One of the greatest challenges surrounding global climate change is in predicting the kind of extreme weather incidents that cause the most havoc. Even a few hours extra warning can make the difference between life and death for those in the firing line.
The trouble with weather and climate is that they are the result of incredibly complex systems interacting with each other in an incredibly complex ways. Climate models are becoming ever more advanced as the various interactions become better understood. Scientists can predict weather patterns no with an accuracy that would have been pure science fiction even as little as a decade ago. That is, they can if they have the raw processing power necessary to crunch the numbers.
There are a number of ‘super-computers’ in the world. Time running programs on these is expensive, and very much at a premium. Fortunately there is another way to get this work done, and it all hinges on using the idle capacity of the Internet with its millions of ordinary desktop PCs around the world.
For the average user the majority of their system resources are idle a lot of the time. Distributed computing projects work by scheduling tasks across the available capacity, typically by the user installing a ‘screen-saver’ type program. Portions of tasks are run when the system is not being placed under high demand by the user – which is actually quite a lot of the time. If the computer has a 3D graphics card this is particularly useful in performing large numbers of complex mathematical operations quickly.
The latest project to use the power of donated computer time is climateprediction.net. Unlike early projects such as climate@home this one uses regional climate models including factors such as winds, temperature, and precipitation. You can get involved yourself today – its good to feel like you are part of the solution every once in a while!
You can leave a response, or trackback from your own site.