NOAA just announced a new projection system that shows rotating spatial data on a sphere. It’s called Science on a Sphere. Four computer controlled projectors (one computer per projector, plus a fifth coordinator computer) project images that appear to move on a six foot spherical movie screen.
Science on a Sphere takes flat, two-dimensional images and data taken from spherical objects like planets and moons, and synchronizes and blends them into animated presentations. Most of the almost 100 presentations created so far are silent displays meant to illustrate lectures.
This is something you have to see, so the NYTimes has a short video on the subject. The NOAA site has the best video, though. My favorite is the x-ray sun. The 500-year CO2 simulator is pretty scary and in-person it must be one of the best visualizations of climate change available.
The system costs a whopping $180,000 for the hardware and software. I’m sure it could be done for cheaper and it could be better as well. Start with a weather balloon for the screen. The stick of gum-sized Linux machines could function as the “computers,” although I wonder if this couldn’t be done on a single computer (if necessary, could we do it with virtual machines?). Projecting the 2-D to 3-D data (that’s geographic projections, guys) is the mathematical stumbling block but most GIS software can handle it now. The specs state that the software accepts most graphics formats, but these are static .gifs, .jpegs, etc. Integrating the system with a GIS platform would allow the user to add/modify layers and create annotations on the fly (think of a sketch map, except rotating and 6 feet in diameter). I’ve simplified some of the details, but it’s doable.
Of course, this division of NOAA probably worked out a very nice GUI and, of course, they’ve managed the coordination of the projectors. The site mentions that they’ve developed an API and they adopted an Open-Source Software License so perhaps we will shortly be able to download the source code so we can create our own applications.
(The oblate spheroid is a geography reference. Planets tend not to be spheres. Most are oblate spheroids. That’s why in addition to projections, you also need to worry about things called datums too.)
I’d hazard a guess the majority of the cost was in writing the custom software, and less so in the hardware (five decent computers with four projectors shouldn’t be running you into the hundreds of thousands of dollars…) and subsequent iterations of it would be less expensive. I’m not sure the gumstix would be
I think it’d be pretty cool to have a huge dyanmic globe sitting in the lobby of one of the buildings here at McGill.
I don’t know about that. Take a look at NOAA’s site. They have some really fancy arced girders supporting that screen.
Then again, if one were to use a weather balloon, you might not need as elaborate a method for holding it up…