Stellar Photometry


   To measure brightnesses of point sources in CCD images, there are basically two methods: aperture photometry and PSF fitting. In aperture photometry, one defines one aperture (usually circular) enclosing the source, and another (usually a ring outside the first) that contains only sky. One obtains the mean counts per pixel from the sky aperture, subtracts that mean from each pixel in the source aperture, and sums the remaining counts to find the total in the stellar image. The advantage of this technique is that it is simple; the primary disadvantage is that the images of stars are big---probably a good deal bigger than you think, and certainly bigger than they appear on an image display. The point-spread function (PSF) has long tails, and a shockingly large percentage of the light of a star can actually get deposited on the detector many arcseconds away from the center of the image. If your stars are not substantially isolated on the sky, you will have overlapping PSFs, and aperture photometry will fail. The remedy is PSF fitting, which can be done simultaneously for many stars in an image. If you are doing star cluster photometry, this is the route to take.

   The procedure for aperture photometry is described in the User's Guide to Stellar CCD Photometry with IRAF. Even if you are doing PSF fitting, read this manual first. The preliminary steps (through the end of section 2) need to be done either way. This page describes only those places where the User's Guide procedure has to be changed because we are dealing with non-NOAO data. Keep in mind that, because of revisions in IRAF since the manuals were written, some of the tasks may have slightly different parameters than those shown in the examples.

   If you are doing absolute photometry---that is, trying to measure calibrated magnitudes---then you will have to go through the procedures to measure instrumental magnitudes of your standard stars and derive the coefficients of the photometric transformation equations.

   The recommended package for PSF fitting and cluster photometry is DAOPHOT. DAOPHOT is included in IRAF under the noao.digiphot package. (If you are running your own IRAF installation, you will need to make sure the TABLES package is also installed or DAOPHOT will not work.) This page makes no attempt to explain DAOPHOT; however a detailed Reference Guide to the IRAF/DAOPHOT Package is available.

2.1 Fixing your headers

Obviously, you need to edit the image headers only for your object frames. The only header keyword you have to add is AIRMASS. However, since the GOT does not write its own position into the headers, you will have to determine the airmass for each image manually. This is not difficult using XEphem.

   First, using your log sheets and the UT and EXPTIME keywords in the image headers, list the images for each target and the Universal Time at mid-exposure for each image. (The UT written in the header by CCDOPS is the start of the exposure.) Next, start XEphem. This time it is important that XEphem knows that the location is Athens, OH. Open the Sky View window, and use the Options window to set the orientation to RA-Dec. Center and zoom in on one of your target fields. (Refer to your finder charts to remind yourself where you placed the CCD field.) In the main XEphem window, set the date and time to the mid-exposure UT of the first image, and hit the Update button. Right-click on the center of the CCD field, and read the zenith angle (ZD). The airmass is the secant of the zenith angle; write this down. Reset the time for the next exposure, hit Update, and repeat for all the exposures on this field, then go on to the next field.

   When you have determined the airmass values for each image, you can add them to the image headers using hedit; for example: 

hedit n2024.fits "AIRMASS" 1.345 add+ ver-
If you like, you can use a text editor to write (and carefully proofread) a list of all of the hedit commands you have to execute, one command per line. Then you can execute it as a batch file by typing 
cl < heditlist.cl

2.1.1 Correcting the exposure time

   We have no knowledge of a shutter correction on the ST8 camera, so there is no need to do this step.

2.1.2 Computing the effective airmass

If you added the AIRMASS keyword using XEphem and hedit as describe above then there is no need to run setairmass.

3.2 Picking an aperture size

   Make sure that you examine your images carefully and use the imexamine task to measure the point-spread function (PSF) of stars in your images. This can be done most easily by putting the blinking cursor on a star and hitting 'r'. The three numbers on the right of the plot are the results of 3 different algorithms for measuring the FWHM of the PSF.

3.3 Setting things up

   You need the PSF FWHM that you measured above to decide how to set the aperture and box sizes for the phot task. Follow the advice in this section of the manual but DO NOT USE THE DEFAULTS. These assume a PSF 4 pixels wide. It is more likely that your PSF is in the range of 6 to 7 pixels. If you use the defaults you will only measure a fraction of the light of your stars and your results will be wrong.

3.8 Making the standard star observations file

   The example of the matching observations file (standstars) in Figure 17 is wrong. You need to give the image names without the extension (.imh in the manual, but .fits for our purposes), or the task will not run.

Additional Online Advice

Guides to aperture photometry and PSF fitting from Harvey Mudd College.
Last updated 2008 May 28. Written and maintained by Tom Statler