Distributed Computing World Climate Simulation

Burnt Offerings writes: "The BBC reports that scientists at climateprediction.com are nearing the completion and public release in late summer of a distributed computing project that simulates the world's climate from 1950-2050 AD. It seems that each user's simulation will have different initial conditions built into their runtime simulation and a single completed simulation from 1950-2050 AD takes on average eight-months (Doh!), assuming average household computing power. The results will be sent back to the project's team, where they will select the models that resulted in the 'real' climate patterns that have occured since 1950-2000. I presume they will then use these validated models to help extrapolate the world's climate from 2000-2050. Pretty cool (or should I say warm? or hot?)."

Comments

end result

[Graspee_Leemoor]

The end result of the project:

"On 1st January, 2050, it will start rather cloudy with outbreaks of rain, mainly in the north. These will clear up by late afternoon, leaving it warm with mild breezes in most of the country."

graspee

Re:Except

[shogun]

I bet I can take just 48 years without even using a PC and have an ever more correct answer then everyone using the lastest pc.

Infeasible

[Enonu]

Blame the climate changes from 1950 to 2000 on the expanded use of the automobile and unregular industrial waste. Do you think any scientist in 1950 could have known about our current situation? How can we in 2000 know about the new problems that'll creep up between now and 2050?

Spend your extra CPU cycles computing the cure for cancer or finding ET. I doubt this will prove useful.

Re:Infeasible

[Papineau]

I think what they want to do is, given the state in 1950 and what we know about the inputs (use of automobile, etc.), which models predicts correctly what we see in 2000, so that those models can then be used, along with some new inputs, to forecast what 2050 will be. Either prospective inputs, to get a glimpse of our possible futures, or actual inputs, to further validate the model or get a sharper view of the future climate.

That being said, 8 months is way too long to get something useful. I know a couple friends who reinstall their OS (and apps) in shorter terms than that, and don't really bother with bringing all data along, just some backup on CDR "in case I really want it again". I think they could at least chop it in periods of a few years, so that if you finish a "unit", somebody else can then pick up where you left. I'd like to see the completion efficiency of whole units in a few months.

Weather != Climate

[cperciva]

Weather is chaotic, but climate is ... well, ok, climate might be chaotic, but we really don't know -- and if it is chaotic, it is still only chaotic on timescales of more than 50 years.

Predicting climate 50 years in the future is a computationally difficult task, but it isn't impossible the way that predicting weather would be.

This is called "Boostrapping" and it is practical

[sam_handelman]

It's generally regarded as a Bayesian technique. Actually, there's far more to Bayesian statistics that bootstrapping, but it's the part I spend a lot of time working with. In fact, I suppose that bootstrapping isn't fundamentally a Bayesian process, but it is highly empirical so it appeals to the same "crowd" as more decidedly Bayesian techniques. By the by, "Bayesian" statistics are statistics that make heavy use of Bayes' Rule to incorporate prior knowledge not included in your measured data.

My background - you develop a program to predict something biological. Let us say, to pick a problem on the same order of difficulty as predicting the weather, that you're trying to predict the three dimensional confirmation that proteins assume, based on their sequence.

Now, okay, you have a bunch of known sequences, which other people (personally, I do both the data mining and some crystalography) have attached to known structures. So, what do you do?

Well, you could fiddle with your program until it predicts really well on those sequences, and announce that it was good. This is "Bad Science", as the parent-poster points out, since the criterion are arbitrary - you have a tendency to "discover" random noise in the data, and you have no way of validating your results.

So, second option. Instead, you split the data in half at random (actually into more than 2 pieces, but conceptually in half.) You take one half, and you make the model predict as well as you can on that data. Then, you VALIDATE ON THE OTHER HALF OF THE DATA. You *never* change the model on the basis of the second half of the data - that is arbitrary/bad/cheating. This is called "bootstrapping". It has nothing to do with compiler installation.

So, as far as most scientists (as opposed to mathematicians) are concerned, the important question is - does this work? In the biological sciences, I can say categorically, yes, this bootstrapping technique has a proven track record. It does work. Obviously, you can screw up (using non-representative data is a good start) but the technique, when properly applied, is sound.

So, I assume it would work for predicting the weather, as well. By work I mean - you would know how well your software predicted the weather. Bootstrapping is not a means of predicting the weather in and of itself, merely of honestly evaluating the effectiveness of a weather prediction mechanism you already have.

Something isn't right.

[blair1q]

They're starting with different initial conditions and hoping that some subset results in 50 years of weather?

Shouldn't they use the last 50 years of weather as initial conditions and vary parameters of the model instead?

What they're doing is like flipping an imaginary coin 500 times hoping to match the first 250 flips of a real coin to predict the the last 250 flips (albeit in a system with non-independent trials). But then they're taking those 500 flips and matching the first 250 to weather reports (might as well be coin flips) and then imagining the next 250 flips will approximate the future weather reports. What they need to do is fix the initial conditions and modify the model (coin flips vs. rolls of the die vs. LCRNG, etc.) to find a model that approximates the dynamics of the system.

Am I making sense here? How are these bozos not just going to apply their effective innumeracy to waste a few trillion CPU hours that could otherwise have been used to do protein folding or cancer-killing molecule matching?

--Blair

Re:Something isn't right.

[astroboy]

They're starting with different initial conditions and hoping that some subset results in 50 years of weather?

No. The term `starting conditions' appears in the BBC article, but if you go to the website it says:

The only systematic way to estimate future climate change is to run hundreds of thousands of state-of-the-art climate models with slightly different physics in order to represent uncertainties.

In large-scale simulations such as these, there are often bits of physics/chemistry/weather that have to be put in by hand because, usually, the relevant bits of science would be too expensive to calculate, or couldn't be seen on the resolution of the simulation. While it's usually pretty doable to come up with reasonable models for the unresolved effects, there are often parameters in the models that could take a range of values.

This ensemble of models allows for the callibration of the model parameters against 50 years of data; this gives some confidence in the predictive power of the models for the next 50 years.

This sort of parameter estimation based on calibration is very common for models of complex systems, and not just for computer models. Ideally, one wants to get to the point where such things aren't necessary and you can directly calculate all the science a priori of course, but these model calibrations are often useful steps along the way.

Extrapolation not pratical with chaotic systems

[lkaos]

As one poster has pointed out, weather is a chaotic system (and climate is also chaotic by definition).

Chaos is gravely misunderstood though so let me real quick through in my explaination for why this experiment will just generate FUD.

Chaotic equations are chaotic not because of the number of variables involved but because of the interdependency on themselves (each iteration requires the former iteration). This leads to extreme sensitive dependency on initial conditions (a.k.a. the Butterfly Effect). I should have probably emphasized the word extreme because even the slightly deviation will produce dramatically different results.

Even the best climate prediction algorithm would be crap if the initial condition was off by 10^(-20). The fact that we cannot measure temperatures exactly means that we could never feed a perfect initial condition.

Chaotic equations do have a given period before divergence gets extreme when initial conditions are altered. The original equations that Lorenz used (the pioneer of weather forecasting and the father of Chaos theory) showed divergence after about three days (which is why five-day forecasts still suck to this day).

I find it very hard to believe that these folks have developed an equation that doesn't show divergence for 100 years. Not to mention the fact that the number of initial conditions are much larger than the project makes them out to be.

Summary: Some PhD is looking for research money and figures that mixing "scientific" proof for global warming, chaos, and SETI-style distributed computer has to be good for a couple million at least.

INFORM yourself with the FACTS

[guanxi]

... or at least the best science has come up with so far, are downloadable from the Intergovernmental Panel on Climate Change (IPCC).

I'd start with the Summaries for Policy Makers, as a way of becoming very well infomrmed in just ~20pp.

AFAIK: It's a UN organization that is the center of research. Their reports are a consensus of almost all the leading scientists from every country on the globe, and their policy statements are approved line-by-line by governments. Even with all that, there are pretty strong statements.

Here's better background.