Thursday, 21 February 2013

Dash-cam astronomy

One of many YouTube-posted videos. A GPS- or GLONASS-set
clock would have provided accurate time information though and
location and direction would have also helped us enormously.
The recent Chelyabinsk event demonstrated the power of distributed sensors for reconstructing incidents after the event. With cheaper and cheaper consumer electronics, miniaturisation, lower power requirements and the increasing distribution of these systems across the world, it is clear that we are entering a new era of Earth (and sky!) observation.

So, what is needed?

Firstly, there is the image itself. This is now widely available with resolutions and linear-responses far exceeding anything that astronomers could only dream of a few decades back. Generally speaking, there is no need to "flat-field" images or remove artefacts. The dynamic range continues to improve as well, making them increasingly sensitive to relatively faint events.

Then there is time. Most of these systems now have built in clocks which can stamp the image with the date and time. Although may have this information on the associated image files, it is embedding the information visually within the image that then allows others to derive useful scientific information. Some require manual date and time setting, which is subject to error and inaccuracy. Better still is a system which synchronises to satellites (GPS, ГЛОНАСС/GLONASS, etc.).

Which brings us on to position. Knowing the view point of the observation is also an extremely useful piece of information. It was the association of  Chelyabinsk videos with known land marks that made it possible for us (with no local knowledge whatsoever) to find references and therefore establish positional information. Again, by embedding this information visibly in the image it removes the need for the camera user to post this meta-data separately (thus saving time, making it easier and also reducing risk of errors).

Direction is also a useful piece of information and can supplement information, especially if other records of the event are scarce. For our own assessment of the Russian meteor, it certainly would have helped in the early stages (when were we trying to dis-associate the meteor with 2012 DA14), although later data helped clear this up. In some cases, this can be done by hand. If lucky, it may even be able to be solved automatically (such as with the astrometry.net project).

Image scale is the final thing that lets us work out the reference frame. This is especially important in wide-field images, which often have large amounts of pin-cushion distortion.

Whether dash-cams, all-sky cameras, CCTV, or from the hand-held devices of quick-reflexed users it is inevitable that there will be a rise in the quantity and distribution of sky data. No doubt, the ubiquity of computing, the influence of social internet media and a growing awareness of the public will contribute to a very new and fascinating era of incident astronomy.

No comments:

Post a Comment