Friday, June 23, 2006

Astronomy Internship - Day 14

Today I spent most of my day working on playing with the parameters for DAOPHOT's star finding algorithm to see if I couldn't get it to pick up some of the fainter ones.

It seems that biggest problem for detecting them is that their "shape" is slightly different. One of the input parameters is telling the program how far from the center of the star it is till its light has been reduced by 50% (this is called the Full Width, Half Max or FWHM). Ideally stars are infintely small, point sources. The reason they get spread out is due (primarily) to the atmospheric turbulance. But since we're looking at a small section of sky, they should all be spread out in the same way, which is why every star should have the same FWHM. Why some of these fainter ones don't is a bit strange. But they're not so far off that we think something went wrong. Just far enough that the program isn't feeling the love.

Once we got something that we were satisfied with, the next task was to do a rough first estimate on the brightness of each star. Since we're looking at a crowded cluster, this is very rough because stars are so close together that sometimes their light overalaps and thus, you're looking at two stars at the same time. This will be sorted out later in a process I might try to go into detail on at a later date (it's rather complex).

The program does that for us for the most part, so it didn't take much time.

Our next step was to find some stars that were good candidates for modeling the "shape" of so that we could attempt to estimate the shape for all other stars and then effectively subtract them out so we can look at one star at a time (note: when I say "shape" I don't mean 3D shape, but rather how the light is spread out and how bright it is on the CCD).

Unfortunately, the finding program detected 2989 things it thought were stars. Sorting through each of them manually does not sound fun. Thus, another set of programs are used that automatically cuts the options down to a few dozen based on how bright they are, how idealized their "shape" is, how nearby other stars are, etc...

But as luck would have it, the first of the three selection programs crashed repeatedly. Dr. Sandquist went to go check it out but wasn't able to immediately figure out why.

After about an hour, he fixed it and I only had to go through ~150 stars. I decided that I'd gotten enough accomplished for the day so I headed back to the dorm. That evening a few of us watched V for Vendetta in the lounge downstairs.

No comments: