Supercomputing and Climate Research

Mr. Obvious writes: "It must have already been submitted, since the article is over a day old (gasp!) but there's a good round-up on the state of the art in supercomputing, as it applies to modeling the weather --- that is to say, modeling the planet --- over at the NYTimes. They go into lots of interesting things concerning how hard it is, what progress has been made lately, why the US researchers feel themselves to be hamstringed in comparison to those in Europe or Japan, and even into some things you probably didn't know (I didn't, at least) about the weather."

Comments

Annoying Slant

[Somnus]

The article notes the objection of global warming skeptics as if there is scientific consensus that a) the build-up of so-called "greenhouses gases" causes the Greenhouse Effect (probably true) and b) that an increase in the concentration of greenhouse gases is anthropogenic (probably false):

So even as the evidence grows that earth's climate is warming and that people are responsible for at least part of the change, the toughness of the modeling problem is often cited by those who oppose international action to cut the emissions of heat-trapping gases.

Yes, the Earth is warming in some areas, e.g. Siberia. But, this is totally expected if you look on a geological timescale, vis-a-vis the Ice Age cycle. The debate is centered on whether or not man or natural processes (cycles of flora and fauna, volcanoes) are driving the current trend. I have not seen any convincing evidence to support the existence of anthropogenic phenomenon, and plenty to support the existence of natural phenomenon.


*** Proven iconoclast, aspiring epicurean ***

Re:Annoying Slant

[blakestah]

The debate is centered on whether or not man or natural processes (cycles of flora and fauna, volcanoes) are driving the current trend. I have not seen any convincing evidence to support the existence of anthropogenic phenomenon, and plenty to support the existence of natural phenomenon.

There have been about a dozen articles published in Science in the last year in which model after model of climate has been tested. Time after time the models have converged on one and only one solution: increases in greenhouse gases are responsible, and the increases parallel those produced by man. The jury is in. The decision is done. The only issue left is whether mankind can do anything about it, and whether we can live with it.

Seriously, see the EPAs opinion http://www.epa.gov/globalwarming/climate/index.htm l

Also see the scientists commissioned by the UN to look into the problem - they also concur you are wrong

http://www.ipcc.ch/

TO paraphrase as at http://www.uic.com.au/nip24.htm

* Over the 20th century the global average surface temperature has increased by about 0.6 degrees C, more than earlier estimated to 1994. This appears to be the largest increase in any of the last ten centuries.

* Globally it is likely that the 1990s was the warmest decade and 1998 the warmest year recorded (since 1861). Certainly this seems to be the case in the northern hemisphere not simply since 1861 but for the last ten centuries.

* On average, between 1950 and 1993, night time daily minimum air temperatures over land increased by about 0.2 degrees C per decade, lengthening the freeze-free period in many mid to high latitudes.

* Since the 1950s the lower part of the atmosphere has warmed at about 0.1 degrees C per decade, as snow and ice cover have decreased in extent by about 10%, and Arctic sea ice thickness more than this.

* However, some important aspects of climate appear not to have changed, including storm frequency and intensity and the extent of Antarctic sea ice.

Re:Inputs one of the problems

[yellowstone]

<wag-about-the-future>
Within the next, say, 100 years (prolly a lot sooner than that), we'll have the ability to release millions, even billions of nano-probes into the atmosphere and oceans (c.f. Stephenson's The Diamond Age). The air-borne probes can measure temperature, windspeed, and humidity. The water-borne probes can measure water temperature, currents, evaporation.

Now imagine all these probes sending their observations back (in real time, perhaps using each other as repeaters to carry the signal) to a centralized data storage and analysis facility.

Now imagine a massively parallel computer running simulations based on these observations... As another poster observed, there are bound to be limitations on any system that doesn't have perfect observations at infinitely fine granularity. Whatever those limitations are, I suspect we are not too far from finding out what they are.
</wag-about-the-future>

(for those who are wondering, "wag" is a technical term used in estimating -- it stands for Wild-@$$ Guess)

--

This is scary

[Rosco P. Coltrane]

These crazy scientists are going to modify the weather pattern on Earth : as they progress in their weather simulations, they'll need more and more supercomputers, which in turn will run so hot they'll raise the temperature world-wide, which will make it harder for the scientists to simulate the weather, so they'll need more supercomputers ...etc... ARGHH, SOMEBODY STOP THEM !

Re:Supercomputer Envy

[Anonymous Coward]

Ok, I don't have a slashdot account, so probably nobody will read this, but you should know a few things about the top 500 list. The Top 500 list is not a measure of how fast a machine is at running weather codes. It is a measure of how fast a machine runs a benchmark called ``Linpack''. If you're trying to use this list to compare the capabilities of various computer with respect to predicting climate and weather changes, you will be mislead. The main complaint of U.S. climate scientists was that they did not have access to a decent Shared-Memory Vector computer like NEC's SX-5. That may be changing soon. The reason they want these is that they deliver real performance on real performance on real applications. The Top 500 list is a better indicator of theoretical performance. What you actually get varies widely depending on the types of tasks you ask the computer to do. Computer's like IBM's SP are good at running Linpack benchmarks, but they are less capable at running climate simulations. It is partly a matter of programming, but it is also a limitation of the SP's high interconnect latency and low interconnect bandwidth. I wouldn't call the T3E weird. It's basically the same as the SP (hundreds of commodity CPUs, each with their own memory, connected by a network) however, it's interconnect network is more sophisticated. Anyway, I can't explain everything in a little message but I hope you at least understand that Top 500 is not the whole story!

If only you knew the complexity

[zavyman]

If the Americans would suck it up and learn to use their amazingly fast IBMs we would hear whining from the other side of both ponds.

Great, what the hell do you think we are doing over at Argonne National Labs? I mean, have you tried to paralellize an atmospheric climate modeler? I don't think so. Coding for a vector-based machine is pretty straight forward. You concentrate on once machine as if it had one processor and one bank of memory, and code away, occasionally noting if your loops are not easily vectorized. The compiler magically does the rest, and your program runs really fast. That's why the Japanese machines are nice.

On the other hand, imagine designing an atmospheric climate modeler on a large cluster. The current paradigm being used and developed is MPI Let's see what you have to worry about. One, since the processors are not sharing memory, messages have to be passed between them to share memory. No biggie. But now consider that the whole atmosphere has to be broken up into pieces of a grid. On the boundaries, grid points must be shared by two or more processors. At each timestep, those points must be synchronized. Code must exist to know the processors that border one another and that know which points to share.

Now what happens if you want to combine different models together, as in the atmosphere, ocean, land, and sea-ice models. This is known as a climate coupler. Well, now you have differing grids for each of the models because they were developed independently. Now your program must handle interpolation of the grid points and must again know which processors border one another so that data is efficiently transfered. Finally, there must be decent load handling so that each processor is doing its fair share of the work.

I'm now working on the climate coupler project here at Argonne. Vector machines are quite easy to program, but we do know that they will lose out to massively parallelized clusters. It's just that the programming is much more difficult, since the messaging is always the bottle neck. Communications development follows a rate similar Moore's Law, but with a longer doubling time. For maximum efficiency, the programmer must handle the messaging model directly.

The modeling group here at Argonne understands the issue, and we are working on a general climate system to run quickly on parallized machines. No you know why you can't just do a simple code rewrite. You need to redesign the whole system.

Accelerated Climate Prediction Initiative.

semantic but important difference

[s20451]

as it applies to modeling the weather --- that is to say, modeling the planet

The article's a little misleading. It starts with a discussion of the weather, then moves on to discuss modelling of the climate. It's basically impossible to predict the weather -- meaning the exact temperature, rainfall, cloud cover, etc. -- more than a week in advance, because you have to specify the model with essentially infinite precision or chaotic effects take over. In fact weather prediction was one of the earliest manifestations of chaos theory. The climate -- meaning long term averages -- can exhibit stable behaviour that is possible to model in the long term. Don't look for this technology to dramatically improve weather prediction.