Differential Interferometry (DI) combines high spectral and high spatial resolution. On non resolved objects, it yields the angular variation of the source photocenter as a function of wavelength which has been shown theoretically and experimentally to complement very usefully both interferometric and spectroscopic data in a large number of astrophysical problems. This paper presents the general characteristics of DI which are likely to allow improvements of the Doppler images of stellar surface structures as soon as interferometers with large apertures and baselines approaching 100 m are available.

The capability of optical long baseline interferometry for measuring the angular diameter of stars or binary separation is now well established. However, for the imaging of photospheric disk structures one needs very long baseline interferometers operated in the multi-telescope phase-closure technique. In this paper we will stress the capability of spectro-interferometric measurements to constrain the physics of hot stars. We will report our study of the interacting binary, β Lyrae, and the mass-losing Be star γ Cassiopeiae. We will look at the interpretation of both the modulus and phase data recorded by the long baseline interferometer GI2T in the southern France. The performances and limitations of spectro-interferometric techniques will also be discussed through some of the most exciting prospects within the reach of current interferometers.

The magnetic field of the Sun is mainly concentrated into intense magnetic flux tubes having field strengths of the order of 1 kG. In this paper an overview is given of the thermal and magnetic properties of these flux tubes, which are known to exhibit a large range in size, from the smallest magnetic elements to sunspots. Differences and similarities between the largest and smallest features are stressed. Some thoughts are also presented on how the properties of magnetic flux tubes are expected to scale from the solar case to that of solar-like stars. For example, it is pointed out that on giants and supergiants turbulent pressure may dominate over gas pressure as the main confining agent of the magnetic field. Arguments are also presented in favour of a highly complex magnetic geometry on very active stars. Thus the very large starspots seen in Doppler images probably are conglomerates of smaller (but possibly still sizable) spots.

The scope of this paper will be starspots, and by that I mean regions cooler than the surrounding photosphere (like sunspots) but I also mean the whole range from very small (like sunspots) to very large (such as seen on very active stars like RS CVn).

Some stellar spectral lines are quite sensitive to temperature. With current techniques, we can, for example, measure temperature modulation of a few degrees. But high temperature sensitivity can also be a detriment when it masks variations in other physical parameters of interest. In such cases, we must have a way of determining the temperature precisely so that the residual differences or variations can be seen and interpreted. The situation with stellar surface features is particularly challenging because several physical variables interact simultaneously, each impressing their signature on the spectral lines. It is frequently a good plan to compare a highly temperature sensitive line to one that is insensitive or to one that has an inverse sensitivity. A good way to find suitable lines is to compare exposures of hotter and cooler stars, as illustrated in Fig. 1. Here we see lines that hardly vary (Si I), some that show modest changes (Fe I), some that show large changes (V I), and one that shows a reverse change (Fe II). Some caution is needed because differences in surface gravity and metallicity can also produce changes in line strengths. So the general idea is to play off the differences of one line compared to another and avoid any direct dependence on absolute line strengths. Yes, this is indeed nothing more than a careful refinement of determining a spectral type.

High resolution, high S/N spectra have been successfully modeled to yield information on the magnetic properties of cool stars for some 15 years now. These analyses yield estimates of the intensity weighted surface filling factor of active regions, f, and the mean unsigned field strength in these regions), B. The measurements are difficult, though, hampered by the small magnitude of the Zeeman effect, small f values for most stars, and the unknown spatial and thermodynamic properties of the active regions. Recent data, mostly in the infrared (IR) where the Zeeman effect (∞ λ2) is larger, are yielding better measurements than ever before. I summarize the new observations here, and show how they are expanding and modifying our understanding of magnetic fields on cool stars.

We present 23 Doppler images of the spotted RS CVn star HR 1099 (V711 Tau) from 1981 to 1993. All of the images show a large variable-shape cool spot straddling the pole. Many of the images also show a number of isolated low-latitude spots. This image set has been used to track the emergence and evolution of spots on this star over the 12-year interval. The Doppler images are compared to previous ‘few-circular-spot’ photometric model solutions when available. These photometric solutions do not well-represent the spot distribution. We discuss, among other things: the location on the stellar surface where spots first appear, their subsequent movement with time, differential rotation from spot motions, and spot cycles and dynamos within HR 1099.

Knowledge of the solar sunspot cycle extends back to the mid-19th century with the work of Schwabe (1843) and Wolf (1856). The mean cycle period of the Sun is 11 years, however, individual cycle lengths range from 7 to 13 years (Eddy 1977). In this century, however, the length of the solar cycle has been closer to 10 years (Donahue and Baliunas 1992a). A complete explanation of the solar magnetic activity and its variations has not yet been produced, although a hydromagnetic dynamo is frequently posited as the source of solar (and therefore stellar) magnetic activity. Empirical measurements of those stars in the H-R Diagram which have convective zones and surface magnetic activity provide the boundary conditions and the range of behavior which must be explained by any all-encompassing theory explaining stellar magnetic activity, and activity cycles.

A theoretical study of storage, instability and rise of magnetic flux tubes in the outer convection zone of a cool star is presented. Special emphasis is laid on their emergence latitudes at the surface of magnetically active stars. We apply the ‘solar paradigm’ and assume toroidal magnetic flux tubes to be stored in force equilibrium within the overshoot layer underneath the convection zone. A non-axisymmetric (undulatory) instability leads to the formation of flux loops, which rise through the convection zone and emerge at the surface to form bipolar magnetic regions and starspots. Our approach combines the analytical determination of the linear stability properties of flux tubes with numerical simulations of the nonlinear evolution of the instability and the rise of magnetic flux tubes through the convection zone. It is found that for sufficiently rapidly rotating stars the magnetic flux emerges at high latitudes since the Coriolis force leads to a poleward deflection of rising flux loops. The latitude distribution of the emerging flux is determined for a number of stellar models along the evolutionary sequence of a star with one solar mass, from the pre-main sequence evolution up to the giant phase. Rapid rotation and deep convection zones favour flux emergence in high latitudes. Starspots right at the stellar (rotational) poles form either directly by flux eruption from small stellar cores (as for T Tauri stars or giants) or by a poleward slip of the sub-surface part of the flux tube after flux emergence in mid latitudes. The latter process explains the simultaneous existence of polar spots and spots at intermediate latitudes as observed on some stars.

“Are results from ill-posed problems, like Doppler-imaging, conclusive at all?” and “Could polar spots be simple image-reconstruction artifacts?” are often asked questions. Although I can not definitively answer them, I will present observational evidence for the existence of cool starspots at or very near a stellar rotation pole.

Polar spots in Doppler Imaging maps of rapidly rotating late-type stars, especially the active components of RS CVn binaries, represent a major break with the solar paradigm, within which spots are only found at low latitudes. I examine critically the evidence for polar spots and point out that they are particularly sensitive to a number of systematic effects relating to determining reference “zero-points”.

Doppler imaging results for the surface abundance distribution on Ap stars are reviewed. The silicon distribution on these stars is often characterized by depleted spots near the magnetic poles and enhancements in ring-like structures at intermediate magnetic latitudes. These results are consistent with the predictions of diffusion theory. On two stars, γ Ari and CU Vir, the silicon enhancements appear very close to a magnetic pole. This can be explained if silicon is enhanced where the magnetic field has its maximum horizontal field strength and these stars possess a decentered dipole field. Doppler images of abundance distributions on Ap stars may be used to accurately measure the obliquity angle and to estimate the decentering parameter for a dipole field. The chromium distribution on at least four Ap stars show a depleted band coincident with the magnetic equator and depleted spots at the magnetic poles. Such a distribution can only arise if the star has a predominantly quadrupole field, or if horizontal diffusion is playing a role in the distribution of elements.

The various mechanisms leading to the appearance of horizontal inhomogeneities at the surface of ApBp stars are critically reviewed. The effect of magnetic fields is essential but is it more to control locally, on the surface, the appearance of anomalies that are caused by the vertical transport of particles, or is it to create directly the horizontal inhomogeneities by horizontal transport? The time scales for horizontal and vertical transport will be discussed in this context.

The processes discussed include: a) ambipolar diffusion of protons and hydrogen in the presence of magnetic fields; b) the guiding of diffusion by magnetic fields; c) the horizontal component of radiative accelerations; d) mass loss; e) light induced drift.

We focus here on those T Tauri star properties that make it worthwhile to Doppler-image these young stellar objects, namely magnetic structures and accretion features. We then comment on recent results obtained for two bright, fast rotating weak-line T Tauri stars.

Of the steadily increasing number of Doppler Images appearing in the literature, the majority are of the RS CVn subclass of binary stars. The large rotational velocities and dynamo action of these stars is probably attributable to tidal spin-up of the active subgiant star by its binary companion. Their dynamos and resultant surface features (starspots) are thus almost certainly influenced by the companion. As such, these objects are not the ideal testing ground for comparison with single-star solar-like activity. Recently, a handful of Doppler images of single late-type main sequence and pre-main sequence stars have appeared in the literature. Pre-main sequence stars include V410 Tau (Joncour, Bertout, & Menard 1994), (Strassmeier, Welty, & Rice 1994), (Hatzes 1995) and HD 283572 (Joncour, Bertout, & Bouvier 1994). The main sequence stars include AB Dor (Collier Cameron & Unruh 1994), (Kürster, Schmitt, & Cutispoto 1994) and LQ Hya (Strassmeier, et al. 1993), (Saar & Piskunov 1994). These studies suggest that high-latitude dark spots are common to all rapidly-rotating stars, although, the spots found in the above work do not symmetrically straddle the pole as do polar spots of rapidly-rotating RS CVn binary stars.

Observations of lithium in sunspots resulted in the prediction of significant lithium equivalent-width variability in chromospherically active stars. Lithium observations of several types of active stars, such as pre-main-sequence stars and late-type dwarfs in very young clusters, whose members are assumed to be nearly coeval, as well as post-main-sequence objects, show a significant range of abundances. Thus, a number of researchers over the past decade have investigated star spots as the possible cause of the various observed lithium-abundance spreads. Observationally, some evidence has been found for lithium equivalent-width variations. Computed synthetic spectra indicate that under certain conditions significant variations should be detected, but such conditions are not often met except perhaps in the most active stars. While variations have been claimed for some T Tauri stars, simultaneous photometry and spectroscopy will be needed to explore the complex relation of spots and lithium-abundance variations. The fact that some post-main-sequence stars have lithium abundances similar to those of pre-main-sequence stars complicates the identification of isolated pre-main-sequence systems.

The solar surface contains of bright regions (plages) and dark regions (sunspots) superposed on the photosphere. If the solar analogy is valid, then active late-type stars should also exhibit bright, spatially distinct plages. These plages can be detected by rotational modulation of chromospheric flux, or by Doppler imaging in chromospheric/transition region lines. I review the evidence for the existence of plages, with particular emphasis on two very active systems: AR Lacertae, for which we have sequences of Doppler imaging observations since 1984, and AB Doradus, which we observed extensively in 1994. I conclude that we need to be cautious when relying on the solar analogy to interpret observations of the most active late-type stars.

Cool gas in the heart of the hot solar chromosphere betrays itself through anomalously strong absorptions of 4.7 μm CO lines at the extreme limb, and off-limb emissions. The origin of the “cool clouds” is thought to be a thermal instability driven by the CO itself. Questions concerning the pervasiveness and detailed structure of the “thermally-bifurcated” zones are being addressed through stigmatic spectroscopy of the solar CO bands with the NSO McMath-Pierce telescope. Here, I report recent progress, and implications for surface thermal inhomogeneities on other stars.

This brief review advertises a breakthrough in solar dynamics: the identification of chromospheric Ca II K2V “grains” as manifestations of acoustic shocks. The successful numerical reproduction of the intricate spectral characteristics of K2V grain formation by Carlsson & Stein shows that the non-magnetic solar chromosphere is pervaded by outward traveling shocks to such extent that static modeling is no longer acceptable. Dynamical modeling should soon reconcile the long-standing Ca II/CO dichotomy. The chromospheric dynamics of the magnetized regions remains enigmatic.

The aim of this paper is to evaluate recent observational and theoretical results concerning the physics of chromospheric heating as inferred from IUE, HST-GHRS and ROSAT data. These results are discussed in conjunction with theoretical model calculations based on acoustic and magnetic heating to infer some conclusions about the magnetic and nonmagnetic surface structure of cool luminous stars. I find that most types of stars may exhibit both magnetic and nonmagnetic structures. Candidates for pure nonmagnetic surface structure include M-type giants and supergiants. M-type supergiants are also ideal candidates for identifying direct links between the appearance of hot spots on the stellar surface (perhaps caused by large convective bubbles) and temporarily increased chromospheric heating and emission.