Doppler imagery of the spotted RS canum venaticorum star HR 1099 (V711 Tauri) from 1981 to 1992

We present a set of 23 Doppler images of the spotted RS CVn star HR 1099 (V 711 Tauri;HD 22468) obtained from 1981 to 1992. HR 1099 shows a large, cool polar spot that has persisted for the 11 yr of this study and other low-la titude spots that come and So on relatively short (less than 1 yr) timescal es and can emerge anywhere on the star. The polar spot has variable protube rances that look very similar to the time-variable vertical extensions of t he Sun's polar coronal hole. The area of the polar spot and its extensions shows marginal evidence of being variable, with a period of about 3 yr and an amplitude of about 1% that is perhaps indicative of a weak cycle, but th is is not yet conclusive. Comparison of our Doppler images with previously published "few spot" model fits to the light curves shows that such simple spot-model solutions, whil e sometimes in agreement, are often misleading and nonunique, particularly when the light-curve amplitude is small. Moreover, these spot-model fits do not recover the existence of the polar spot. The Doppler images show quite good agreement among multiple images at a given epoch and between differen t Doppler imaging research groups using completely independent data sets an d imaging software. Our (cool spots only) Doppler imaging solutions, when p roperly thresholded, generally well reproduce the published light curves. H owever, in one instance the difficulty of fitting light curves suggests tha t at least one hot spot was present on HR 1099 during one observing season. Variations in the mean brightness of the system at the observed 0.05 mag l evel seem to correlate with spot area, particularly the polar spot, indicat ing that the mean light level is a pretty good proxy of spot area on HR 109 9. While the polar spot with variable extensions was always present, isolat ed spots also frequently appeared at both mid- and low latitudes. On severa l occasions isolated prominent spots emerged and then disappeared on or nea r the equator. The "migrating photometric wave" ion HR 1099 is due not to a simple longitu dinal migration of spots on a differentially rotating star but rather to ch anges in the spatial distribution of a few spots (some of which move but mo st of which are fixed in longitude) that emerge and then disappear. So, at least for HR 1099, the phase drift of this migrating photometric wave minim um contains very little unique information about differential rotation or s pot migration. While the tracking of individual features involves some unce rtainty and speculation because of our limited time sampling, the tracks of two long-lived spots suggest that some spots that emerge at low or interme diate latitudes may migrate up to the pole in a clockwise spiral (slower th an the orbit), then apparently merge with the polar spot. If these dark spo ts trace magnetic flux, we speculate that some of the magnetic flux emergin g at lower latitudes migrates poleward and merges with the polar spot flux. It is not yet clear whether this flux is of the same or opposite polarity to the polar spot and thus whether these poleward-migrating, low-latitude s pots reinforce or cancel the polar spot field. One of the high-latitude spo ts also appeared to get stretched in longitude as it approached the polar s pot, and its overall track is quite reminiscent of the annulus of toroidal field found by Donati et al. encircling the polar spot of HR 1099 in 1990.9 . In general, the spots appear to be very tightly locked to the orbital frame of the system, and most disappear before they have had a chance to migrate significantly; Like solar coronal holes, they show very little evidence fo r shear due to differential rotation. A few selected, long-lived features g ave longitudinal migration rates of 1 part in 300 to 1 part in 3600 of the rotation period, in the sense that intermediate and low latitudes rotate sl ightly slower than the orbital angular velocity, while the pole and highest latitudes appear to be synchronized to the orbit. The implied differential rotation is thus of opposite sign and about a factor of 56 less than for t he Sun. The rotation rate versus latitude behavior can be well fitted with a variety of formulae, including the PI launder formula. One of the best fi ts is provided by a rotation period versus latitude that is proportional to the surface strength of a centered axisymmetric magnetic dipole field, wit h the pole synchronized to the orbit and lower latitudes rotating more slow ly. We believe that these starspots are not measuring photospheric differen tial rotation. Instead, like solar coronal holes, their relatively low degr ee of shearing and nearly solid body rotation may be enforced by a multikil ogauss, axisymmetric, nearly current-free quasi-potential global magnetic f ield. Our Doppler images also agree very closely with the Zeeman-Doppler im agery of Donati et al. and support their finding that regions around the ed ge of the polar spat and within bright spats show largely monopolar fields of at least 300-700 G strength, The large, permanent cool polar spots, the very low observable differential relation and shearing of starspots, and th e evidence of strong, essentially unipolar magnetic fields associated with them leads us to believe that NR 1099 and other rapidly rotating RS CVn sta rs harbor quite strong (multikilogauss) axisymmetric global magnetic dipole fields. These fields have historically been largely hidden from view by th eir high degree of rotational symmetry, by being concentrated in the low su rface brightness dark spots, and by these stars' high degree of rotational line broadening. We propose that the starspots on HR 1099 and other rapidly rotating RS CVn stars are, by analogy with solar coronal holes, large unip olar, magnetic regions that are tightly frozen into multikilogauss, axisymm etric global dipole fields in these stars. Since the large cool polar spots , the signature of these dipoles, are not present on more slowly rotating R S CVn stars,we believe that they must be dynamo-induced fields rather than remnant fossil fields.