Times series with a close companion star

Magnus, if you would share one of your best images with HQA, I'd appreciate it.  I can't give you any help with VPHOT, but I'm very interested in seeing a current-epoch image of this field.

In VSP, you can select a DSS image for this target.  However, as in the case of many CVs, this comes with a bunch of caveats.  CVs are generally intrinsically faint (except in outburst), and so they tend to be very close to us.  For V627, the parallax indicates a distance of around 100pc.  Because of this, they also tend to have high proper motion, as any relative velocity with respect to the Sun will show up easily.  For V627, it is moving about 0.1arcsec per year to the northeast.

The DSS comes from scans of Palomar plates from the 1950's.  Therefore, the time difference between when the plates were taken and current date is about 60 years.  At 0.1arcsec/year, this means V627 Peg has moved about 6 arcsec from the position shown on the DSS.  The DSS shows the variable to the SW of a companion star.  The current view is more likely that V627 Peg is to the E or NE of the companion star, basically reversing the view you see in the DSS.

We recommend using the DSS to give a realistic view of the sky, as the star catalogs (UCAC, USNO-A) that are used for the dot-plot representation do have some missing stars or phantom stars that can confuse the observer (we've found a few dozen over the years, so it is not a big effect).  However, you can see from this example that the DSS is not a perfect solution.  It would be nice to replace the DSS with a modern-epoch all-sky survey!

When V627 Peg was recently in outburst at V=9.8, the companion was not a problem, as it is V~15.1 and therefore contributed less than 0.01mag to the total.  Now, when V627 Peg has faded back to quiescence, you have to be careful about the companion.  This problem also occurs for novae, which preferentially occur in the galactic plane with lots of crowding.

For photometry of V627 Peg, there are four basic methods that I know of.  Others might know more.

- increase your aperture size and include both the variable and its companion.  This has the advantage of simplicity.  If you know the magnitude of the companion, it can be subtracted from the measurement by using a tool such as the spreadsheet Nemesis, written by Lew Cook and available on the AAVSO web site.  You can submit the combined magnitude rather than splitting out the companion, but if you do, please add a note to the submitted magnitude that tells the researcher that this is a combined magnitude.  Your combined magnitude measure will likely be 0.1mag or more brighter than someone who removed the companion, but it is ok as long as you tell people about the process.

- use a very small aperture and try to just measure the variable.  Depending on how close the companion is, and seeing conditions, you may or may not be able to get a clean measure of the variable, as flux from the companion is likely to contaminate the measurement in an inconsistent way.  It may not be possible to do this with a time series with VPHOT; Ken or George can give you some guidance.

- use an analytical means to split the two stars.  One common method is psf-fitting, where an analytical model of a star profile (like a Gaussian) is fit to both the variable and its companion simultaneously, thereby giving good photometry of both stars.  This works best when both stars are well exposed (high S/N), and you have a very fine pixel scale that gives several pixels across the FWHM.

- use an analytical method called image subtraction.  This technique takes a reference image, and then subtracts each image of the time series from the reference image, scaled so that the background stars are cleanly subtracted out, leaving just the difference of the variable's brightness between the reference image and the time series image.  This is a very effective technique, but more complex than the other methods listed above.

To get good magnitudes for the companion, you could use a good-seeing telescope such as TMO61 of the AAVSOnet robotic telescope network.  It often has seeing around 1arcsec.  If you are an AAVSO member, you can just submit an AAVSOnet proposal to get an image set or two.

Arne