Twitter has proven itself a veritable as-it-happens earthquake detector.

As such, I think it would be very interesting to be able to geo-locate folks who are reporting shaking that might be related to an earthquake -- both as a function of time and of location.

There are number of ways to go about this.

The Big Brother Can See You method

The first would be to hook Twitter up on the backend to the USGS database and watch for quakes -- if you're able to either identify folks by the cell tower they're on (and thus geolocate them to a specific area) or if they've filled out location data in their profile or if they use Plazes, we could automatically generate a map with people's Twitter's plotted on them.

The Opt-in Model

If we wanted to make the system less accurate but possibly more fun, we could simply create a keyword (or letter) for Twitter... say "quake" or "q". Folks who preface their their tweets with either of these would be thrown into the "Twitterpation Quake Feed" and folks could do all sorts of interesting, albeit inaccurate, stuff... like show Twitterers running from a massive crack in the ground in some Flash visualization. Whatever.

Resources to use

Well, I think it goes without saying that we need quake data, so let's get that from USGS.

Next, we'll need Twitterers, and that should be obvious too. I wonder though if it would help to geocode each Twitterer's page if they specify a location. All it would take is the addition of a few metatags. For example:

<meta name="ICBM" content="37.783731, -122.394247" />
<meta name="DC.title" content="Chris Messina" />
<meta name="geo.position" content="37.783731;-122.394247" />  

I think it'd be fairly trivial to tie up a JSON feed to either the Simile Timeline or Exhibit tools...

...And then there's always Google Earth that would look really neat with Twitter speech bubbles popping up all over the place...

In any case, seeing quake Tweets popup in certain locations as quakes are really happening would both be interesting but also useful. Frankly, I'm

Inspiration

• Props to Nick Douglas for requesting this.
• Props to Tara for capturing an early example of this behavior.
• Props to Kris Tate building a piece of this.
• Props to anyone who builds something based on this idea.
• And props to Obvious and crew.

Feedbacks

Anyone gots feedbacks or suggestions?