Photoelectric Photometry Newsletter

What My Students Did This Summer

89 Herculis (V441 Her, HR 6685, HD 163506) is a fifth-magnitude F2Ia "supergiant", noteworthy as it is one of the few to be at high Galactic latitude (though some believe that it is a low-mass star which is "masquerading" as a supergiant).

The data available for the present analysis consisted of 567 AAVSO photoelectric V observations, and 2229 VRI observations from the "Fairborn-10" APT in Arizona. The data covered the interval JD 2447000 to 2451360. The errors of the observations were typically better than 0.01 magnitude.

Fourier analysis (Figure 1) was performed with the AAVSO program TS (available at www.aavso.org/software.shtml) The strongest peak in the power spectrum was at 65.2 days. A similar analysis, yielding the same result, was conducted using the program Period98 (dsn.astro.univie.ac.at/period98).

A noteworthy discovery was the appearance of a peak at around 283 days, appearing in analyses of all three passbands ( VRI). This period had already been detected in radial velocities of the star, and ascribed to binary motion.

Phase diagrams were found to be particularly useful in the further analysis of the data. To simplify matters, averages were taken in 20 phase bins. For both the 65.2 and 283 day periods, the V and (V-I) variations were in phase. The light amplitudes were such that delta V > delta R > delat I.. The V amplitude for the 283-day period was actually greater than that for the 65.2-day period, so it is strange that the 283-day photometric period had not been observed before.

Further analysis was performed with (O-C) diagrams, using the 65.2-day pulsational period. Maxima were extracted from the present data, as well as from the literature. Although there are several short gaps in the data, which result in some uncertainty in the cycle numbers for the maxima, the (O-C) diagram appears to be a parabola, opening upward, which indicates an increasing pulsation period.

The 65.2-day period was not as pronounced after JD 2450200, whereas the 283-day period continued to be prominent. Continued observation and analysis of V441 (89) Her would be beneficial, especially an analysis of the short-term period from JD 2450200. Additionally, it would be of great use to obtain further times of maxima, where the cycle numbers are precisely known, so as to remove some of the ambiguity regarding the long-term change in period.

Figure 1. The power spectrum of the V observations of 89 Her, showing peaks at periods of 65.2 and 283 days.

We thank the AAVSO photoelectric observers, and Greg Henry at Tennessee State University for their data. We also thank the Natural Sciences and Engineering Research Council of Canada, and the Summer Career Placement Program, for their support.

Rho Cas, by David Kolin

Rho Cas is a peculiar SRd variable. SRd stars are a mixed bag of stars, that usually have long periods ( >30 d) and large changes in amplitude. The category contains some low-mass Population II stars, as well as some Population I hypergiants. Contemporary research has shown the possibility of chaotic variations in the light curves of SRd stars. Although SRd's are rarely observed now, their importance will likely increase in the future.

The variability of rho Cas was first observed by L. D. Wells in 1900. 643 observations from the AAVSO's photoelectric photometry program, and 1597 observations from the 10 inch ("Fairborn-10") APT were examined (Figure 2). From JD 2445641 to 2451361, the star's behaviour was semi-regular, with variations from 4.4 to 4.9 magnitudes. The APT data also contained measurements of the I band, which allowed for the calculation of (V-I) colour observations. The colour of rho Cas changed from 1.0 to 1.4 magnitudes. Interestingly, the colour of the star changed in tandem with the brightness. Previous investigations of rho Cas estimated its period to be 298.5 days (Percy and Zsoldos 1991 A&A 246, 441). However, in light of more recent data, this estimate needs to be revised.

A program called Period98 was used for Fourier analysis which yielded several peaks, the highest of which corresponded to a period of 815 d. Noteworthy peaks were also present at 382 d, 645 d, and 512 d. An autocorrelation analysis was performed using the program Astrolab. The minima in the autocorrelation graph did not suggest a clear period; an 800 d period was most prominent, and a 500 d variation also appeared to be present. Phase diagrams of all of the previously mentioned periods had a lot of scatter and were unpersuasive.

The most useful information was gained by examining the light curve. At the beginning of the light curve, rho Cas underwent a relatively long cycle of 660 d, and reached its maximum at about JD 2446186. The star then went through a series of shorter cycles, until approximately JD 2448000. (It was from these cycles that Percy and Zsoldos measured the period to be 298.5 d.) After these shorter cycles, the star reverted to longer cycles of about 800 d. However, these 800 d cycles contain interesting variations within them. Fourier analysis supports the presence of these longer and shorter cycles. When the observations before JD 2448000 were analyzed, the strongest peak was at 294 d. An analysis of the observations after this date yielded a strongest peak at 763 d. Continued analysis of rho Cas is important. It will be interesting to see how the star behaves in the future.

We would like to thank the AAVSO photoelectric observers, and Greg Henry at Tennessee State University for their data. We are also grateful to Employment Canada for its support of the Summer Career Placement Program. David Kolin participated in the University of Toronto Mentorship Program, which allows senior high school students to work on research projects at the University.

Figure 2. The long-term V light curve of rho Cas, based on AAVSO and APT photometry.

Small-Amplitude Red Variables, by Joseph B. Wilson

For the past few months, I have been working for Dr. J.R. Percy at the University of Toronto, analyzing data from the Automatic Photoelectric Telescope Fairborn-10 in Arizona. The data consists of V, R, and I band observations of 34 red variables that have amplitudes under 2.5 magnitudes - too small to be considered Mira variables. Most of the stars in the current dataset are classified by the General Catalogue of Variable Stars as SRa or SRb. The Mira stars have periods longer than 90 days, while these small-amplitude stars have periods as short as 25 days. What is interesting is that the majority of these stars are multiperiodic, and have at least one long-term period at about 10 times the length of the short one. A second phenomenon observed with these red stars is the tendency for their periodicity to change over time - sometimes very quickly - to periods that can be drastically different from the previous one.

I analyzed the Fairborn-10 data using light curves, autocorrelation and Fourier analysis methods to determine any patterns in this typical multiperiodicity. Figure 3 shows a light and color curve for EU Del, which could be considered a prototype for the small-amplitude red variables (SARV's). Note the amplitude of 0.7 magnitude, and the period of about 62 days. EU Del appears to be mono-periodic; this analysis did not turn up any evidence of long-term variability. The 62-day period appears to be fairly consistent throughout the 12 years of data taken from the telescope. This is rather uncharacteristic for these stars.

Theorists have suggested that the SARV's and Miras are similar kinds of stars, at two ends of a continuum with well-defined, large-amplitude periodicity at one end (the Miras), and low-level variability with poorly-defined periodic structure at the other (the multi-periodic SARV's). There are some stars in our sample with up to five periods, as revealed by Period98, the Fourier analysis program which we used. [This program, and instructions on how to use it, are on the WWW at dsn.astro.univie.ac.at/period98.

There is clearly a lot more work to be done on these stars. This same dataset can be interpreted more thoroughly when some astrophysical work is done in determining the pulsation constants and pulsation modes for these stars. The results could lead theoreticians to further conclusions about the nature of SARV's, and their relationship to other classifications of red variables - Mira, SRa, SRb, and Lb variables.

Figure 3. The light and (V-I) color curve for the prototype small-amplitude red variable EU Del. Note that the light and color curves are in phase.

Editor's Note. Long-term AAVSO photoelectric photometry of about two dozen small-amplitude red variables was analyzed by Percy et al. (1996, Publ. Astron. Soc. Pacific 108, 139-145). This study provided an excellent model for analysing the APT data - JRP.