Sunday, November 29, 2009

Dumbest climate education move ever

In the process of looking up information on edGCM, the educational version of NASA's General Circulation Model 2, I found out that they now charge $250 bucks.

What the fuck?

The whole purpose of this project was to provide a transparent, accessible climate simulator, so that people who are wary of trusting their future to voodoo climate models can actually download one, pick apart the code, and see how they work.

Only now, it costs hundreds of bucks to do so.

What kind of moron expects that skeptics who think global warming is all a hoax are going to shell out this kind of money? Or indeed any money at all? Convincing people that it was even worth a mouse click was hard enough, but this is fucking absurd!

In fact, putting it behind a paywall plays right into the hands of the conspiracy theorists, who will point to this as a coverup attempt.

So did Columbia sell the entire project to Marc Morano to cover their budget shortfalls? Or did they simply donate their brains to science (presumably nanotechnology)? With friends like these, who needs denialist shills?

Edit: Also, GCM stands for "General Circulation Model", as the Goddard site attests. Revising the acronym to "Global Climate Model" makes them look like they have no idea what their own product is.

9 comments:

Anonymous said...

TPB is your friend in these situations.

Chuck said...

Trailer Park Boys?

James Annan said...

Well I guess the question is, who should pay for this? The model is a dead duck scientifically, an interesting piece of history but little more. We managed to compile and run it, but some changes to the code mean it no longer replicates the original output.

Chuck said...

did you run the open source code at GISS or just DL the edGCM executable? I have done the latter with previous (free) versions, but not really put the effort in to make the output comprehensible. Do the changes make it wrong? Or is it just too crude to tell us anything we don't already know?

As for paying for it, why not just throw it in a wiki and let the public develop it as they like? They paid for the original with their tax dollars, after all...

James Annan said...

I got the real edGCM fortran code. I think one of the changes was to the grid, there may also have been some minor bug fixes. The overall result is the climate sensitivity is significantly higher than the original.

There are other more modern models that are more compatible with modern compilers...I remember it being a bit painful to get running (actually, most models are).

Chuck said...

Was the original too low, or is the new one too high (assuming that 3 +/- 0.5 is correct)?

Do any other more compilable models run on the sorts of computers that normal people own?

James Annan said...

The original was about 4.2C, it is now 4.9, and the transient response (eg to 1% pa increase) is also particularly high, probably due to the simplistic ocean. It also has no facility to account for tropospheric aerosols.

I suspect that GISS model E would be a good starting point if you want to compile and run your own. It claims to run on various systems with a range of compilers, and is computationally pretty cheap.

Chuck said...

Is there a facebook app for compiling fortran-based climate models? If not, then I am probably in over my head.

James Annan said...

Is there some fortran code for facebook? If not, ditto :-)

Not that I'm any good at compiling complex things - I just start up Wife_1.0 and feed her regular cups of tea while she gets on with it...