Eclipsing Binary Update Number 7

Number 7 November 1998

Notes from the Chairman

The scenery and the mountain streams in the vicinity of Boulder, Colorado, were a bit distracting. Nevertheless, those of us in attendance at the Spring Meeting of the AAVSO managed to renew our contacts and tend to the business at hand. Since the Fall Meeting, 15 AAVSO observers have reported more than 5,500 observations of eclipsing binary stars.

In the previous issue of this newsletter, we listed a number of eclipsing binaries on the AAVSO program that have been neglected by our observers. The response has been encouraging. We have even heard from astronomers not directly involved in our program, who provided their recently observed times of minima on some of these stars. I would urge observers to dig out that list again and continue the pursuit of those stars that have now rotated into position for viewing. We will plan to review the status of this list again at the Fall Meeting.

The importance of maintaining continuous observation of eclipsing binaries is illustrated by Chris Stephan's recent discovery that YY Del is currently running very late compared to our published ephemeris. Thirty years of published AAVSO data on YY Del would indicate that this star has a stable period. Yet, three recently observed minima, two by Stephan, would be completely off the scale of the O-C diagram published in "Observed Minima of Eclipsing Binaries No. 4." The fact that many eclipsing binaries change period, seemingly without warning, increases the significance of any minimum timing that fills what might otherwise be a gap in the data. --MEB

Now you see it, now you don't - let's play eclipsing binary peek-a-boo

What's a variable star fanatic supposed to do when he is handed a key to Indiana University's old Link Observatory facilities, which include a 25-cm f/6.3 astrograph and a decrepit but still workable blink comparator? I think you'll agree, Your Editor had no choice but to begin taking photos and searching them for new variable stars (Williams, JAAVSO 25, 30). The first success was just too easy - and it was an eclipsing binary! I had accumulated several exposures of the field around M46 and M47 in Puppis. So one day I put the two best-matched photos into the comparator to test its performance. I was adjusting the focus and fiddling with the knobs to bring the two photos into alignment when I noticed that one of the stars near M47 was blinking on and off.

The blinking star was about magnitude 12.8 ptg on one photo and completely invisible, fainter than 15.0, on the other. This variable object was so obvious that I assumed it would prove to be a known variable. I found the star on the chart of this region that I had printed from the Guide CD-ROM program, so it wasn't a stray asteroid near its stationary point. It didn't have a variable star label, and no new variable was announced near its position in the newer Name Lists. So the star must be a previously undetected variable, unless some odd emulsion defect prevented the image from registering on the one film where it was invisible. New photos in the following weeks didn't show any variation. I finally confirmed variability when I was able to consult the vast Harvard plate collection. After examining the star on more than 30 plates, I found one where it had disappeared again. After looking at another 500 plates, I had found almost a dozen minima. Now only one question remained: How could an eclipsing binary with such a deep minimum have been missed by all the earlier searchers? Discovery dates of other variables in the same field showed that Hoffmeister at Sonneberg had surveyed this region at least twice.

When I calculated the period from the minima I had found, I had the answer. The period was 10.98 days. This longish period makes observable eclipses relatively rare. But worse, 10.98 days is very nearly equal to 11 sidereal days. From any one longitude, minima are observable near the meridian in a dark sky only one season out of three. Throw in cloudy nights and the occasional full moon, and many years could pass before a minimum is caught on a photographic plate at a particular observatory. This is the game that eclipsing binary observers have to play not just the hide and seek with clouds and moonlight, but peek-a-boo with eclipsing binaries whose periods join in odd syncopation and complex synchronicities with the Earth's rotational period.

One of the most remarkable recent examples of this pesky phenomenon is MP Pup, whose period of 0.9989 day was first revealed by Our Chairman (see Baldwin et al. 1994, JAAVSO 23, 14). A synoptic light curve (right) compiled from observations on Harvard plates shows clumps of minima every 600 days. This light curve can be decrypted when you realize that the period of MP Pup is 1+1/600 sidereal day.

Gerry Samolyk, the Eclipsing Binary Committee's long-time chart distributor and ephemerides compiler, went to extraordinary lengths to capture a minimum timing of Z Her one season. This star's period is 3.99 days - say no more! That year, the minima of Z Her were ideally scheduled for observation from the Mid-Atlantic Ridge. Gerry was up to the challenge, however. At that time he was making frequent trips from Milwaukee to Europe on business, so he took along his binoculars and observed one branch of the eclipse light curve from one side of the Atlantic and the other branch a few days later from the other side of the ocean.

As a final case in point, consider V1061 Tau, an early-type, overcontact binary system discovered photographically in 1990 by AAVSO Vice President Dan Kaiser. It's period is 1.3852 days. One of Your Editor's goals is to obtain a high-precision photoelectric light curve in order to improve upon the published light curve solution (Terrell et al. 1995, PASP 107, 653). I hate to observe in cold weather, so I wanted to get as many observations as I could before winter.

In early October, Taurus doesn't rise high enough for high-quality photometry until about 2 am. Alas, I have to work for a living, so early morning observations can only be made on weekends. On Saturday, October 10, at 2 am the orbital phase of V1061 Tau was 0.73. The next weekend, on Saturday, October 17, at the same hour the phase was 0.79. Note that after a full week, we have progressed only 1/20th of the way along the light curve. Taurus rises earlier each night, of course, and by the third weekend we could start observing at 1 am. On Saturday, October 24, at that hour, the phase was 0.80 - ouch! No progress at all.

A quick calculation reveals that seven sidereal days are equal to 5.04 orbital cycles of V1061 Tau. So if observations are started on the same night each weekend when Taurus reaches the required altitude, it would take almost 25 weeks to record the complete light curve. Of course, by January Taurus will be rising before sunset and observations can be made during evening hours on every night of the week. Nonetheless, it looks like Your Editor will be observing for a long time before obtaining data at all phases of the orbital cycle. But that's the nature of the game when you play peek-a-boo with an eclipsing binary through the picket fence we call night and day. --DBW

Eclipsing Binary News and Notes

In 1955, astronomers at the Remeis Observatory in Bamberg, Germany, began publishing a list of new variable stars that eventually exceeded 1,700. Recently, W. Braune, M. Dahm, and H. Meyer published a list of the Bamberg Variables with their GCVS or NSV designations and their identification with stars in the Guide Star Catalogue (BAV Mitteilungen Nos. 108, 109). A quick review of this list shows how valuable the Bamberg discovery program was for eclipsing binary astronomy. Many bright eclipsing variables were found. BV stars that will be familiar to AAVSO eclipsing binary observers include PV Cas, RV Tri, OX Cas, XZ UMa, AL Cam, AI Dra, FZ Del, EG Cep, HU Tau, RU UMi, V1010 Oph.

The most recent program to produce a rich harvest of eclipsing binaries was a byproduct of the Hipparcos satellite, which in addition to measuring high-precision parallaxes also detected variability in many stars brighter than 10th magnitude. A number of these new variables are eclipsing binaries with amplitudes of 0.4-0.6 magnitude:

                  V372 And	(EA)	9.1-9.5		2.94 d
                  DN Cam	(EB) 	8.2-8.7, 	0.50 d
                  DP Cam 	(EA:)	9.9-10.4,	--
                  CD Lyn	(EA) 	9.8-10.4, 	1.53 d
                  CL Lyn	(EA) 	9.8-10.3, 	1.59 d
                  CN Lyn	(EA) 	9.1-(9.6, 	1.96 d
                  FN Cam	(EW)	8.6-9.1, 	0.68 d
                  FK Dra	(EA) 	9.3-9.8, 	2.00 d
                  CI CVn	(EB) 	9.4-9.9, 	0.82 d
                  IO UMa 	(EA) 	8.2-8.8, 	5.52 d
                  EF Boo	(EB) 	9.4-10.0, 	0.42 d
                  YY CrB 	(EW) 	8.6-9.1, 	0.38 d
                  V792 Her 	(EA) 	8.1-8.7, 	27.5 d
                  V441 Cep	(EA)	8.7-9.2,	3.30 d

Many more eclipsing binaries with amplitudes less than 0.5 magnitude were discovered, which will be good candidates for photoelectric investigation. According to VSNET 1643 from the Hipparcos News & Feedback Service, Hipparcos data and light curves can be searched online at http://astro.estec.esa.nl/Hipparcos/research.html.

The 7th magnitude star HD 143213 (=SAO 121294 = GSC 353:301) is a bright eclipsing binary detected by the Tycho instrument aboard the Hipparcos satellite. Nine of the 103 measurements were 0.3-0.5 magnitude fainter than the rest, suggesting that this was an eclipsing binary. But the period could not be determined due to the small number of fainter observations and their poor time distribution. In a masterwork of visual observation (right), E. Born in Erlangen, Germany, made 310 estimates of the new variable with 10x70 binoculars between June 1996 and August 1997. Analysis of these data revealed that HD 143213 is a detached binary system with a period of 3.45 days and a displaced secondary minimum (Bastian and Born, IBVS 4536).

With the period found, the Tycho data could be plotted to phase, confirming the period and the displaced secondary. The eclipses of HD 143213 are about 5.5 hours in duration, and the secondary minimum occurs at phase 0.54. But the two light curves now presented a mystery - the primary minimum at phase 0.00 of Born's light curve appeared at phase 0.59 of the Tycho light curve. Both data sets agreed on the period, and no period change had occurred during the six years between the two sets of observations. Additional observations by Born showed that the period was correct and constant. The answer to the mystery was revealed in IBVS 4590 by the same authors. The epochs of the Tycho data were in error. When the data were extracted from the Tycho internal database, an obsolete routine for calculating the Julian dates had been used inadvertently. When the JD times of the Tycho observations were decreased by 1.5 days, the two data sets agreed perfectly.

In 1997, Klaus Bernhard in Linz, Austria, put his CCD camera and 20-cm telescope to work, searching for new variables in close proximity to bright stars. This plan calls to mind the dwarf nova HL CMa, which is so close to Sirius that it wasn't discovered until its X-ray emissions were detected by satellite instruments, even though its maxima are 11th magnitude and frequent.

After six months of searching, Bernhard discovered the variability of GSC 1062-0033 about 0.5 degree NE of Altair. Observations by Bernhard, Wolfgang Quester in Germany, and Seiichiro Kiyota in Japan revealed that the new variable is an eclipsing binary with an Algol-type light curve (left) and period of 1.616 days. The visual magnitude range is about 10.5-11.0. The preliminary report appeared in IBVS 4540.

According to VSNET 1691, the southern star HDE 331015 (= CoD -48 10986 = GSC 8329:3364) is a large-amplitude Algol variable. The discovery was made by Australian nova hunter Paul Camilleri on his search photos. Follow-up observations during the past two years by several amateurs in Australia and New Zealand revealed that this eclipsing binary varies from visual magnitudes 11.0-13.4 with a period of 9.417 days. Eclipses last 7.2 hours with a total phase at minimum of 1.2 hours. Details will appear in Publications of the Variable Star Section, Royal Astronomical Society of New Zealand.

Featured star: RZ Cassiopeiae

RZ Cas is probably the most frequently observed eclipsing binary in the sky. The AAVSO Eclipsing Binary Committee receives an average of about 15 minima each season, and other observing groups receive similar numbers. RZ Cas is popular for many reasons. With a range of variation from 6.2 - 7.7 V, it is bright enough for observation with binoculars. At a declination of +69 degrees, it can be observed by Northern Hemisphere observers for about 10 months each year. And the eclipses are deep and rapid. In the middle phases of eclipse, RZ Cas dims and brightens by 1 magnitude per hour. This is illustrated by the excellent photoelectric light curve (right) recorded in October 1995 by Kikuichi Arai at Hanyu-Jitsugyo Senior High School in Japan (from IBVS 4502). The period of 1.195 days means that there are several observable minima each month.

RZ Cas is a classic Algol-type eclipsing binary with a hot, A2.8V primary star and a cooler, evolved secondary component. This binary system is highly active. The O-C diagram (below) covers the interval 1954-1997 and is compiled from AAVSO data after 1963. The O-C residuals show an unusual "saw-tooth" pattern of 6-7 years duration before the period increased substantially around 1993. It will be interesting to see whether this short-cycle phenomenon continues to manifest itself around the new mean period. Even though RZ Cas is well-observed, we still need more data to define and confirm this kind of detail in the O-C diagram.

Most light curves of RZ Cas at minimum have the rounded form of a partial eclipse, as shown in the Arai light curve. But occasionally, the light curve seems to show a brief interval of constant light at minimum, leading some photometrists to believe that the eclipses could be total. However, these light curves must be interpreted with the greatest caution. The moment of minimum is the most sensitive portion of an eclipsing binary's light curve. All minima are "flat" at the moment when the direction of the light change is reversed. The displacement of just one or two data points by observational scatter can strengthen the impression of a brief interval of constant light. In addition, the shape of the minima of highly interactive binaries like RZ Cas could be affected by circumstellar gas streams, accretion disks, and starspots.