ASTRONOMY & ASTROPHYSICS NOVEMBER II 1999,PAGE29 SUPPLEMENT SERIES Astron. Astrophys. Suppl. Ser. 140, 29–53 (1999) Starspot photometry with robotic telescopes. UBV (RI)C and by light curves of 47 active stars in 1996/97? K.G. Strassmeier, E. Serkowitsch, and Th. Granzer Institut f¨ur Astronomie, Universit¨at Wien, T¨urkenschanzstraße 17, A-1180 Wien, Austria e-mail: [email protected], [email protected], [email protected] Received March 15; accepted September 3, 1999 Abstract. We present continuous multicolor photometry their appearance and disappearance marking the begin- for 47 stars from October 1996 through June 1997. ning and ending of a spot cycle similar to the Sun’s 11-year Altogether, 7073 V (RI)c, UBV,andby data points, each cycle. the average of three individual readings, were acquired In the past decade several studies were presented that with three automatic photoelectric telescopes (APTs) at attempted to model long-term light-curve variations with Fairborn Observatory in southern Arizona. Most of our an evolving spot distribution on the basis of one stellar ro- targets are chromospherically active single and binary tation to the next. Much literature could be cited here but stars of spectral type G to K but there are also four instead let us selflessly draw your attention to our own se- pre-main-sequence objects and three pulsating stars ries of papers on “Time-series photometric spot modeling” in our sample. The light variability is generally due and the many references therein. Paper I on the RS CVn to rotational modulation of an asymmetrically spotted binary VY Ari (Strassmeier & Bopp 1992) included a de- stellar surface and therefore precise rotational periods scription of the modeling technique and were followed by and their seasonal variations are determined from Fourier Strassmeier et al. (1994a) on 15 years of photometry of analysis. We also report on photometric variations of γ HR 7275, Ol`ah et al. (1997) on 30 years of HK Lac and CrB (A0V) with a period of 0.44534 days. All data are Ol`ah et al. (1999) on 10 years of V833 Tau. A partic- available in numerical form. ularly important spot parameter extractable from these data is the spot’s lifetime and one result, still lively de- Key words: stars: activity — stars: late-type — stars: batted, is that there is no linear decay law for starspots as rotation — binaries: spectroscopic — techniques: there is for sunspots, i.e. the larger a sunspot the longer photometric it lives (dA/dt =const;A being the spot area). Various physically similar stars in terms of rotation period and spectral type can host spots or spot groups with a wide range of lifetimes. For a set of four spotted RS CVn stars, Henry et al. (1995a) observed individual spot lifetimes 1. Introduction between 0.5 years and over 6 years. The same range of life- times was found for the stars in our previous papers men- tioned above. Maybe not surprising, the spotted star with Photometric monitoring of late-type stars has proven to the longest photoelectric history (RS CVn with 45 years be a powerful tool to discover cool starspots and to mea- of broad-band data) also shows the longest time scale: a sure precise stellar rotation periods with, in some fa- cyclic change of the total spotted area with a period of vorable cases, even differential rotation on the stellar 20 years (Rodon´o et al. 1995). All these phenomena are surface (Hall 1972, 1994). The way the light and colors suggestive of a significantly non-solar magnetic field topol- of a spotted star vary with time traces the evolution of its ogy and a dynamo process that might operate on many starspots. Time-series light and color curves can then be different timescales. used to determine the approximate distribution of spots on Clearly, continuous photometric monitoring for a long the stellar surface, their temperature contrast, and even period in time is needed for conclusive analyses and au- ? All data are available from CDS via anonymous ftp to tomatic photoelectric telescopes (APTs) are ideally suited cdsarc.u-strasbg.fr (130.79.128.5) or via http://cdsweb.u- for such a task. In this paper, we present multicolor data strasbg.fr/Abstract.html from our two 0.75-m University of Vienna twin APTs 30 K.G. Strassmeier et al.: Starspot photometry with robotic telescopes (Strassmeier et al. 1997b) and the 0.25-m Phoenix-10 APT 2 millimag for Wolfgang and approximately 4 millimag (Boyd et al. 1984) for 1996/97. Altogether, 47 program for Amadeus. Between May and June 1997, the Amadeus stars were observed in either by, UBV,orV (RI)C.Some APT had gotten repeatedly out of focus and the internal special cases were targeted in all of these bandpasses but standard deviations raised to 7 − 8 millimag in V during with varying time and phase coverage on the individual this time. telescopes. Others had been observed only briefly and their aim was mostly to support spectroscopic observations for Doppler imaging with large telescopes (see, e.g., Rice & 2.2. The Phoenix-10 0.25 m APT Strassmeier 1998). Table 1 presents a summary of the pro- gram stars and their commonly known stellar parameters. The Phoenix-10 APT is now located at the same site in A period analysis for all stars from each individual tele- Washington Camp as Wolfgang-Amadeus and thus fol- scope gives a homogeneous set of stellar rotation periods lows the same weather and extinction pattern. It is al- and the associated rotational light variations. ready in routine operation since 1983 and is managed by Mike Seeds at Franklin & Marshall College as a multi- user telescope (see “Phoenix-10 Newsletter” and Seeds 2. Instrumentation and data quality 1995). Strassmeier & Hall (1988a) examined its data qual- ity from its first four years of operation and found ex- All three telescopes are located at Fairborn Observatory ternal uncertainties of 10, 20, and 28 millimag in V , B, near Washington Camp in southern Arizona and operate and U, respectively. For the stars in this paper, integra- fully automatically (see http:// 24.1.225.36/ fairborn.html tion times were set to 10 s for all targets. More recently, for a description of the observatory). Henry (1995) compared the long-term external precision of the Phoenix-10 APT with APTs of larger aperture (the Vanderbilt/Tennessee State 0.4 m and the Tennessee- 2.1. The two 0.75-m APTs “Wolfgang” and “Amadeus” State 0.8 m) and verified the telescope’s long-term stabil- ity. A further comparison was made with Wolfgang and The Wolfgang APT is optimized for blue wavelengths with Amadeus as well as with the University of Catania 0.8 m APT on Mt Etna (Strassmeier et al. 1997a). All relative an EMI-9124QB photomultiplier and Str¨omgren by filters zeropoints in the V bandpass agreed to within their formal while Amadeus is optimized for red wavelengths with an EMI-9828 tube and Johnson-Cousins V (RI ) filters (for errors. In order to eliminate datapoints grossly in error, we C applied a statistical procedure that excluded all data with details we refer to Strassmeier et al. 1997b). All measure- an internal standard deviation greater than 20 millimag ments were made differentially between the variable and a comparison star (the latter is sometimes referred to as the in V . “Comp” star). A check star is used to verify the stability of the comparison star (we sometimes refer to it as the 3. Results “Check” star). Stars brighter than ≈5m were measured with a neutral-density filter in front of the spectral filter. Table 1 lists the program stars and summarizes their Integration time was usually set to 10 s except for stars most relevant stellar properties taken either from the re- − m − fainter than 8 9 where 20 30 s for the broad-band cent literature or from the second edition of the cat- − filters and 30 60 s for the intermediate-band filters were alog of Chromospherically Active Binary Stars (CABS, used. The data reduction was based on nightly extinction Strassmeier et al. 1993) if appropriate. Table 2 identifies coefficients obtained from a set of standard stars or, for the comparison and the check stars, the observing inter- the nights when an insufficient number of standards was val, whether a neutral density filter was needed or not, observed (less than 20), an average of the previous three the useable number of differential data points – each the good nights was adopted. Otherwise, the procedures were mean of three readings on the variable and the compari- the same as described in Strassmeier et al. (1997a). son –, and which telescope was used. After the obviously deviant data points due to misiden- tifications and clouds were eliminated, we computed ex- ternal uncertainties, σext, for all check-minus-comparison 3.1. Seasonal light curves magnitudes. Such uncertainties allow an examination of the long-term data quality expected for the variable- The individual panels in Fig. 1 through Fig. 45 present minus-comparison data. The mean external standard de- the photometric data in one bandpass (usually V or y), viation of a “nightly mean” from a yearly mean was for the periodogram from these data, and the phased seasonal Wolfgang 4 millimag in b and y, and for Amadeus 6, 8, and light curve from the best-fit period. We emphasize that the 10 millimag in V , R,andI, respectively. The mean inter- sometimes abnormally large scatter in the phase plots is nal standard deviations from three readings of the variable almost exclusively due to intrinsic spot changes and not and four readings of the comparison star were less than due to instrumental scatter.