Open Source personal Tsunami Warning System
The president of Finland announced, that the Asian disaster which also affected some finnish tourists is the biggest disaster in our country during the slightly over 50-year period of peace-time. In Sweden things are even worse, as they haven't had any major events (last war was fought in 1814) in mid-term memory other than the sinking of Estonia, resulting in major difficulties for people to handle the issue.
I tried hard not to write anything about the Tsunami. Everyone else is doing it already in their weblogs and the level of information I'm able to provide to the collective pile is fairly low. Once again Wikipedia is a great source for updated "facts" of all of the noise.
Well, I guess I have found an element which connects to my overall interest so here it goes.
Robert Cringely has written a very interesting article about how to build a Global Internet Tsunami Warning System (or more conveniently, GITWS). The basic idea is that letting the governments build the multi-government warning system involves many many years of politics, new technology and great piles of money to build. The alternative?
We, of course - In a similar manner as weblogs provide alternative coverage to main-stream news and individual Open Source developers building great software from scratch.
A multi-government system requires a great ammount of data to process. Seismic sensors are monitored and as activity is detected, the data is combined and simulated automatically. From the simulation we can see how the event could affect certain locations and we could warn those locations beforehand. I expect a simulation like this requires a lot of calculation power.
If we attack the problem from another angle, i.e. "is my beach going to be devastated by a tsunami?", the ammount of information and seismic sensors you require to figure that out are fairly low. I expect there are many things you could leave out of your calculation compared to a full simulation. It's possible that this information could be processed fairly quickly on a single computer or on a local p2p application of some sort as a full simulation is not required.
If seismographs are online, all the data could be gathered electronically and a formula to calculate a disaster from your viewpoint exists, the system could be online in months, not in 10 years. The software could be released under Open Source for public good, of course. Everyone interested in if a tsunami is going to hit your location could have a nice tsunami icon in their taskbar, sending a horde of IM's, SMS's, email and flashing your screen to notify you in event of a disaster and give you plenty of time to sail with your precious cruiser to sunset.
It could be done?