[espadrine]

Congrats!

How do you automate screening the sky?

Do you have robots that map stars, diff the images, and send an email when the diff has a new spot of light?

How frequently does the sky get fully screened? Could there be events fast enough that we don't detect them?

[knice]

Apologies, I'm not one of the researchers involved in the discovery. I'm just the web developer who built the announcement site for UCSC. Ryan Foley will participate in a Reddit AMA tomorrow (Tuesday, 10/17). I imagine your questions might get answered there.

[knice]

This paper by one of the team members has more detail about the process used to narrow down the search field to a list of galaxies and identify possible locations of the event:

[edit]: link to the exact paragraph that discusses the method http://science.sciencemag.org/content/early/2017/10/13/scien...

[semaphoreP]

Not involved in this, I'm guessing this is how it goes: there are algorithms that automatically register the image that was just taken, diff it with a reference image of the sky from before, and if there's a significant difference that passes some false positive tests, then there's some notification for human intervention.

Indeed there could be events we are missing right now. Astronomers are building instruments that have larger fields of view (e.g., LSST), so that they can scan the sky ever few days.