PEP Newsletter v22n1

Photoelectric Photometry Newsletter

Volume 22, No. 1, January 2003

Editor: John R. Percy

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 1C6; e-mail: All material in this Newsletter has been written by the Editor unless otherwise indicated.


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 1C6; e-mail: All material has been written by the Editor, unless otherwise indicated.

We extend our sincere sympathy and support to our colleagues in Australia, where the Mount Stromlo Observatory was recently destroyed by bush fires. Mount Stromlo was not only of immense historical importance, but it was actively involved in research, and in the construction of instrumentation for the Gemini telescopes. Observatories near to their universities have special capabilities: I recently received a preprint from Peter Wood which used the Mount Stromlo 1.88m telescope for long-term spectroscopic study of Z Eri -- a star on the AAVSO PEP program. Only ``local" observatories can easily achieve such long-term coverage. And the Mount Stromlo 1.88m telescope is a ``twin" of the 1.88m telescope here at the University of Toronto, so we were especially moved by this astronomical tragedy.


I recently attended the 201st meeting of the American Astronomical Society, in Seattle WA, 5-9 January 2003. There were literally hundreds of research papers presented. Here are a few which may be of interest to photoelectric observers.

The Millenium Outburst of the Yellow Hypergiant Rho Cas

by A. Lobel, A.K. Dupree, R.P. Stefanik, G. Torres, G. Israelian, N. Morrison, I. Ilyin, C. de Jager, H. Nieuwenhuijzen.

Rho Cas is one of the brightest (V=4.6) and most popular stars on the AAVSO PEP program. It is a yellow hypergiant which is one of the most luminous stars in our galaxy. Long-term AAVSO and APT photoelectric photometry was recently analyzed by J.R. Percy, D.L. Kolin and G.W. Henry (2000 PASP 112, 363) who found several periods, the most prominent being 820 days, but the behaviour of this star changes noticeably with time.

Lobel et al. report that, in the fall of 2000, the star dimmed by more than a magnitude, and the spectral type changed from early F to early M -- a change in temperature from 7000K to 4000K in less than 200 days. The star is also losing mass at the prodigious rate of 0.05 solar masses per year, the highest rate directly observed in any stellar object. This is the third such outburst in the last century; the pervious ones were in 1946-7 and 1985-6. They state that ``The recent observations may therefore signal that a new and stronger outburst of $\rho$ Cas is imminent". Photoelectric observers: be sure to measure this star every week or two!

This paper made it to USA Today, Wednesday January 8, 2003, under the headline ``Dying Star's Huge Eruptions Analyzed", with an accompanying graphic captioned ``Hypergiant -- fascinating scientists since 1946".

The details of this paper have now been published in the Astrophysical Journal, 583, 923-954 (1 February 2003). We are presently analyzing recent (2000-2003) photoelectric observations of this star.

Click image to enlarge.
Figure 1: The photoelectric light curve of Rho Cas, based on AAVSO and APT photometry. Top panel: V magnitude; bottom panel: (V-I) colour. From Percy et al. PASP 112, 363 (2000).

The Scale of Granulation in Red Giant Atmospheres

by P. Demarque, F.J. Robinson, K.L. Chan, Y.-C. Kim, D.B. Guenther, S. Sofia

The majority of the stars in the AAVSO PEP program are pulsating red giants. These stars have one or more radial pulsation periods, typically 20-100 days. In addition, many of them have long secondary periods, an order of magnitude longer than the radial periods. The cause of these long secondary periods is uncertain. One hypothesis, first put forward by Irene Cummings (University of Canterbury, New Zealand), is that red giants have large, bright (or dark) starspots which produce light variations as the star rotates. These spots could be convection cells. It was suspected that, on cool giant stars, the convection cells were large and few, compared with the small, numerous convection cells (granules) on the sun.

Demarque et al. have carried out three-dimensional radiative hydrodynamical simulations which suggest that the convection cells on red giants are smaller and more numerous than previously suspected. In this case, they would not produce any significant brightness variations as the star rotated, since the effects of bright and dark cells on the visible hemisphere of the star would cancel out. So rotation is no longer a straightforward explanation for the long secondary periods. Which leads to .....

Model Atmospheres for Irradiated Red Giant Stars with Winds

by J.P. Aufdenberg and T.S. Barman

This paper presented exploratory model atmosphere calculations applicable to ``symbiotic binary systems" where a hot white dwarf illuminates the extended atmosphere of a red giant. The model was applied to EG And -- one of the stars on the AAVSO PEP program. The cool star in this binary system is an M2-type red giant, with a 29-day pulsation period. There is also a 242-day long secondary period, but this is exactly half of the orbital period of 483 days! So the long secondary period must to connected with the binary nature of this star. Are all long secondary periods due to binarity?

A New Class of Pulsating Stars: Gravity-Mode Pulsators among Subdwarf B Stars

by E.M. Green and G. Fontaine

It's always exciting when a new class of variable stars is announced. In this case, the class is not entirely new. Subdwarf B stars are hot stars, smaller and less luminous than normal B stars, which are approaching their eventual fate as white dwarfs. Some of them are already known to vary on time scales of a few minutes; they are called EC 14026 stars, after the prototype. The variations of these are thought to be due to ordinary radial pulsation.

The EC 14026 stars are interesting because their pulsational variability was ``discovered" independently by theoretical calculation (by Charpinet, Fontaine, and Brassard), and by a team of observers at the South African Astronomical Observatory. Now, Green and Fontaine have discovered, in several subdwarf B stars, variations on a time scale of about an hour -- about ten times greater than in the EC 14026 stars. They are suspected to be non-radial pulsators, but the actual driving mechanism for the pulsations is not yet known. This study has now been published in Astrophysical Journal Letters, 583, L31-L34 (2003 January 20).

[If you were thinking carefully about the items above, you might ask: could the long secondary periods in red giants be due to non-radial pulsations? Unfortunately, the periods are not quite right, and non-radial pulsations tend to have very small light amplitudes -- an order of magnitude smaller than observed in red giants.]

[Or maybe you were wondering what is the difference between radial and non-radial pulsations? In radial pulsations, the motions are strictly radial, or in-and-out. All layers of the star which are the same distance from the center move in the same way. This is not true of non-radial pulsation. The poles of the star could be moving in, while the equator of the star was moving out. The equator of the star could be divided into an even number of zones such that the even-numbered zones were moving in while the odd-numbered zones were moving out.]

``But I Am Constant as the North Star": The Return of Polaris as a Low-Amplitude Classical Cepheid"

by J.J. Davis, J.C. Tracey, S.G. Engle and E.F. Guinan

Polaris is a Cepheid pulsating variable star with a period of 3.97 days. It is the nearest and brightest Cepheid (but, contrary to popular belief, it is not the brightest star in the sky). Its other claim to fame is that, during the last century, its amplitude has decreased to almost zero!

This paper reports recent photometry of Polaris. The V full amplitude is now 0.028 mag, slightly larger than was observed in the early 1990's. It appears that the century-long decrease in amplitude has halted, or at least paused. The pulsation period of Polaris continues to increase at +3.2 seconds per year.

By the way: for many telescopes, Polaris is not an easy star to observe, situated so close to the north celestial pole!

I presented a paper on Multiperiodicity and Amplitude Variations in Five Small-Amplitude Pulsating Red Giants, based on work done by my students Gurtina Besla and Vince Velocci. This work was described in the previous issue of this Newsletter. One of my undergraduate students -- David Hou -- is analyzing several more pulsating red giants, to see whether multiperiodicity and amplitude variations occur in them also.

One of the impressive aspects of this meeting was the large number of papers which were based on undergraduate research -- about 100 in total. That number includes only papers in which the first author is an undergraduate (and must therefore be present at the meeting). My paper was also based on undergraduate research but, since I was the author who was present and presenting the paper, my name had to go first. So there may have been 150 or 200 papers, in total, with undergraduate co-authors.

There was, in fact, a special session on Undergraduate Research, which I co-organized. It dealt less with the actual research done, and more on the ways of funding and supporting it. Three undergraduates discussed their experience: how research had enabled them to learn not just astronomy, but how to do astronomy, and what astronomy was like as a career. With archival variable star data from the AAVSO, and from today's massive sky surveys for gravitational lenses, and exoplanets (see below), there is more than enough variable star data to keep hundreds of undergraduate research students busy!

One of the two or three most newsworthy stories from the meeting was the first discovery of an exoplanet -- a planet around another star -- by the ``transit method". A hundred exoplanets have been discovered in the last decade, in each case by observing their tiny gravitational effect on their star. The star moves in a tiny orbit, with the same period as the planet's orbit. The star moves at a speed of typically a few tens of meters a second in its orbit, and this can be detected and measured by powerful spectrographs on large telescopes, using the Doppler effect. This work has produced a remarkable and unexpected result: there are Jupiter-mass planets whose orbits are ten times smaller than Mercury's, with orbital periods of a few days. In one case, the exoplanet, in its orbit, passed in front of its star -- an event called a transit -- and the light of the star dimmed by about 2 per cent.

But can exoplanets be discovered by monitoring millions of stars for slight, periodic dips in brightness due to transits? About 20 groups (including one at my own university) are undertaking projects of this kind.

At the AAS meeting, one such team announced its discovery, using the transit method but confirmed with the Doppler-effect method, that a faint star called OGLE-TR-56 has a Jupiter-sized planet which orbits only a few million km from its star -- 14 times closer than Mercury is to the sun -- with a period of only 29 hours! OGLE-TR-56 is 5000 light years away -- by far the most distant exoplanet yet discovered.

One of the invited review talks at the meeting was by Sara Seager, now at the Carnegie Institute of Washington, on the nature of exoplanets. If an exoplanet is so close to its star, its atmospheric temperature must be very hot. What will the atmosphere be like? Cloudy? Windy? And can we find ways of studying the nature of these distant planets, and even finding ones with life? [Incidentually, Sara Seager was an undergraduate at the University of Toronto, and worked on several photometric research projects with Don Fernie, including a comprehensive study of the variability of Polaris, mentioned above.]

One of the exciting things about the transit method is that it could in principle be used by a backyard astronomer with a small telescope to discover an exoplanets. Already, one of the (relatively bright) exoplanets discovered by the Doppler-effect method was observed, with a small telescope, to transit.


One of the mysteries of pulsating red giants is the nature and cause of the long secondary periods in some of these stars. Could they be due to binarity? There is a whole class of red giants in binary systems -- the Z And, or symbiotic stars -- which might provide a clue.

Symbiotic stars are defined as stars whose spectra show the simultaneous presence of a cool giant star, and a hot plasma. The hot plasma may be a hot main sequence or white dwarf star, or it may be a disc of material being accreted from the cool giant by a small companion star.

The orbital periods of symbiotic stars are typically 500-1000 days. The light curves usually show evidence of eclipses with this period. There are shorter-term variations as well. The cool giant should pulsate with a period of tens of days, though not many symbiotic stars have been carefully observed and analyzed for this kind of variability. There may also be rapid variations in brightness -- especially in blue or violet light -- which are due to the hot component.

There is one famous Z And star in the AAVSO PEP program: CH Cyg. The pulsation period of the red giant component is about 97 days. The binary period is suspected to be 5750 days, so it does not fit the ``binary period is ten times the pulsation period" relation. This star should continue to be monitored.

There is another enigmatic binary star in the AAVSO PEP program: SS Lep. Its components are A0Ve and M1III. The Yale Catalogue of Bright Stars classifies it as ``Z And?", states that it is a shell star (which would make it a relative of the Be stars -- note the A0Ve spectral type), and states that it ``varies like R CrB"! The light amplitude is given as 4.82-5.06 in the General Catalogue of Variable Stars, and at least some of this comes from the pulsation of the red giant. The most recent detailed study of this star, by A.D. Welty and R.A. Wade (Astron. J., 109, 327 (1995)), concludes that it is a pre-main sequence binary system (as suggested earlier by Polidan and Shore) with an A-type main sequence star and a low-mass M4III star, still contracting to the main sequence. The orbital period is 260.34 days. Every 2-3 months, the A-type star casts off a shell of gas. The orbit turns out to be almost circular. This star needs more monitoring -- even though it is rather far south for many of our observers.

The interesting connection between symbiotic stars and the long secondary periods of pulsating red giants is that, in many cases, the ratio between the long period (the orbital period in the symbiotic star, and the long secondary period in the pulsating red giant) and the pulsation period (the observed period in the pulsating red giant, and the expected pulsation period for the cool giant in the symbiotic star, given its spectral type) is about ten. Could there be a connection?

Unfortunately there is evidence that the long secondary periods are not due to binary motion. K.H. Hinkle et al. (2002 AJ 123, 1002) have measured radial velocities for several pulsating red giants with long secondary periods. They find radial velocity variations with the long secondary period, but the amplitudes, eccentricities, and orientations of the ``orbits" are very peculiar, and not at all like the orbits of bona fide binary systems. They conclude that the long secondary periods are due to some as-yet-unknown phenomenon.

An e-mail message from Doug West reminded me of the unusual nature of this star, which is on the AAVSO PEP Program. It is a spectroscopic binary with a period of 2984 days. Hinkle et al. (see above) accept that this is a bona fide orbit. The magnitude of Eta Gem varies with a period of about 233 days (so here is another case of a period ratio of about ten). This star is also a visual binary with a period of 474 years; the companion is just over an arc second away. It is also reported to be an occultation double with a separation of 0.03 arc seconds.

But there is something slightly peculiar, still. The 233-day photometric period of Eta Gem is considerably longer than the expected radial pulsation period for an M3III star; the expected period is about 55 days. In fact, there is some suggestion of a time scale of about 20 days in the light curve. So is the 233-day period really a pulsation period??

This star is also suspected to be an eclipsing binary, but the intrinsic variations of the M giant make it difficult to be sure.

The AAVSO's interesting and useful ``Variable Star of the Month" web article has now gone seasonal (or quarterly) -- not surprising in view of the considerable work which goes into these comprehensive profiles of some of the AAVSO's favourite stars. The first Variable Star of the Season is U Mon -- an RV Tauri star on the AAVSO visual and PEP programs; thanks to Kerri Malatesta for producing this article. RV Tauri stars are yellow supergiants which show alternating deep and shallow minima. Their periods (deep minimum to deep minimum) are typically 50-150 days. Their amplitudes are a magnitude or more.

U Mon is a member of an enigmatic subclass of RV Tauri stars called the RVb stars; they have long secondary periods of 600-2500 days. As with the pulsating red giants, the cause is not entirely clear, but there is increasing evidence that the RVb stars are binary stars. A dust disc around the non-RV Tauri component might somehow explain the long secondary periods. In U Mon, the light curve of the long secondary period looks suspiciously like an eclipse curve. Binarity might explain another strange property of some RV Tauri stars -- their atmospheres seem to be depleted of the chemical elements which form dust.

A recent paper ``RU Cen and SX Cen: Two Strongly Depleted RV Tauri Stars in Binary Systems: The RV Tauri Photometric b Phenomenon and Binarity", by T. Maas, H. Van Winckel, and C. Waelkens, Astron. Astrophys., 386, 504-516 (2002) concludes its discussion by stating ``only long-term multi-wavelength spectral and photometric monitoring will be able to study these enigmatic objects in detail".

----- And now for something completely different -----

You will remember, from past issues, that the AAVSO PEP program is participating in an exciting project to monitor IM Peg -- the guide star for the Gravity Probe B satellite. The purpose of this satellite is to make a sophisticated test of certain aspects of relativity theory. The latest estimate for the launch date for GP-B is 21 July 2003, but there is a possibility that launch may be delayed until September. In the meantime, here is an update on IM Peg, and a brief review of RS CVn stars -- the species to which IM Peg belongs.

An RS CVn star is used as a guide star because it appears starlike, and is both an optical source and a radio source. The positions of radio sources can be determined very precisely relative to a background of radio point sources -- namely quasars -- which provide an absolute frame of reference. IM Peg is being observed by very long baseline radio interferometry. Its resolving power is sufficient to observe the orbital motion of the radio source within the binary system!

The RS CVn stars have been defined by Douglas Hall as ``binaries with orbital periods between 1 and 14 days, with the hotter component F-G IV-V and with strong H and K line emission seen in the spectrum outside eclipse". There are similar stars with orbital periods less than a day (``the short-period group") and greater than 14 days (``the long-period group"). It is probably appropriate to lump all these stars together and call them ``stars showing the RS CVn phenomenon".

Individual members of this group have been known and studied for decades, but they first began to attract wider attention in the mid-1960's on account of their their unusual photometric variability outside eclipse. In the mid-1970's, even more exotic properties were discovered: radio emission and flaring, thermal X-ray emission indicating temperatures of 107K, and strong and variable calcium II H and K, hydrogen, and magnesium II h and k line emission. It is now accepted that these properties are due to ``stellar activity" -- starspot groups, a thick chromosphere, and coronal magnetic loops. In many ways, they are like solar activity, but on an even grander scale.

Many of the RS CVn stars show photometric variability other than that which may be due to eclipses. The variability consists of a more-or-less sinusoidal ``distortion wave" whose amplitude and phase (relative to the orbital period) slowly vary. The phase variability causes the distortion wave to migrate relative to the orbital period or -- if the system eclipses -- relative to the eclipse light curve. The distortion wave is thought to be due to the starspot groups: darker regions near the equator of the star. The rotation of the star, which is usually synchronous with the orbital revolution, then produces the distortion wave, as the darker regious are turned toward or away from the observer. The variability of the amplitude of the distortion wave is due to the changing area of the spots; this measure of the stellar activity changes in long-term cycles, as it does on the sun and on other sunlike stars. The variability of the phase of the distortion wave is due to the differential rotation of the star, which affects the position of the starspot region, relative to the companion star. The accompanying figure shows a schematic diagram and light curve of RS CVn itself.

Click image to enlarge.
Figure 2: Schematic light curve and schematic model of RS CVn. RS CVn is an eclipsing binary. The primary and secondary eclipses are shown. There is also a ``distortion wave" which extends through the light curve. Its range and phase can vary with time. In the schematic model of the RS CVn system, the sun is shown to scale. The distortion wave is produced by a dark, spotted region on the cooler star. Changes in the range and phase of the distortion wave are due to changes in the size and position of the spotted region. Source: light curve based on one by E.W. Ludington, published in Sky & Telescope, February 1979.

Why is the level of activity so high on RS CVn stars? Because stellar activity is caused by magnetic fields on a star, and magnetic fields are produced by the star's rotation. RS CVn stars rotate rapidly because tidal interactions with their companions has ``spun them up" to rotation rates several times faster than the sun.

Because of the complex photometric variability of RS CVn stars, it is desirable and necessary to monitor these stars systematically over many consecutive observing seasons. This requires extensive observing, and Douglas S. Hall in particular recruited the help of dozens of amateur astronomers with photoelectric equipment. Most RS CVn stars are bright. Their colours are not extreme. The amplitudes can be 0.2 magnitude or more. They are ideally suited for observation by amateur astronomers.

As mentioned in Phil Manker's report below, IM Peg has been observed 433 times by the AAVSO PEP program. About two-thirds of the measurements are by Lou Cox, who has been observing it several times a night to look for flaring activity. Our first priority is to get other others observing it, especially at different sites where it might be clear when Lou Cox' site is cloudy, and especially at different hour angles and longitudes so we will have continuous coverage of the star.

During the last year, the distortion wave of IM Peg has shown a period of 25 +/- 1 days, and a V amplitude of just over 0.2 magnitude. The rise from minimum to maximum is three times faster than the fall to minimum. No obvious flaring behaviour has been observed. There are significant gaps in the light curve, so let's put the pressure on this star!


There are several Be stars in the AAVSO PEP program; these are hot stars which have shown hydrogen emission in their spectrum on at least one occasion. The emission arises from hot gas in an expanding disc around the star. The ``problem" is -- what causes the disc to form and disperse?

In a recent paper, Petr Harmanec (Astron. Astrophys., 396, 937 (2000)) lists the various hypotheses: (i) it is flung off by the rapid rotation of the star, the rotation rate being unusually large; (ii) outflow in the form of an equatorial stellar wind, driven by the star's own radiation; (iii) non-radial pulsations in the star, perhaps several different modes which, when they reinforce each other, can drive gas outward; (iv) magnetic field loops in the photosphere; there is some evidence for such loops in these stars; and (v) the effect of a binary companion. [Possibility (vi) would be ``some or all of the above".] Harmanec's paper deals specifically with the role of binarity, which is undoubtedly a factor in some of these stars.

John P. Manker, Chair

During the fiscal year 2001-2002, seventeen observers contributed heavily to the AAVSO PEP database i.e. 3451 observations. The grand total is now 33935.

Over the fiscal year, our observers provided data in support of the Gravity Probe B satellite. The designated star to be observed was IM Peg. This star is a guide star for GP-B. To this end, the guide star was observed 433 times. Lou Cox (Canada) contributed 64 per cent of the data -- 277 observations.

Also during the fiscal year, we have six new observers. Unfortunately only one is very active -- Patrick Wiggins (Utah).

The following are the observations submitted between 1 October 2001 and 30 September 2002: name of observer, location, number of observations in 2001-2002, and observer's grand total: T. Beresky (MO), 9, 779; W. Clark (MO), 52, 449; L. Cox (Canada), 620, 1183; F. Dempsey (Canada), 54, 602; S. Dallaporta (Italy), 58, 1792; F. de Villiers (South Africa), 20, 418; J. Fox (MN), 97, 223; B. Grim (UT), 35, 82; W. Jones (South Africa), 633, 2450; P. Kneipp (LA), 17, 186; G. Lopata (CA), 9, 36; K. Luedeke (NM), 640, 4069; N. Stoikidis (Greece), 337, 1615; R. Thompson (Canada), 778, 7238; J. Wood (CA), 12, 2833; P. Wiggins (UT), 70, 70; H. van Bemmel (Canada), 10, 80. Total 2001-2002: 3451. Grand total: 33935.