AAVSO: Photoelectric Photometry Newsletter Volume 21, No. 1

Contributions to this Newsletter are gratefully received at any time. Please send them to: John Percy, Erindale Campus, University of Toronto, Mississauga ON, Canada L5L lC6; e-mail: jpercy@erin.utoronto.ca

Arthur J. Stokes

Janet A. Mattei has circulated the following sad message, shortly after the 2001 Fall Meeting:

I sadly inform you of the passing away of a long-time AAVSO member Arthur (Art) Stokes, from a heart attack, on Nov. 6, 2001. Art and his family had just come back from the recent 90th Anniversary meeting of the AAVSO held this past weekend.
Art had been a member of the AAVSO since 1962; he served in the AAVSO Council for several terms, and was the AAVSO President in 1981 and 1982. He pioneered the establishment of the AAVSO Photoelectric Photometry Committee and was its chairman for over ten years. He was the recipient of the AAVSO Merit Award in 1987. In recent years he had been contributing very valuably to the AAVSO Solar Division, designing and building SID (sudden ionospheric disturbance) receivers for solar observers and editing the SID Technical Bulletins. He will be greatly missed in the AAVSO.
Funeral services for Art will be held in Hudson, Ohio on Nov. 10, 2001, Saturday, at 10:30 am. Art's family has requested that in lieu of flowers contributions in his memory be made to the AAVSO or the Emergency Medical Services of Hudson.
Other AAVSO members responded through the AAVSO electronic discussion group, mentioning the many ways in which Art contributed to their lives, through his wisdom, his mentoring, and his company.

Where the Editor Went This Summer

As usual, the editor's summer plans included one or two conferences on variable stars and/or education. At variable star conferences, I aim to present the research results which my students and I produce, often using AAVSO data. There were two variable star conferences in the summer of 2001 in Brussels and Leuven, Belgium.

The larger and more general one was International Astronomical Union Colloquium #185, on Radial and Non-Radial Pulsation as Probes of Stellar Physics. These ``pulsation conferences" are generally held every two years; the next is in New Zealand, in 2003. This meeting was held in the historic university town of Leuven, famous for its mediaeval and renaissance architecture. The University of Leuven had just celebrated its 575th birthday. For this reason, and because Belgian astronomers have made notable contributions to our understanding of stellar pulsation, Leuven was an appropriate (as well as enjoyable) venue for this meeting. The meeting was dedicated to the memory of the late Paul Ledoux, one of the founders of stellar pulsation theory. The meeting also honoured Paul Smeyers on the occasion of his retirement; Paul has been an active researcher in stellar pulsation theory, and mentor to dozens of younger Belgian astronomers in the field of stellar astrophysics. For some images which convey both the serious side and the less-serious side of the meeting, see the web site:

http://www.ster.kuleuven.ac.be/~iau185/gallery.html

At this meeting, I presented a summary of my students' new results on pulsating red giants (see text below). Art Cox (Los Alamos National Laboratory), who gave the introductory lecture on ``Unsolved Problems in Stellar Pulsation", spent considerable time discussing the problems of pulsating red giants --- stars which have been studied extensively in the AAVSO visual and PEP programs, as well as by my students.

The second meeting was held in Brussels, immediately before the Leuven meeting. It was a workshop dealing with pulsating A and B type stars. Since my students who do not work on red giants tend to work on pulsating A and B stars, I was able to present a completely different set of results at this meeting! Most of this work deals with very small amplitude variables which are not suitable for the AAVSO observing programs. They are relatively normal-sized stars which pulsate in radial or non-radial modes with periods of up to a day or two. The combination of the {\it Hipparcos} epoch photometry database, and our autocorrelation analysis technique, are well suited for the study of these stars, and that's what we presented at this meeting --- me and six students.

PEP Variable Stars of the Month

In order to highlight some of the stars in the PEP program, and in order to provide some easily-available reference material on classes of stars well represented in the PEP program, and in order to help the AAVSO HQ staff with the very useful Variable Star of the Month section of the AAVSO web site, I have prepared articles on ``EU Del and the Small Amplitude Pulsating Red Giants" and on ``Gamma Cas and the Be Stars". You can find them in the /vstar/vsotm/ section of the AAVSO web site. Look under August 2001 and under October 2001. EU Del is the ``prototype" of the small-amplitude red variables (pulsating red giants), and has been observed for almost two decades by the AAVSO PEP program. Gamma Cas is the prototype of the Be stars --- B stars with emission lines in their spectra. The emission lines were discovered by Father Secchi, using a visual spectroscope, over a century ago. One useful consequence of this project is that we have added gamma Cas to the AAVSO PEP program, and improved the charts for the other Be stars in the program. So please consider observing these stars! The nature of the ``Be phenomenon" remains enigmatic, even today --- especially the connection (if any) between the short-period variability mentioned above, and the long-term variability which is studied in the AAVSO PEP program.

I thank Kerriann Malatesta for her help with this project.

P Cygni 2000

The proceedings of P Cygni 2000: 400 Years of Progress, edited by Mart de Groot and Christiaan Sterken, have now been published by the Astronomical Society of the Pacific, as Volume 233 in its Conference Series. Among the many interesting articles is one on ``Long-Term VRI Photometry of P Cygni", by John Percy, Trevor Evans, Greg Henry, and Janet Mattei. It utilizes the AAVSO's long database of PEP observations, and its even-longer database of visual observations of this famous star, as well as observations from a robotic telescope. The P Cygni conference marked the 400th anniversary of the discovery of the three-magnitude brightening of this star --- the prototype of luminous blue variable hypergiants. It now varies by 0.1 to 0.2 magnitude, on time scales of months to decades, and is a popular star in both the AAVSO PEP and visual programs.

Gravity Probe B

Gravity Probe B (GP-B) will test two predictions of general relativity by measuring the precessions of gyroscopes in an earth-orbiting spacecraft with respect to an external frame of reference. That frame will be defined by the guide star IM Peg = HR 8703. GP-B needs a guide star whose proper motion can be measured with better than 0.15 milliarcsecond/year accuracy with respect to an extragalactic frame which is defined by pointlike sources called quasars. These are radio sources whose positions are stable to the necessary accuracy. The guide star must also be a nearly pointlike radio source whose position can be measured to milliarcsecond accuracy. For more information about GP-B, see the web site

http://einstein.stanford.edu/
The radio results can be found at:
http://aries.phys.yorku.ca/~bartel/GPBresults.html

IM Peg = HR 8703 is an RS CVn star --- a binary system with a K2 giant star which shows a high level of "activity". This activity includes compact radio emission, whose position can be compared, using radio interferometers, with the extragalactic reference frame. IM Peg is a 24.65-day spectroscopic binary. Radio interferometers can actually detect the motion of the radio star in its orbit! The most recent optical study of IM Peg is by Berdyugina et al. (Astron. Astrophys. 360, 272 (2000)), which presents surface images of the star. The period of spot rotation is 24.73 days.

IM Peg varies in brightness because of the rotation of spotted regions into and out of our line of sight. This changing brightness, relative to the non-variable binary companion, or other stars or nebulosity within the 140" diameter field of view of the spacecraft's guide telescope, could possibly bias the "average" proper motion of the star. In addition, the observation of any relatively rare, perhaps once per year, short-lived, bright flares on the active giant, as have been seen on other RS CVn binaries, would also be quite interesting.

Astronomers at Harvard-Smithsonian Center For Astrophyiscs have requested our assistance in providing close monitoring of IM Peg. What is needed from our observers is monitoring of IM Peg regularly. GP-B is not planned for launch until 2002, but the monitoring should begin as soon as possible to test the procedure. The star will also be monitored with a robotic telescope, but this is in operation for only part of the year. (Usually no robotic observations are made from late January until early May nor from early July until mid September of each year. Observations in these periods would be especially valuable. If you are interested in participating in this project, please notify AAVSO Headquarters (aavso@aavso.org).

The AAVSO Near-IR Photometry Project

Another exciting new development is a plan for an AAVSO near-infrared photometry program. The idea for this program came out of a ``futures" planning session after the 2001 AAVSO Spring Meeting. Very few observatories do long-term near-IR monitoring of variable stars. Yet pulsating red giants such as Mira stars, and the small-amplitude red variables in the AAVSO PEP program, have most of their flux in the near-IR. The plan has been pushed forward through the effort of Doug West, and a small working group including Janet Mattei and John Percy and several other AAVSO members and observers. With support from the AAVSO, OPTEC Inc. is producing five prototype near-IR photometers, which will be tested by selected volunteer observers --- mostly on stars which are already on the AAVSO PEP program. In this way, the near-IR data can be combined with the V data which is already being obtained.

Pulsating Red Giants - New Results

J.R. Percy, A.G. Bakos, H. Dunlop, G.W. Henry, A. Hussain, L. Kassim, R. Kastrukoff, Z. Nyssa, J.B. Wilson

This is a summary of a paper which we presented at International Astronomical Union Colloquium #185 (see above). --- John Percy Abstract. We have (1) accumulated long-term (up to 15 years) photoelectric photometry of 76 bright pulsating red giants (PRGs) and related stars, with V amplitudes of 0.04 to ≥ 1 magnitude; (2) used a variety of techniques including autocorrelation to determine both ``short" radial pulsation periods, and long secondary periods; (3) determined pulsation modes using the Q-value method; the modes range from fundamental to second overtone, independent of temperature; (4) tested wavelet analysis on small-amplitude PRGs; (5) made a special study of PRGs with V amplitudes less than 0.1 magnitude; and (6) called further attention to the problem of the long secondary periods of these stars. Our studies provide a ``model" for the analysis and interpretation of larger, fainter surveys such as MACHO and OGLE.

Introduction
Mira, the first pulsating red giant (PRG), was discovered over 400 years ago, and many more Mira and semi-regular variables were subsequently discovered. Stebbins & Huffer (1930) published the first photoelectric survey of red giant variability; they found that detectable variability set in at about M0III, and increased with decreasing temperature. In the 1970's, Eggen published a series of papers in which he surveyed and studied large numbers of PRGs, and classified them as small-, medium-, and large-amplitude red variables (SARVs, MARVs, LARVs). In the 1990's, several groups identified ultra-small-amplitude (0.01 to 0.1 magnitude) red variables (USARVs); in particular, the comprehensive survey of GK giants by Henry et al. (2000) extended the USARV phenomenon to as early as K2III. The present paper deals almost entirely with SARVs and USARVs.

In a recent paper, Wood (2000) states that ``red giant stars are probably the least understood of all variable stars". Wood may be biased, but Art Cox, in his opening lecture at this colloquium, seemed to support this view, on the grounds that the structure, evolution, and pulsation of PRGs is dominated by convection --- a process which is poorly understood in astrophysics.

Data and Analysis

Our data come from three sources: the American Association of Variable Star Observers (AAVSO) Photoelectric Photometry program, the AAVSO visual program (for a small section of this paper only), and the robotic telescopes of Tennessee State University. See Percy et al. (1996, 2001ab) for further details.

We use light curves, Fourier analysis, and a form of autocorrelation analysis, as complementary tools for period analysis. The first two of these are widely used and well understood, but autocorrelation is less so. It is useful for identifying characteristic time scales of variability, especially in stars which are not strictly periodic. It determines the cycle-to-cycle behaviour of a star, averaged over all the data. When used with Fourier analysis, it can be useful in choosing between the true period and alias periods, when the data have seasonal (or nightly) gaps. Burki et al. (1978) published a simple autocorrelation algorithm in which the data are shifted sideways in time, and the goodness of fit is determined. We have recently implemented this algorithm for the study of PRGs (Percy et al. 2001c).

In our own algorithm (Percy et al. 1993), we calculate Δmag and Δtime for every pair of measurements, and plot the first (averaged in bins) against the second over an appropriate range of Δtime. Minima occur at multiples of the characteristic time scale "tau", and can be used to measure "tau". The persistence of the minima is a measure of the degree of periodicity. The height of the first maximum is a measure of the amplitude. The height of the minima depends on the average error in the measurements, and the degree of irregularity of the star. Our method cannot deal easily with complex multiperiodicity, though it can do so if there are two periods with a large ratio, which is often the case with PRGs. Our algorithm is similar to (but not as statistically elegant as) the ``variogram" method of Eyer & Genton (1999), but is not the same as ``classical" autocorrelation analysis (Scargle 1989).

Results

We have accumulated a large (76 stars) database of long-term (up to 15 years) photoelectric photometry of bright SARVs and USARVs, mostly M giants, but also a few class II and I stars, and chemically-peculiar stars. The first two dozen stars from the AAVSO Photoelectric Program were published by Percy et al. (1996). Another two dozen stars were published by Percy et al. (2001a). Long-term VRI photometry of 34 stars, obtained with a robotic telescope, was published by Percy et al. (2001b). In each case, periods and amplitudes were carefully determined by the three techniques mentioned above. The data are suitable for studying both the short-term and the long-term variability of PRGs; see Percy et al. (2001b) for 200-day and 5000-day light curves of 34 stars.

USARVs. In Percy et al. (2001d), we made a special study of 11 K5-M0 III stars (Henry et al. 2000) with amplitudes less than 0.1 mag. They have periods as short as 4.8 days, but their Q-values are similar to those of SARVs. Several also have multiple periods. We are presently analyzing an additional season of data on these stars, to investigate the consistency of their behaviour. Unlike Koen & Laney (2000), we do not find periods less than 10 days in later-type M giants. We believe that the short periods in their paper are spurious, and are due to the strong aliasing characteristics of the Hipparcos epoch photometry.

Pulsation Modes. Percy & Parkes (1998) determined pulsation modes through Q-values for a small set of SARVs which were studied by Percy et al. (1996); they found the stars to be pulsating in radial modes between the fundamental and the third overtone. We have now determined Q-values for 47 M giants with well-determined ``short" periods (some stars have two short periods): the radii are estimated from both spectral types and (V-K) colours; the masses are assumed to be 1.4 solar masses; the theoretical Q-values were taken from Ostlie & Cox (1986) which should be adequate for these ``warm" red giants. About half appear to pulsate in the fundamental, about a quarter in the first overtone, and about an eighth in the second overtone. The rest have Q-values larger than the fundamental; two are USARVs, one may be II-III class, one is W Boo if the period is 55 days (see below). There is no obvious correlation between pulsation mode and temperature.

Wavelet analysis can be used to track the changing amplitude and period of variable stars, and it has been used extensively on LARVs and MARVs (e.g. Szatmary et al. 1994). We have used the wavelet software on the AAVSO web site (www.aavso.org, Foster 1996). We have investigated its application to SARVs, especially to study possible amplitude variation and mode switching (Percy & Kastrukoff 2001). We first applied it to AAVSO photoelectric photometry and 5-day means of AAVSO visual photometry of the well-studied SARV EU Del, with a period of 62.5 days. We found that the method usually gave the 62.5-day period, but the ``most significant period" occasionally switched to a one- or two-cycle/day alias. This seemed to occur during seasons of irregularity, in which the measurements (and the adjacent seasonal gaps) could be adequately fitted by the alias period. We also applied wavelet analysis to W Boo --- a SARV which, according to Fourier analysis of AAVSO photoelectric photometry (Percy & Desjardins 1996) had switched between periods of 25, 35 and 50 days. The wavelet analysis confirmed that the most significant period clustered around values of 25±2, 35±2, and 55±2 days. We conclude that wavelet analysis of SARVs can be effective, even if there are seasonal gaps, as long as it is used with care and caution, in conjunction with light curves, autocorrelation, and Fourier analysis.

Long Secondary Periods --- whose cause is unknown --- have been previously been found in LARVs and MARVs. We find them in at least half of our SARVs. The finite length of our datasets may bias against finding long periods in some stars. The ratio of the long period to the dominant radial period is remarkably constant --- between 9 and 12. This is consistent with the ``parallel sequences" found by Wood (2000) in the period-luminosity diagram. The ratio does not appear to correlate with temperature. In EG And, the long period is exactly half the binary orbital period, so the long period appears to be ellipsoidal. However Hinkle et al. (these proceedings) seem to rule out binarity as a general cause of this phenomenon. Cummings (1999), on the basis of a spectroscopic and photometric study of southern PRGs, suggests rotation as the most likely cause. Wood (2000), in a cogent review based on observations of LMC PRGs, rules out dust, non-radial pulsation, and rotation as a cause. He suggests that, if binarity is not the cause, then a new convectively induced oscillatory thermal mode must be invoked. Long secondary periods occur in other types of pulsating stars, such as RV Tauri stars and the SRd variable 89 Her.

Future Plans
In the near future, we plan to: (1) extend our database to other SARVs in the literature with well-determined periods; (2) use other means of radius determination (such as angular diameters), and semi-empirical masses, to improve our Q-values; (3) use wavelet analysis, and other techniques, to study the nature of the mode switching, amplitude variability, and irregularity of SARVs; (4) seek clues to the nature of the long secondary periods.

Acknowledgements. JRP thanks NSERC Canada, and GWH thanks NSF and NASA, for research support. We thank the AAVSO, especially Howard Landis and Janet Mattei, for providing access to data. JRP thanks his Toronto co-authors, who are students at the senior high school and undergraduate level; variable stars are ideally suited for student projects (Percy 2000).

References
Burki, G., Maeder, A., & Rufener, F. 1978, A&A, 65, 363
Cummings, I. 1999, J. Astron. Data, 5, #2
Eyer, L., & Genton, M.G. 1999, A&AS, 136, 421
Foster, G. 1996, AJ, 112, 1709
Henry, G.W., Fekel, F., Henry, S.M., & Hall, D.S. 2000, ApJS, 130, 201
Koen, C., & Laney, D. 2000, MNRAS, 311, 636
Ostlie, D.A., & Cox, A.N. 1986, ApJ, 311, 864
Percy, J.R. 2000, in Amateur-Professional Partnerships in Astronomy, ed. J.R. Percy & J.B. Wilson, ASP Conf. Series, 220, 310
Percy, J.R., & Desjardins, A. 1996, PASP, 108, 847
Percy, J.R. & Kastrukoff, R. 2001, JAAVSO, submitted
Percy, J.R. & Parkes, M. 1998, PASP, 110, 1431
Percy, J.R., Ralli, J., & Sen, L.V. 1993, PASP, 105, 287
Percy, J.R., Desjardins, A., Yu, L., & Landis, H.J. 1996, PASP, 108, 139
Percy, J.R., Dunlop, H., Kassim, L., & Thompson, R.R. 2001a, IAU IBVS, #5041
Percy, J.R., Wilson, J.B., & Henry, G.W. 2001b, PASP, 113, 983
Percy, J.R., Hussain, A., Gomez-Forrellad, J.M., Garcia-Melendo, E. 2001c, IAU IBVS, submitted
Percy, J.R., Nyssa, Z., & Henry, G.W. 2001d, IAU IBVS, submitted
Scargle, J.D. 1989, ApJ, 343, 874
Stebbins, J., & Huffer, C.M. 1930, Publ. Washburn Obs., 15, 138
Szatmary, K., Vinko, J., & Gal, J. 1994, A&AS, 108, 377
Wood, P.R. 2000, PASA, 17, 18

In addition, the project work of another student, Nathan Leigh, was presented at the 2001 annual conference of the Canadian Astronomical Society, in Hamilton, Ontario, in May. His project was on ``Autocorrelation Analysis of RV Tauri Stars". Most of the data came from the MACHO database, which is a massive database of variable star photometry which was gathered as a by-product of the (successful) search for massive compact objects in the halo of our galaxy, using gravitational micro-lensing. However, AAVSO visual and photoelectric photometry of RV Tauri stars was used successfully to test the method.

What Causes the Long Secondary Periods in Pulsating Red Giants?

As mentioned above, one of the more interesting results of the AAVSO PEP study of small-amplitude red variables is the realization that over half of them have long secondary periods, about ten times the short pulsation period. Also as mentioned above, one hypothesis is that the stars are members of binary systems, with binary periods about ten times longer than the pulsation period. Why? Presumably because both periods are, in some way, connected with the radius of the pulsating red giant.

In a recent publication, my colleague Petr Harmanec (Charles University, Prague, Czech Republic) included a table of binary systems with hot components. I noted that, in 17 of these systems, the other component was an M-type giant. In most cases, the pulsation periods of the M-type giants are not known, so I assumed they were the same as the pulsation periods of stars in the AAVSO PEP program with the same spectral types. When I did this, I found that, in 10 out of 17 stars, the ratio of the binary period to the estimated pulsation period, was close to 10. In 13 out of 17, it was between 6 and 15. This is suggestive, but a lot more work has to be done, to observe and analyze these systems in more detail, and to learn how the binary nature causes the long-period variations.

What My Students Are Doing This Year

As usual, I have several undergraduate and senior high school students working on research projects:

Akos Bakos is investigating the pulsation modes in a large sample of small-amplitude red variables; the periods were determined from AAVSO and robotic telescope photometry of these stars.
Gurtina Besla is using AAVSO and robotic telescope photometry to try to understand the apparent irregularity of small-amplitude red variables. Is it true irregularity, or the result of several different simultaneous periods in the stars?
Chris Harlow is using our autocorrelation technique to ``mass-produce" the determination of the short periods in Be stars (mentioned above), using data from the Hipparcos epoch photometry database. The long-term variations of several Be stars are being studied through the AAVSO photoelectric photometry program.
JoAnne Hosick is using robotic telescope data to study the variability of two high-luminosity B-type supergiants --- HR 8020 (HD 199478) and $\alpha$ Cam. The first star is being studied in collaboration with Dr. Nevena Markova, from Bulgaria. It is an extension of the P Cygni project described above.
Farisa Mohammed is using our autocorrelation technique to study known beta Cephei stars --- hot stars which pulsate in radial modes, with periods of a few hours --- and search for new beta Cephei stars in the Hipparcos epoch photometry database.
Emily Redelmeier is continuing Ryan Kastrukoff's project, to use AAVSO wavelet analysis software to study the changing periods and amplitudes of small-amplitude red variables, using AAVSO and robotic telescope photometry.
..... and two keen senior high school students are eagerly awaiting the assignment of their projects.

search engine | site map | links | contact us

Photoelectric Photometry Newsletter

Volume 21, No. 1, November 2001

Contents: