Friday
Jan102014
by 
Bishop Hill
Bishop Hill Friday funny
Jan 10, 2014 Climate: WG2 According to Peter Stott of the Hadley Centre, it was fair of David Cameron to suspect that the recent UK storms were linked to climate change. Myles Allen agrees.
Cameron may not be right, but he has the right suspicions.
Reader Comments (71)
Absolutely. If this problem is so earth-shatteringly important, let's get it all up on GitHub so the community can peer review the code and start improving it.
A good start would probably be some unit tests. And don't give me that "it won't run on a PC" nonsense - if a single method (function) can't be tested in isolation on a reasonably spec'd piece of hardware, that's a good indication that some refactoring is required.
I suspect a better approach however, as is often the case when software is broken or not fit for purpose, would be to use the existing code as a lesson in how not to do it and start again. Of course, we'd need all the specs and surrounding documentation as well, but those are published, right?
"Within a couple of years, they could be assessing "in real time" whether weather extremes the country was experiencing had become more or less likely to occur as a result of climate change."
That one -yes 51:49, that one -yes 100:0, that one -yes 0:100 (It is possible to discern a link - and for a few dollars more we could be even more accurate).
Well, I'd be shopping there every day before I built my flood defences.
Competition lead for the title "Head of Climate Seance" is shifting back from Chris Turney to Myles Allen.
Jan 11, 2014 at 8:45 AM throg
Throg - the following may be of interest...
Engineering the Software for Understanding Climate Change
The abstract says: Climate scientists build large, complex simulations with little or no software engineering training, and do not readily adopt the latest software engineering tools and techniques (...) Their software practices share many features of both agile and open source projects, in that they rely on self-organisation of the teams, extensive use of informal communication channels, and developers who are also users and domain experts. These comparisons offer insights into why such practices work.
Having spent time in an organisation that made strenuous efforts to improve how it developed software and seen what it takes to do that, the foregoing sounds pretty dire to me.
However, I would say that the problems of creating the software are the *least* of climate modelling's problems. If the whole exercise is fundamentally flawed, does it matter whether or not the software is crappily constructed?
There is nothing wrong with models, if then give them what they 'want' even if they actual fail to work.
But still a more public insight into the way the MET and others have been able to use AGW to tap into funding they otherwise would never have got . Now I wonder why their very keen on keeping 'the cause ' on track ?
It would appear that Myles Allen has been contributing to the videogame "The Fate of the World"
"Accurate real-world data is used in many videogames. Take a motor racing game: you need to know how a car responds when a player brakes or turns too sharply. A climate-based game is similar, but you can't test drive the climate. So we needed the expertise of a real scientist. That's why Dr Allen's input is so important"
http://www.ox.ac.uk/research/research_impact/oxford_impacts/fateoftheworld.html
Yes Dr Myles Allen knows a lot about fantasy worlds. One of them is called ClimatePrediction.net.
It is so good to see our hard-earned tax going to good use!
Haha... Back on message.
Scientists back Cameron over a link between flooding and climate change
I think one Met Office spokesperson has probably been recommended for additional training.
You can find more on Allen here: http://scienceandpublicpolicy.org/originals/playing_climate_games.html
"Dr Myles Allen has stated quite clearly in the past that his scientific duty is to provide “evidence” for lawyers wishing to take up climate litigation claims against companies and countries for their perceived climate crimes in emitting carbon dioxide as part of supplying energy, or in a manufacturing process."
He was the consultant for a computer game in November 2010 called “Fate of the World”.
"Dr Myles Allen, who provided the climate model, pushed us to make sure we included methane as well as CO2, but he did so for game play reasons as well as scientific ones. He pointed out that if we included methane we could include a lot of exciting/scary geo-engineering technologies. That opened up a new set of features for players: who wouldn’t want to be able to risk plunging the Earth into an Ice Age, or cause floods of biblical proportions?"
Jamspid (Jan 11, 2014 at 8:13 AM):
I think he is getting enough of our money as it is, thank you very much.Its been said before on this blog that Myles Allen will end up on the winning side - whichever one it turns out to be.
10 million, eh! not 9 or 11....just ten measly million....to test a theory he already knows to be reality...sounds like a bargain..tease out the attribution .... 10 million quids worth of calculations. Good grief!
Jan 10, 2014 at 10:14 PM | 52
==========================
I lived in South Oxford from 1978 to 83. I recall it flooding at the bottom of Vicarage a couple of times back then, and certainly the streets around the river there would flood from time to time. Oddly, nobody attributed this to global warming or climate change. It's just what happened in a low-lying area by a major river. Whooda thunk it?
"Why Oxford floods" - http://www.oxfordfloodalliance.org.uk/page21.html
"Oxford's major floods are a result, not of water falling onto the immediate area, but of water which has fallen as far away as the Cotswolds coming down the valley of the River Thames. Severe flooding occurs in the plain to the west of Oxford. There is a marked geological narrowing of the Thames valley immediately south of Oxford city, around the old Abingdon Road red-brick bridge over the railway at Redbridge in South Oxford /Kennington. The natural narrowing has been made far worse by the building of the railway, roads and buildings, and by landfill. In some cases these seem to have been done with little or no regard for the possible effects on flooding, and lack of proper maintenance over many years has made things even worse. Further up, the Botley Road is a particular problem. It is a causeway across the floodplain from the City to Botley and Cumnor beyond. There is not enough capacity to carry water under it, or immediately to the south of it, so it acts as a dam. This leads to flooding."
New ITV 'period' murder mystery, perhaps - 'The Suspicions of Mr Cameron'...?
New ITV 'period' murder mystery, perhaps - 'The Suspicions of Mr Cameron'...?
SORREE - pressed the 'create post' button twice....
Just reading one of my son's excellent Christmas presents to his dear old Dad, who happens to revel in Bill Bryson's take on the world - this one entitled 'One Summer America 1927'...
Just got to the bit about the Mississippi floods in May of that year, when over SIXTEEN MILLION ACRES were flooded.
Now - call me cynical, but I don't think 'climate change' had been invented then....
Competition lead for the title "Head of Climate Seance" is shifting back from Chris Turney to Myles Allen.
Jan 11, 2014 at 9:06 AM | Ivor Ward
"Climate Seance" is good. It may have a future.
Myles Allen already has some form here. He was the Zebra of the Week once, a now-defunct prize that was intended to commemorate Michael Mann's observation in "Climate Wars" that skeptics identified and attacked individual climate doomsday scientists like lions separating zebras from the herd in the Serengeti. By Mann's reckoning last week we trapped the entire herd of zebras (52 of them, the whole deck of cards) in Antarctica, far from their natural habitat. That he got captured with his entire intrepid herd would have earned Chris Turney the Zebra of the Year award, if the prize continued.
Here is Myles Allen looking like a tasty zebra.
Ah, those were the days!
SORREE - pressed the 'create post' button twice....
Do that again and you're out of the hunt for a week. ;)
"These climate models are some of the most complicated pieces of software in the world".
If so, it can be said with absolute certainty that they will be riddled with errors. Software must never be allowed to become complicated.
In particular, for coding to be intelligible EVERY branching point in any software MUST be coded by either "IF condition DO x else DO y" or DO z repeatedly until condition". Any instance of a "bald" GOTO instruction renders the software inescapably complicated and the problem of understanding the code immediately ramifies into utter incomprehensiblity.
To judge the quality of their software just ask the climate modellers if their programs make use of GOTO instructions. If they say it is so then their programs are JUNK, fatally and unavoidably.
simon abingdon...
The following is not entirely irrelevant:
New peer reviewed paper finds the same global forecast model produces different results when run on different computers
Also
Assessing climate model software quality: a defect density analysis of three models
As I said before, the undoubted fact that the software is crap is essentially irrelevant, since what they are trying to do is doomed to failure anyway.
To judge the quality of their software just ask the climate modellers if their programs make use of GOTO instructions. If they say it is so then their programs are JUNK, fatally and unavoidably.
Jan 11, 2014 at 2:37 PM | Unregistered Commentersimon Abingdon
...... Climate Muddlers, shuerly