gravitational waves and VOEvent

[John Swinbank]

On 25 May 2011 20:08, Roy Williams <roy.williams at ligo.org> wrote:

> The observation from LIGO will not be well localized, rather a wide
> probability map on the sky that may span several constellations -- see for
> example (*) . It could be represented in several ways: FITS on an RA/Dec
> grid, FITS on a Healpix grid, a set of contour lines of probability, a set
> of 'likely galaxies', or something greatly simplified from the full skymap,
> what?

Ed Daw and I experimented with triggering LOFAR using VOEvents
generated by LIGO during the run last August/September.

We adopted the "likely galaxies" table. Given LOFAR's relatively huge
field of view, I wrote some simple code to produce "optimum"
observations (ie, including as many high-likelihood galaxies per
pointing as possible). The whole exercise was relatively
straightforward and seemed to work well.

Of course, it would also be possible for us to take a probability map,
cross correlate it with a galaxy catalogue, and generate our
observations based on that. I guess an important question here is
where the "astronomical insight" should be happening. For instance, I
think it's a reasonable working assumption that LIGO systems know more
about the physics of gravitational wave emitters than we do at LOFAR,
so it makes sense for them to choose the most likely sources and just
point us at them.

I think that logic particularly apples when we're trying to do fast,
automated followup: a simple message from an expert giving a list of
possible targets is going to be easier to handle than a complex
special-case algorithm for each potential source we might receive
triggers from.

On the other hand, I'd imagine there's another use case to consider:
the human astronomer, examining the data at leisure, is presumably
going to be interested in getting as close to the raw probability map
as possible, so they can make their own decisions about which
particular galaxies or other sources might be of most interest to
them.

Cheers,

John