COMMISSION 27 OF THE I. A. U.
          INFORMATION BULLETIN ON VARIABLE STARS
                       Number 3304

                                         Konkoly Observatory
                                         Budapest
                                         15 March 1989

                                         HU ISSN 0374 - 0676


                    LIGHT CURVES FOR XY UMa


The variable nature of XY UMa was first noted by Geyer et al. (1955), after
which Geyer began a prolonged series of photometric observations, the results
of which he reported in 1976. He gave a general explanation of the
peculiarities of the data by employing the concept of evolving starspots. This
idea has been reinforced by a wealth of corroborating evidence in this and
similar cool "short period RS CVn" stars. (See e.g. Baliunas and Vaughan,
1985). Indeed, XY UMa, in terms of its chromospheric surface flux, may well be
the most active of such systems (Gurzadyan, 1987).

Recently Heckert and Zeilik (1988) have reported new observations of the star,
which they analysed in terms of a single spot group centred at a longitude not
far from 270deg, a minimum radius of about 10deg and appearing at some
intermediate latitude (approx. 40deg in an unspecified hemisphere).

The purpose of the present article is to report some similar studies on the BVR
light curves of Jassur (1986), which he observed in Egypt during March 1979.


                               [FIGURE 1]

 Figure 1: Model Maculation Wave Fit to the Initial Fit residuals

                               [FIGURE 2]

 Figure 2: Final Fit to the corrected data. The "raw" light curve can be
           seen in Jassur (1986).


The starting parameters were taken from Budding and Zeilik (1987). Initial
fits to the distorted light curves produced Geometric parameters r_1=0.346,
r_2=0.175 and i=83.2deg, not too far, but significantly different from the
values of Budding and Zeilik. The fractional sizes of the two stars are larger
after the re-run, and the inclination lower.

We next looked at the difference curve after subtracting this first order model
from the light curve. This is analysed for maculation ("starspot") effects.
Two spot groups were used. They appear to be centred at around longitudes
132deg and 220deg. They are of comparable size (approx. 15deg in radius) and
have been fixed at intermediate latitudes. These spots were given small, but
nonzero surface fluxes in the three colours, unlike the minimum area black
spots of Heckert and Zeilik.

Finally the "cleaned" light curves were formed by taking out the calculated
maculation effects from the original data. The basic geometric parameters
specifying the fit to this new light curve do not change by very much from the
initial fits (r_1=0.345, r_2=0.190, i=84.0deg). Correlated errors calculated
for these quantities are of the order of 1%. These geometric values only
varied within the expected error estimates in the three wavelength ranges.
Although these are closer to the "adopted" final values of Budding and Zeilik,
there is still some appreciable difference (r_1=0.327, r_2= 0.168, i=88.2deg).

The curve-fits were also checked by the codes of Wilson and Devinney (1971).
The geometric parameters derived for the fit to the "clean" curve are
essentially similar to the ones we gave before (i.e. rather different from
those derived by Budding and Zeilik), apart from the fractional luminosity
values. Banks (1989) has noted some slight systematic difference between the
fractional luminosities obtained in the Wilson-Devinney code from those of the
Budding program in various curve fitting experiments for eclipsing binaries,
though geometric parameters tend to be effectively similar. The difference in
luminosity values may be related to differences in the roles of assigned
temperatures in the two procedures.

The conclusion arising from these present efforts therefore seems to present
some challenge to the expectation given by Budding and Zeilik (1987) that the
derived geometric parameters for "clean" light curves of the same system should
always be essentially the same. They expected this to provide a general
confirmation of methodological adequacy. On the basis of these present results
this cannot be confirmed. Perhaps some systematic effects are at play in the
case of XY UMa other than those which the maculation wave and eclipsing binary
variation separating procedure of Budding and Zeilik takes into account, at
least with a single iteration of the procedure.


T. BANKS                       E. BUDDING
Physics Dept.                  Carter Observatory,
Victoria University of Wgtn.   Wellington,
New Zealand                    New Zealand

References:

Baliunas, S.L., and Vaughan, A.H., Ann. Rev. Astron. Astrophys., 23,
   pp. 379-412, 1985. [BIBCODE 1985ARA&A..23..379B ]
Banks, T.S., Unpublished M.Sc. Thesis, Victoria University of Wellington,
   New Zealand. (1989).
Budding, E., and Zeilik, M., Astrophys. J., 72, 369, 1987. [BIBCODE 1987ApJ...319..827B ]
Geyer, E.H., in "Structure and Evolution of Close Binary Systems", edited by
   P. Eggleton et al., p 313, 1976. [BIBCODE 1976IAUS...73..313G ]
Geyer, E.H., Kippenhahn, R., and Strohmeier, W., Kleine Veroeffentl, Remeis
   Remeis Sternwarte Bamberg, #9, 1955.
Gurzadyan, G.A., Astrophys. Space Sci. 123, 67, 1987. [BIBCODE 1986Ap&SS.123...67G ]
Heckert, P., and Zeilik, M., IBVS #3253, 1988.
Jassur, D.M.Z., Astrophys. Space Sci., 128, 369, 1986. [BIBCODE 1986Ap&SS.128..369J ]
Wilson, R.E., and Devinney, E.J., Astrophys. J., 166, 605, 1971. [BIBCODE 1971ApJ...166..605W ]