The only thing we can be sure of about the future is that it will be absolutely fantastic. So if what I say now seems to you to be very reasonable, then I have failed completely. Only if what I tell you appears absolutely unbelievable, have we any chance of visualizing the future as it really will happen.

- Arthur C. Clarke (h/t Brin)

Sunday, June 28, 2009

For Those Who Underestimate Climatology



This is climate model output! See the movie here. Watch for the tropical storm in the Indian Ocean toward the end of the simulated month.

If anyone has any advice how I can include this Quicktime in a Keynote presentation I'd be very much obliged.

Update: It's a matter of paying $30 to Apple to ransom the ability to download a Quicktime.

12 comments:

James Annan said...

Ah, very truthy :-)

Michael Tobis said...

It depends on what you think a climate model is.

If someone thinks a GCM is a mere exercise in curve fitting, this should be curative, in ways that a coarser model could not be.

I am not sure science is advanced as a consequence of this feat, but perhaps the communication of science might be.

(I have been searching high and low for the equivalent Jamstec movie which I have seen, but could not find it, by the way. Too late though, this one is prettier.)

James Annan said...

A little googling finds this which dates from 2002 or so. It's not as graphically pretty though. The domain is cropped to the west Pacific inc Japan, in case the outline is not clear. But it's a global model.

Pedants may complain that these are both actually atmospheric models...

Hank Roberts said...

Pretty! I'd love to see the same time period from the weather satellite imagery.

Now if someone could warp the animation to fit over a sphere so it would actually line up with the real satellite imagery ... paging Pixar ...

Pico said...

Can you fill us in on what the model is being used for?

Michael Tobis said...

This is CAM, the community atmosphere model. The "community" is quite literal. This is the most commonly used atmospheric model by US researchers for many purposes, including being the atmospheric component of the most commonly used coupled model. (It is normally not used at such a high resolution, though; this run is very expensive.)

Your question thus reduces to what the purposes of climate models are. There are many. Which is a good point which I'll try to fit into my already over-full talk.

I have tried to come up with a description of what climate scientists actually do. It's on Correlations. Maybe that will help.

Pico said...

I'm participating in the climateprediction.net project (which farms out models to be run using idle processing time on people's computers). My computer has three processors, so it runs three model experiments at any one time. Some are short, some are long. Currently I have one which is 85% of its way through an experiment that will take 118 days to run.

It's a great idea, but the people running it have no clue about community engagement. I can watch the model earth spin, switch between views of clouds, temperature atmospheric pressure etc. But there is no way to look at the trends in any parameters over the course of the experiment. If you want to find out about what the experiments you are running are actually doing you need to do some serious detective work on their site. And they almost never provide updates about their experimental results.

It seems to me that this is a golden opportunity to engage people and get them to understand better what climate models are all about, while contributing in a very real way to the advancement of the science. If they realistically set out to make their project more attractive to the average punter, I'm sure they would be able to massively expand their contributor base.

Anyway, this has turned into a rant, but my point it that climate models are inherently fascinating things which I think climate scientists could use far more effectively to communicate the science.

Meanwhile, I hope that 118 day experiment and the dozens of others I've run are useful to whoever has created them.

Michael Tobis said...

I'd like to see a lot more transparency in climate models (and incidentally climate observation analysis, but that's neither here nor there). Unfortunately, it seems like it's mostly the bad guys who agree with me.

If I won the lottery, I would hire a few folks and start with a clean slate, a small language suitable both for human reading and theorem proving, and a physics-optimization friendly architecture.

However, because of the fact that your chances of winning the lottery are not substantially improved by buying a ticket, I don't buy them.

I have come to doubt that any public funding agency or private nonprofit will ever take my propositions in this regard seriously. To be honest, maybe they shouldn't.

It's a shame, though. Whether it's me or not, I think somebody could do better at this point.

Dan Satterfield said...

Michael,
The easiest way to get that in a keynote might be to use Screen Flow. It will make a QT of anything on your mac screen. I think it's more expensive than just sending Apple your 30$ though!
Screen Flow has come in very handy for such things though...

Kooiti MASUDA said...

Please watch NICAM .

coby said...

Michael, with the free version of RealPlayer you can download the .mov file directly from the webpage.

I have it and can post it somewhere for you to download if you like.

Tom Fiddaman said...

Here are some more movies.

Models:
http://www.ccsr.u-tokyo.ac.jp/~satoh/nicam/MJO2006/olr_gl11_061225-061231.mpg (direct link to NICAM cited above)
http://www.earthsimulator.org.uk/movie.php (NUGAM)

Data:
http://www.cira.colostate.edu/cira/Climate/webloop/nvap_webloop.html (NASA water vapor)
http://www.ssec.wisc.edu/data/composites.html (SSEC global satellite composites)