, the 2nd day of February, there will be six more weeks of winter. No tail shadow is sure to predict an early spring.

Emerging from his cosy hutch this morning, into a balmy 10oC and sunny skies, Gnipper sniffed the air, looked all around and then down at the ground. He saw snowdrops already in full bloom but no tail shadow and squeaked the good news to his handler – it’s going to be an early spring!

Gnipper’s prognostication was later confirmed by his trans-Atlantic rodent cousin, Long Island groundhog Malverne Mel. Unusually though, this year Pennsylvania groundhog Punxsutawney Phil at Gobbler's Knob, Punxsutawney, Pa. disagreed with Gnipper and predicted six more weeks of winter.

University of Abington Climate professor Zephyr de Hund said it is not unknown for rodents to arrive at differing prognostications, because of the natural variability in the climate system. It would be good to have an early spring though, he said, as the recent weather has been “rough”.

Speaking at a meeting of houseplants, however, well-known eco campaigner and heir to the British throne Prince Charles said: “this clearly shows climate change deniers are 'headless chickens'. Every year the guinea pig predicts an early spring – the evidence for climate change is overwhelming”.

<ends>

Feb 2, 2014 at 10:00 PM | Unregistered Commentergareth

As that chap Williams said on your radio segment, all the models give similar results so they must be right. Furthermore there is not much point in playing games with cooler models that do not give alarming results as the paper would then not have much impact. In an earlier life I served on a committee approving government research grant applications in natural sciences and engineering. One of the most important criteria for funding approvals was "research impact". By that they meant whether anyone would take any notice.

Feb 2, 2014 at 10:04 PM | Unregistered Commenterpotentilla

So "ice free" is less than 1 million square km - that's about the same as France and Germany combined (1.03 million) - seems like a lot of ice to me !!

Feb 2, 2014 at 10:15 PM | Unregistered CommenterMartin

" The 2013 rebound in ice extent has little effect on these projections."

No comment about the 2013 rebound in ice volume?

Esa's Cryosat sees Arctic sea-ice volume bounce back

"......One of the things we'd noticed in our data was that the volume of ice year-to-year was not varying anything like as much as the ice extent - at least for the years 2010, 2011 and 2012," explained Rachel Tilling from the UK's Nerc Centre for Polar Observation and Modelling (CPOM).

"This is why we're really quite surprised by what we've seen in 2013.

"We didn't expect the greater ice extent left at the end of the summer melt to be reflected in the volume.

"But it has been. And the reason is related to the amount of multi-year ice in the Arctic........"

Feb 2, 2014 at 10:30 PM | Registered CommenterGreen Sand

"Aren't they going to have to sort the aerosols out before they can start to make predictions?"

Absolutely.
I've just written on a blog here Down Under that we are spending inordinate time and effort, not on the raw data, but on "getting the message over" and even highly finessed, voluminous discussion on whether one method of adjustment is better than another.
Goodness, if we had high quality 'climate' science, we would have progress from open data release and intelligent use of the raw data.
Never mind the message, we have read it before, just get the data out in a form that allows audit.

Feb 2, 2014 at 11:12 PM | Unregistered CommenterGeoff Sherrington

For simulations of Arctic sea ice extent to have any merit, the models will also have to be able to simulate the multidecadal variations in the sea surface temperatures in the North Atlantic (critical because the Arctic Ocean is open to the North Atlantic) and the multidecadal variations in the sea surface temperatures of the North Pacific (not the PDO).

Feb 2, 2014 at 11:47 PM | Unregistered CommenterBob Tisdale

I see there is many free courses appearing on climate to educate us all -
http://www.realclimate.org/index.php/archives/2014/02/new-climate-science-moocs/

Feb 3, 2014 at 1:07 AM | Unregistered Commenterdougieh

"Aren't they going to have to sort the aerosols out before they can start to make predictions?"

---------------------------------------------------

No, before the main ice melt drivers (AMO and black soot) are included, the model is still rubbish.

Feb 3, 2014 at 3:08 AM | Unregistered CommenterManfred

""the claim that the Arctic ice will be gone by 2032 just a weeny bit shaky.

Aren't they going to have to sort the aerosols out before they can start to make predictions?"

Gosh no, they'll have retired by the time that prediction fails."

It's about SSIA. September sea ice area

Feb 3, 2014 at 5:30 AM | Unregistered CommenterJon

A model that has spend so much time and effort to become hot probably also have a good idea what to do?

Feb 3, 2014 at 5:38 AM | Unregistered CommenterJon

Apart from the never ending research grants and high salary, the prestige and the foreign travel, one wonders why anyone with any interest in science would want to spend a lifetime working with what they know are unvalidated computer models which have zero ability to calculate the future with any confidence. It would have driven me mad spending all my working life wasting my time and wasting money doing something I knew to be so worthless. It's time for a top psychologist to study why such people do it.

Feb 3, 2014 at 7:25 AM | Registered CommenterPhillip Bratby

Whether, there is lots of sea ice, or no sea ice at the Arctic pole - what is the question?

During probably the Roman and MWP - the sea ice extent was maybe diminished because anecdotal evidence suggests [archaeology] that, the Greenland ice cap retreated back from coastal regions but that, it was not due to man made influence. Whatever happens to the sea ice and it is mainly down to oceanic currents - influence entering and circulating in the Arctic basin and of course the Solar input and particulate pollution - thus, man made CO2 [if it ever was] is irrelevant, then - as was, to as it is, now.

Compute all you like, but get this Hickman - this is an academic exercise for climate alarmists clasping at ectoplasm - statisticians and there is: nothing to see here but idle and useless speculation.

Feb 3, 2014 at 8:13 AM | Unregistered CommenterAthelstan.

Strange when they refer to polar amplification, they leave out Antarctica.

Feb 3, 2014 at 8:51 AM | Unregistered CommenterCharmingQuark

Athlestan -

During probably the Roman and MWP - the sea ice extent was maybe diminished because anecdotal evidence suggests [archaeology] that, the Greenland ice cap retreated back from coastal regions but that, it was not due to man made influence. Whatever happens to the sea ice and it is mainly down to oceanic currents - influence entering and circulating in the Arctic basin and of course the Solar input and particulate pollution - thus, man made CO2 [if it ever was] is irrelevant, then - as was, to as it is, now.

Yes, while there may be some uncertainty about the magnitude of the warming in the Roman and MWP, there is no uncertainty that the Arctic seas were open (at least in summer) on the north coast of Greenland during the Holocene Optima, thanks to very good archaeological, geological and paleontological evidence. (drfitwood on raised beaches and ridges which can only have been formed by wave action - http://www.bbc.co.uk/news/science-environment-14408930

The same seems to be the case in neolithic times: Stewart, T. and England, J., Holocene Sea-Ice Variations and Paleoenvironmental Change, Northernmost Ellesmere Island, NWT., Canada, Arctic and Alpine Research, Vol 15, No. 1, 1983. Stewart and England examined more than 70 samples or Holocene driftwood on Ellesmere at 82° N Latitude. The time distribution of the driftwood indicates “prolonged climatic amelioration at the highest terrestrial latitudes of the northern hemisphere” from 4200 to 6000 years before the present.

Feb 3, 2014 at 9:26 AM | Registered Commenterlapogus

Arctic sea ice is building whilst Antarctic sea ice is at record levels.

The problem with all these model assessments is that the academics think that their work reveals reality whereas it only reveals alarmist claims with no foundation.

Feb 3, 2014 at 11:09 AM | Unregistered CommenterJohn Marshall

"Weighting" models is the same as calibrating them to a particular set of circumstances and this article captures the pitfalls associated with that approach very well.

http://www.scientificamerican.com/article/a-formula-for-economic-calamity/

Here is a quote from this article about models that is very apt .................

"The problem, of course, is that while these different versions of the model might all match the historical data, they would in general generate different predictions going forward--and sure enough, his calibrated model produced terrible predictions compared to the "reality" originally generated by the perfect model. Calibration--a standard procedure used by all modelers in all fields, including finance--had rendered a perfect model seriously flawed. Though taken aback, he continued his study, and found that having even tiny flaws in the model or the historical data made the situation far worse. "As far as I can tell, you'd have exactly the same situation with any model that has to be calibrated," says Carter.

And ............. climate models are not even close to being "perfect models"!!

Feb 3, 2014 at 11:27 AM | Unregistered CommenterRobert Thomson

Your Grace

This being the case the models will necessarily run too hot.

That's Computer Aided Global Warming for you.

DP

Feb 3, 2014 at 11:57 AM | Unregistered CommenterDP

Re: Robert Thomson

You might find this paper interesting:

Our Calibrated Model has No Predictive Value:
An Example from the Petroleum Industry

Abstract: It is often assumed that once a model has been calibrated to measurements
then it will have some level of predictive capability, although this may be limited. If the
model does not have predictive capability then the assumption is that the model needs
to be improved in some way.

Using an example from the petroleum industry, we show that cases can exit where
calibrated models have no predictive capability. This occurs even when there is no mod-
elling error present. It is also shown that the introduction of a small modelling error can
make it impossible to obtain any models with useful predictive capability.

We have been unable to find ways of identifying which calibrated models will have
some predictive capacity and those which will not.

Feb 3, 2014 at 12:14 PM | Unregistered CommenterTerryS

The problem is that Global warming is not supposed to enhance the negative AO and NAO conditions that drive the loss in Arctic summer sea ice extent, quite the reverse. The cause of the increasingly negative AO/NAO episodes since the mid 1990's is therefore not internal to the climate system, but thoroughly typical for weaker solar conditions.

Feb 3, 2014 at 1:07 PM | Unregistered CommenterUlric Lyons

DP: “That's Computer Aided Global Warming for you.

Brilliant – I have long wondered what CAGW meant!

Feb 3, 2014 at 6:33 PM | Unregistered CommenterRadical Rodent

"The large range across models is reduced by weighting them according to how they match nine observed parameters."

Is anyone else bothered by that statement?

But heck, I'm just an engineer.

Feb 3, 2014 at 8:27 PM | Unregistered Commenterke25324

Lapogus

Why are you surprised. Temperatures stayed at levels not far below today's for several millennia. There was time for the Arctic to reach a low extent equilibrium.

We are returning to that state, though it will take a little while for enough energy to be pumped in to melt the ice.

I remain puzzled why you argue that less ice then meant a warmer past, while not accepting that less ice now means a warmer present.

Feb 3, 2014 at 8:28 PM | Unregistered CommenterEntropic man

But heck, I'm just an engineer.

Feb 3, 2014 at 8:27 PM | Unregistered Commenterke25324

That may be your problem. By the time science becomes engineering it has become what Kipling called "the tables at the end" of the textbook.

All that pesky uncertainty has been removed!

Feb 3, 2014 at 8:37 PM | Unregistered CommenterEntropic man

EM -

Lapogus

Why are you surprised. Temperatures stayed at levels not far below today's for several millennia. There was time for the Arctic to reach a low extent equilibrium.

Where did I say I was surprised? Please enlighten me, or have you been making stuff up again?

We are returning to that state, though it will take a little while for enough energy to be pumped in to melt the ice.

I remain puzzled why you argue that less ice then meant a warmer past, while not accepting that less ice now means a warmer present.

Again you are off track. I do not consider the current global sea ice extent (and by current I mean late 20th / early 21st Century) to be significantly lower than they have been in the late 19th / early 20th century -

http://arctic.atmos.uiuc.edu/cryosphere/IMAGES/global.daily.ice.area.withtrend.jpg

http://noconsensus.wordpress.com/2009/06/16/historic-variation-in-arctic-ice-tony-b/

Average temperatures in polar regions are still well below zero C for 10 months of the year. Global average temperatures have risen about 1C since the end of the Little Ice Age in the early 1800s, which was very likely the coldest prolonged spell in the last 10,000 years, and I am very glad temperatures have increased for that matter. For all the incredible 'warming' of the late 20th Century (which in time I am sure will be recognised as just a decade of mild winters with some UHI thrown in), the growing season here in Scotland is still only about 3 months long. Wake me up when global sea ice extent falls below 12 million km^2, or when I can manage to grow some sweetcorn, sunflowers, courgettes, basil or coriander without a fecking poly tunnel.

Feb 3, 2014 at 11:44 PM | Registered Commenterlapogus

EM

but in engineering we run tests to prove our tables (models) are good to go in the real world, or we get locked up or sued.

if this was just an academic discussion then ok, who cares, your point may be valid.
but the real world has different standards from academia.

Feb 4, 2014 at 1:15 AM | Unregistered Commenterdougieh

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Post:
 
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>