Photoelectric Photometry Newsletter

What My Students (and I) Did Last Summer

This was the year when I had to justify my existence to the Natural Sciences and Engineering Research Council of Canada, which generously funds my research. [By the way - a couple of years ago, the NSERC bestowed its "Friend of NSERC" citation upon the AAVSO, its photoelectric archivist Howard Landis, and its active Canadian observer Ray Thompson, in recognition of their contributions to science.] I therefore thought it was time to finish up, and write up all of the interesting projects which my students and I have recently done, mostly with AAVSO and Hipparcos data. These students range from fourth-year university students doing a "senior thesis" to bright high-school students in the University of Toronto Mentorship Program.

Ted Colivas (Percy & Colivas 1999, PASP, in press) studied 75 years of times of maximum of almost 400 bright Mira stars (determined from AAVSO visual observations), and discovered that the (0-C) diagrams of almost all of these stars were dominated by the effects of random, cycle-to-cycle period fluctuations of unknown cause. These fluctuations make it difficult (though not impossible) to detect evolutionary changes in these stars.

Winnie Au (Percy & Au 1999, PASP, in press) used this same database to search for evolutionary period changes in this same set of Mira stars, underneath the fluctuations mentioned above. We believe that we have found them within the "noise"; their magnitude is in good agreement with evolutionary models. These results are possible only because of the existence of many decades of AAVSO visual observations, and because of the foresight and dedication of the AAVSO and its observers.

Dalia Hanafi-Bagby (Percy & Bagby 1999, PASP, in press) used the times and magnitudes of maximum in this same database to look for evidence of multi-periodicity in these stars. We find some with evidence for long-term variations, superimposed on their Mira-type variations, and some which may be pulsating in two adjacent radial modes - the dominant one likely being the first overtone. [Another benefit of long-term visual or photoelectric observations of pulsating stars is the ability to detect multiple periods in these stars.]

Metin Guler (Percy & Guler 1999, JAAVSO, in press) used Hipparcos epoch photometry to investigate the onset of variability in red giants. It was previously known - partly as a result of the AAVSO PEP program's Project SARV- that noticeable variability sets in at spectral type MOIII. High-precision photometric surveys have found micro-variability in K5-K9III stars as well. It is not clear whether this micro-variability is the same as the radial pulsational variability found in M giants as well (see below).

Mike Parkes (Percy & Parkes 1999, PASP, in press) used new Hippar- cos distances, and radii derived from interferometry and lunar occultations and from temperatures derived from model atmospheres, to determine the pulsation modes of small-amplitude red variables (SARV's) with spectral types Ml-5III - mostly stars for which the AAVSO PEP program has determined periods. These stars are pulsating in various modes, from the fundamental to the third overtone. This appears to be one reason why SARV's have such a wide range of periods - from 25 days to over 200 days.

Jonathan Hale continued an ongoing study of period changes in RV Tauri stars, using a variety of existing data. We have been searching (with mixed results) for evolutionary period changes in these stars. Like the Miras and RU Cam mentioned above, however, there are random changes which mask the evolutionary ones, and I am not quite sure whether we have observed the evolutionary changes or not.

Yvonne Tang's analysis of AAVSO photoelectric photometry of RU Cam was reported in the last issue of this Newsletter. Jonathan Hale continued this analysis, and showed that RU Cam shows random, cycle-to-cycle fluctuations in period, and probably multi-periodicity. Before its dramatic decrease in amplitude in the mid-1960's, however, evolutionary period changes were observable in this star; now they are "buried" under the random fluctuations.

Margarita Marinova spent the summer analyzing Hipparcos epoch pho- tometry of small-amplitude red variables, and writing a manual on How to Analyze Variable Star Observations . We immediately discovered that the Hipparcos observations are not well-distributed in time. There are groups of observations only a few hours apart (not much use for studying variables with periods of weeks), and these groups tend to be separated by 20-30 days - again, not well suited to the task at hand. She made a special study of the SARV AG Cet, which was described above.

Margarita Marinova also completed a very interesting study of the Be star variable NW Ser, also using the Hipparcos data. These data are much better-suited for studying Be star variables, because these stars vary on time scales of hours, and weeks to months. The former variations can be studied with the closely-spaced Hipparcos data; the latter variations can be studied with the widely-spaced data. The results are in press (Percy, Marinova, Harmanec & Bozic 1999, A&A, in press). [Petr Harmanec and Hrvoje Bozic are my long-time collaborators in the study of Be star variables, including the ones on the AAVSO PEP program.]

Much of our work makes use of the Hipparcos epoch photometry. This has some advantages but, because of its non-standard wavelength coverage, it is difficult to compare it with ground-based photometry. There is also the Tycho photometry collected on the same mission; it is less accurate, but includes measurements which are close to the standard B and V system.

Joe Wilson is using the above-mentioned closely-spaced Hipparcos data to search for a long-suspected (but never-confirmed) group of variables called the "Maia Variables" (after the suspected prototype in the Pleiades. These variables - if they exist - have spectral types of B7-A3IV-V and periods of a few hours.

Katrina Au-Yong is also using the above-mentioned closely-spaced Hip- parcos data, to search for,beta Cephei pulsating variables in eclipsing binaries where the masses and radii of the components are known. There are very few pulsating variables whose masses and radii are known, and whose pulsation mode and internal structure can therefore be determined. The beta Cephei stars have spectral types of BO.5-211-IV and periods of a few hours. Most of the known examples are very bright "alpha" and "beta" stars in their constellations.

Jonathan Hoss is beginning a study of period changes in Population II Cepheids, to see if we can detect evolutionary changes and/or random, cycle-to- cycle period fluctuations as we did for RU Cam (see above). The Population II Cepheids are believed to be in a relatively rapid phase of evolution. If we can find several decades of times of maximum, then we should be able to "see" the evolution of these stars.

David Kolin is beginning a study of the photometric variability of SRd variables - yellow giants and supergiants with semi-regular variations, but without the characteristic alternating deep and shallow minima which would qualify them as RV Tauri stars. We will use Hipparcos epoch photometry, and AAVSO visual observations (if available). We hope to investigate such questions as: Do these stars vary on both short (weeks) and long (years) time scales, like the RV Tauri stars? What mode(s) are they pulsating in?. Can we detect evolutionary changes?

Liz Nelson is looking for random, cycle-to-cycle period fluctuations in classical (Population I) Cepheids. The (0-C) diagrams of some classical Cepheids suggest that such fluctuations are present. They have already been found in Mira stars, RV Tauri stars, and the Population II Cepheid RU Cam. How do these fluctuations depend on properties of the stars such as radius, gravity, presence of convection cells in the atmosphere? Liz is also an avid observer, so she will be observing some Cepheids visually as well.

Devi Soondarsingh is analyzing the AAVSO photoelectric data on three RS CVn stars - HK Lac, SZ Psc, and lambda And. RS CVn stars are sun-like stars with high activity levels - spots, prominences, flares, and X-ray emission - which are caused by their rapid rotation. The rotation is usually due to the presence of a close binary companion, whose tides "spin up" the RS CVn star. The starspots show up as light variations with the same period as the rotation. Many of these stars are also eclipsing binaries.