Super computers require Super power
Seems to me that the climate scientists are going to accelerate the global warming process, assuming one believes the hype, by producing guesstimates on increasingly more powerful energy sapping supercomputers.
On Thursday, the Met Office launched its new report on global warming: UK Climate Projections 2009, otherwise known as UKCP09. This is based on the output of Hadley Centre climate models that predict temperature increases of up to 6°C with wetter winters, dryer summers, more heatwaves, rising sea levels, more floods and all the …
Ain't it strange how, when their predicted doom'n'gloom failed to show, the Green Lobby instantly announced that it was reality that got it wrong since their calculations must be right, cos they said so.
If I were as vindictive as soon of those loons I have had the misfortune to know, I would suggest that if they are so sure it's Real Life that got it wrong, maybe it should be confiscated from them...
"They are claiming that they can predict what will happen in individual regions of the country - down to a 25km square. You can enter your postcode and find out how your street will be affected by global warming in 2040 or 2080."
It's not exactly complicated to interpolate between areas, not matter what the original resolution. It might not be particularly accurate on a tiny scale from large areas, but it is certainly better than leaving the general public to guess a value from the original data and nearby values (like I have to do for my weather forecast). I don't imagine that there are huge differences between adjacent areas anyway. Plus you can cross reference with more accurate local data with the height above sea level, river level, etc, to work out new values without having to increase your margin of error on the results. It's not rocket science, don't make it out to be so wildly inaccurate (not that I'm claiming that the results of the model are accurate in any way).
The Met Office can't tell you the current weather, let alone predict global flood and famine in decades to come. Ask anyone in the know, meteorologically speaking, and they'll tell you that the Met Office's modelling is completely naff. Mate of mine who lives in Mallorca called me a few weeks back to say that the Met Office forecast on the BBC had it that the Spanish isle was cold, wet, windy and generally unpleasant. The sky was blue, temperature was in the mid-20s Celsius, he told me.
Similarly, the weather here on the South Coast of the UK at the time was supposedly dull and cloudy with a risk of showers. You'll not be surprised to know that the weather was warm, sunny and - as I took the call whilst sitting out in the garden, glugging back my share of a pot of Java - pleasantly uncloudy.
When they can get the day's forecast sorted out with some kind of accuracy, I'll start to listen to their warnings.
"No, madam; there isn't a hurricane on the way."
Phil Jones: the Secret Agent in Hawaii
by Steve McIntyre on June 21st, 2009
As CA readers know, Phil Jones keeps his CRU data secret. Embarrassingly, the UK Met Office relies on this secret data and says that it is unable to provide this supporting data for the most relied upon temperature data set in the world. Their statements in response to FOI requests as to what they actually hold seem contradictory, but most recently they state that they do not hold original data, but only the "value added version" provided to them by Phil Jones. Whether they are entitled to keep the "value added version" secret is something that their FOI officer is presently considering.
Recently, Anthony Watts discovered that the Honolulu Observatory data, which NOAA and NASA lost track of in the 1980s, continued to the present day.
Anthony observed the substantial difference between trends at Honolulu airport and at more rural sites.
When I've done previous benchmarking of GISS data, I've usually tried to use relatively isolated stations so that the effect of data inclusions and exclusions could be simplified. Since Hawaii is relatively isolated, it seemed like it would be an interesting exercise to look at the Hawaiian gridcell, to get a preview of whether the "discovery" of a long data set might have an impact at the gridded level.
As so often, when one goes down a climate science rabbit hole, wonderland awaits. More »
1000x bigger than your or my PC?
I think we can manage that now.
1000x bigger than the biggerst super computer?
A bit more difficult.
But if the software *already* exists (in principal) how about custom hardware?
The EFF did a DES cracker system to prove DES could be cracked in a week at most (3 1/2 days) on average for less than $200k.
So with the structure of the software decided could a special purpose machine runt the job a *lot* faster?
A single chip won't hit more than a few 100 MHz, and a few million gates. So you'd probably need a lot of them, but they would be relatively cheap.
Just a thought.
but as interesting as it gets in the US ....
or (tiny url) http://tinyurl.com/lwa72u
(EPA) "The administrator and the administration has decided to move forward on endangerment, and your comments do not help the legal or policy case for this decision…"
"I don’t want you to spend any additional EPA time on climate change.”
Done deal indeed.
"down to a 25km square. You can enter your postcode and find out how your street will be affected by global warming in 2040 or 2080."
No, you can obtain a probabilistic forecast of the mean (monthly/seasonal) climate conditions within that 25km square. The results are available for different emission scenarios, and each has a high level of uncertainty. As others have mentioned, this is a downscaling exercise from the results of lower resolution climate runs (using either statistical techniques or via nested regional models)....downscaled from something like this:
Even the relatively low resolution (3.75deg x 2.5deg) hadcm2/hadcm3 runs provide some measure of UK regional variability...the newer hadGEM is twice this resolution. This will give some indication of the impacts of large scale climatic factors (e.g., general hemispheric warming, changes in the Arctic Oscillation). The downscaling attempts to account for localized forcing/feedbacks (e.g., topography, land use).
For those of you who are prepared to give the Met Office the benefit of the doubt on this one, it is worth having a look at this:
So far as I am aware, Humanitarian Futures cannot be described as a sceptical source.
Of course you're either purposely or mistakenly confusing climate with weather (or signal with noise).
I'm sure most people would be reasonable accurate at predicting their salary for the next year (with different scenarios indicating cut-in-hours/no-raise/medium-raise etc). Making a prediction of their day-to-day incidental expenses would be somewhat harder.
One thing I've noticed about the whole climate change debate is how very mid-1930s Nazism/Stalinist the language each side uses has become. Those who are sceptical that Man's CO2 production has much to do with climate change are no longer called "climate change deniers", but "but climate change denialists". Cheese above uses the phrase "you can obtain a probabilistic forecast". Probabilistic? Once I hear English being mangled like this I reach for George Orwell's Politics and the English Language and reflect on its essential truth: that not only does sloppy English reflect sloppy thinking, it actually engenders further sloppy thinking. This is not a comment one way or the other on the climate change debate: just a comment on how it is increasingly being framed. And a warning from the man with a pipe and a moustache and quite a few good ideas in the 1930s and 40s.
If you have a problem with the word "probabilistic" then I suggest you consult the OED (not normally an organ of Newspeak). People seem to be quite happy dealing with probabilities when they concern betting odds...what's the difference?
I'd have thought that presenting data with appropriate error bars and uncertainties is an appropriate response to the radicalized (political) language you mention previously. For example, the newest IPCC estimates for cloud feedback on climate range from the negative to the positive...the average being positive. By using probability, the range of uncertainty (betting odds) is shown. You seem to want a deterministic values i.e. the mean or median (the favourite is always going to win in the horse race).
Materials properties for alloys have been know to vary 25% (on a standard alloy composition). That's a failry well controlled environment where most of the variables affecting the outcome are known.
Never mind how big the model would have to be. Do we actually have *all* the relevant parameters known. I'm thinking how long climatologists dismissed solar activity levels as irrelevant (Hint. without the Sun this planet's atmosphere becomes a liquid layer roughly 1m thick and daily living become quite difficult).
And then we have the resolution. on a 2Deg resolution model would any UK city be big enough to register?
And bumps above the surface, or mountains. How much air flow can they possibly change?
So at what point do the probability ranges smear out so much they are virtually at the background (completely random) level?
A good point about sloppy language indicating sloppy thinking. The one that really gets to me is 'the environment'. The term is not capable of being used in this way. There are a vast, if not infinite, number of environments depending on what entity you are refering to, but there is no single environment.
Hilary Benn, another watermelon, another red painting himself in green. Like the Milliband brothers he grew up under a raving Marxist father except in Hilary's case it was Saddam's friend Tony 'Stop The War' Benn. It's no surprise to see so much nepotism at play among these socially concerned socialists. That's all a Red does - grab power then share power with friends and relatives.