Calstar 2006: Seeing stars in M31

by Marek Cichanski

At about 10 pm or so on Friday night, after only about an hour of observing, it was fun to announce to the folks around me "well, I think I've finished my observing project for the weekend."

Actually, if I'd had the time to prepare more data and more finder charts, I could have kept going with the project all weekend.

The project was to visually identify individual stars in M31. I was very pleased at how straightforward it turned out to be.

The idea for this project came during a reminiscence about this year's Australia observing trip. As I probably noted in my ORs from Coonabarabran, I was really blown away by the starclouds of the LMC. These starclouds were fully resolved in a 12" scope, and they were a joy to observe. I was also struck by the completely un-resolved nature of the central 'bar' of the LMC. It really struck me how different the intrinsic brightnesses of these stars must be. I think that I wrote an OR from Coona in which I talked about recapitulating a portion of the discovery made my Henrietta Leavitt.

Given what a thrill it was to resolve individual stars in the LMC, I got to wondering if this would be possible in M31. This would also make for another neat recapitulation of a big discovery in galactic astronomy, namely the first photographic resolution of stars in M31 by Edwin Hubble around the early 1920s. I haven't done the homework on this, but I think that Hubble first resolved stars in M31 on plates taken with the 100" telescope on Mt. Wilson. I can't recall if he used the 100" or the 200" on Palomar Mountain to make the first observations of Cepheid variables. At any rate, the resolution of stars in M31 was a big part of resolving the 'spiral nebula' controversy, and thus establishing the 'island universe' concept. I thought that it would be a real hoot if one could recapitulate this visually, at least to the extent of resolving some individual stars through the eyepiece.

The upshot is that after a lot of work with some recently-acquired data, it turned out to be even easier than I'd hoped.

A research group headed by Phil Massey of Lowell Observatory has recently done a high-resolution survey of stars in the Local Group galaxies:

http://www.lowell.edu/users/massey/lgsurvey/

I also read a PDF preprint of a paper by Dr. Massey and the research group; at the moment I can't recall if it's available from a link on the above-referenced site, or if maybe Dr. Massey sent it to me.

This group was able to do multi-spectral photometry that discriminated stars in M31 from stars in the Milky Way foreground. I noticed that some of the brightest stars had absolute magnitudes brighter than 17, suggesting that it might be possible to see them in my 18" Obsession or by using the 30" at Fremont Peak. I emailed Dr. Massey to see if he could think of any obvious dealbreakers, and he thought that it sounded plausible.

So, the Great Data Workflow began. The research group had put a .txt file of their M31 stellar catalogue on the web, and my first task was to pick out the stars that seemed like they'd be bright enough to see. The big challenge here was file size. The .txt file is 57 Mb, much larger than the 65,000-line limit of MS Excel. So, I opened it in MS Word, and cut it into six pieces that Excel could handle. (The document was over 13,000 pages long in Word!) After some sorting, I had a list of stars brighter than apparent magnitude 16.5.

Then it was on to the images. The research group's images are on the Lowell website, and I downloaded one of their V band images. They had 10 multispectral frames covering M31. Each image, however, is over 230 Mb! After a long download, I was able to open the image using "ds9", a software package from the Smithsonian Astrophysical Observatory. Ds9 is a really slick piece of software for working with FITS files, and each of the Lowell images had a detailed coordinate system built in to it. I was able to mouse around on the image and read out my RA and DEC to a very small fraction of an arcsecond. I was able to pick a star from my list, and find it unambiguously on the FITS image. Although ds9 allowed me to annotate the image, I decided to save time by using a printout of an image from the Hodge atlas of M31. The stars in question showed up quite clearly on the Hodge atlas, and I just circled them with a blue pencil.

After getting out of a meeting at school on Friday afternoon, I grabbed my finder chart, hopped in my vehicle, and battled my way down through the traffic to Calstar. I finished my setup after dark, and set to work looking for the stars. I found the first one within a few minutes, and I was able to do all of the stars on my chart within about 45 minutes or an hour. They were on Hodge plate #25, which is the plate that shows M32 and NGC206. I picked this area simply because it was pretty. David Kingsley also used my finder chart to identify some of the stars.

So, it turns out that recapitulating Edwin Hubble's big discovery is quite doable, thanks to the Local Group Group's new data and images. It's not so much an advance in technology, although I'd be remiss if I didn't give credit to John Dobson, Dave Kriege, and Al Nagler for the technology that I was using. Mostly it was a new dataset that revealed M31 to be lousy with mag 15 stars. If I'd had more time, I could probably have prepared finder charts for hundreds of these stars. And with a scope like the 30" at Fremont Peak, one could presumably identify many more.

I don't know if anyone else has done this. I suppose that it all depends on whether or not anyone had the idea of using the group's data in this way. I wouldn't be surprised if someone like Brian Skiff has done it. But I still entertain the happy notion (illusion?) that I might be the first to do it. Either way, it was a lot of fun.

All in all, it was a real kick, and it made a great memory to go with a great Calstar 2006!

Marek