CAN THE STARPATCH ON ZETA-BOOTIS A BE EXPLAINED BY USING TANGENTIAL FLOWS

We discuss a modification of the starpatch model originally proposed in 1988 by Toner and Gray to explain their observation of periodic line asymmetry and line strength variations for the G8 dwarf zeta-Boo A. The modified model incorporates known properties of large spot groups observed on the Sun. In particular, it parameterizes conditions within the patch in terms of a mass flow which is tangential to the stellar surface (analogous to the Evershed and moat flows seen around sunspots), rather than in terms of an enhanced granulation velocity dispersion (as was assumed in the original model). We performed an extensive search of parameter space, fitting not only the line bisector variations, but also a newly reported approximately 5% variation in the mean line broadening. From this fitting process we infer the following characteristics for the patch: areal coverage 10% +/- 3% of the visible disk, latitude 30-degrees +/- 4-degrees, mean brightness 0.85 +/- 0.05 relative to the "quiet" photosphere, mean tangential flow velocities of 8.0 +/- 1.5 km s-1, and dispersions about the mean of 8.0 +/- 2.0 km s-1. For the line strength variations to be consistent with the observed photometric variations, we find that at least approximately 35% of the patch area must be composed of penumbral-type features. If we assume the temperature difference of approximately 370 K for penumbra on the Sun, this fraction increases to approximately 75%. This suggests, that while the patch on zeta-Boo A may share similar dynamics with sunspots, it has a very different appearance. Our best solution using the tangential flows model gives a reduced X2 measure of the goodness of fit of 0.85, indicating very good agreement. We also computed a grid of solutions using the original, enhanced granulation velocity dispersion model. When these were compared to the observations (including both the line bisector and the line broadening variations) the agreement was found to be somewhat less satisfactory than with the tangential flows model. The best solution using the original hypothesis yields a reduced X2 of 1.29. This model requires a patch at a latitude of approximately 45-degrees, covering approximately 14% of the visible disk. The patch is approximately 10% fainter than the surrounding photosphere and has a granulation velocity dispersion which is enhanced by a factor of approximately 1.7 relative to the rest of the star. Since both models reproduce the observations with adequate precision, we cannot conclusively choose one interpretation over the other. However, the two models predict very different geometry dependences for the shape of the phase curves. This should allow future studies to decide if one interpretation is markedly better than the other.