Open Tsunami Alert System

Saturday, January 08, 2005

 

Weak points in centralized Tsunami Warning Systems

Some good points about weak links in such a system.

Notice the number of weak links here that would be eliminated by decentralized non-governmental OTAS.

Comments:
In lieu of a trackback ping, some related notes comments:
http://dannyayers.com/index.php?s=tsunami&submit=Search+Archives
 
Agreed, security would be both important and difficult considering the numerous sources of data injection. Particularly considering the "default to alert" philosophy. Did I miss an article here, or do you plan to mimic the pressure sensors on 'sea floor -> bouy -> satellite' concept? What happens if your data sources go offline? Verification processes? Just wondering if this is hashed out somewhere?
 
Guys -- I share your security concerns (in fact I wrote something to Bob about script kiddies that's isomorphic with Allan's comment.)

I'll have a longer post on my basic architectural thoughts here shortly; the short answer is that I don't want this to necessarily depend on any single source of data, but that I also dont see this as having its own sensor web; instead, it should use other sources.

My thought is that a "conservative" alert system would be capable of being very simple, because the alerts could be based on seismic data alone. I am, however, by no means a geophysicist; there seems to be some question in others' minds, and they're better informed than I am.
 
First I'd like to say thanks for your concerns on this issue. We have death toll potentially reaching millions on Indonesia alone, which may be actually preventable.

I've written my thoughts here: http://harry.sufehmi.com/archives/2005-01-12-842/

Hope you'll find some of the ideas useful to this project.

Hopefully we can continue utilising this blog to co ordinate our efforts on this.
I'll check the SourceForge page soon.


Thanks,
Harry
 


<< Home

Archives

January 2005  

This page is powered by Blogger. Isn't yours?