by Rex Parker, AAAP
Variable stars are not getting the attention they deserve in the AAAP these days! Deep sky objects, planets & exo-planets, and intergalactic large scale structure are the hot topics. Even double stars get more telescope time at the observatory. Visually observing the changes in star brightness is slow and perhaps difficult. Yet the science behind variable stars is fascinating, and the photomultiplier tube (PMT) detector has been displaced by the CCD as instrument of choice for their measurement. A CCD camera on a telescope with a good tracking mount can accurately measure stellar magnitudes. It occurred to me that for variables, the real obstacle is to actually set up and do the experiment.
Algol (β-Persei) is a variable star 92 light years away in the constellation Perseus which undergoes periodic variations in brightness every ~2.87 days. It is the most famous example of an eclipsing binary star system; when the dimmer star moves in front of the brighter (in our line of sight) the overall intensity decreases markedly. November is the only time when the full light curve of Algol can be observed at our latitude. This is because Algol stays at its minimum for ~ 2 hours and takes about 10 hours to complete the eclipse.
What is needed is a clear November night without a full or gibbous moon, and a minimum occurring around midnight. Sky & Telescope publishes the “Minima of Algol” monthly in the magazine and website. I had been keeping an eye on these tables and noticed a minimum predicted for Friday night (11/13) at 12:52 AM.
Non-anti-blooming CCD chips are linear photometric devices. The wells (pixels) of the chip fill with electrons in linear proportion to the photon flux, as long as the full well capacity is not exceeded (controlled by exposure time). For the data below, images at 0.2 sec each were taken with SBIG ST10-XME and Tak FS-128 at f/8 on a Losmandy G11 mount, with images taken every 5 minutes all night long and downloaded in real time to my laptop. The data was processed the next day using a program (in Maxim DL) that compares the brightness of the target star to an invariant control in the same image field whose magnitude is precisely known (in this case, from NASA’s Hipparcos). The reference star in this case was mag 8.6, so that probably gives rise to some of the noise in the plot (exposures had to be short to keep Algol’s pixels within full well capacity of CCD). The resulting data are normalized brightness, given as magnitude, which allows construction of a light curve plot of magnitude vs time (figure below). Recall that magnitude is a holdover from historical units in which each magnitude is 2.5 times brighter than the next, and the span of all stars visible to the naked eye are 1 to 6 (1 is brightest).
At last, a minima of Algol near midnight on a clear Friday night in November when I was home!