Looking at the Sun forges the framework within which we try to interpret stellar observations. The stellar counterparts of spots, plages, flux tubes, chromospheres, coronae, etc., are readily invoked when attempting to interpret stellar data. This review discusses a selection of solar phenomena that are crucial to understand stellar atmospheric activity. Topics include the interaction of magnetic fields and flows, the relationships between fluxes from different temperature regimes in stellar atmospheres, the photospheric flux budget and its impact on the measurement of the dynamo strength, and the measurement of stellar differential rotation.

The process of recovering the geometric information encoded in the time varying line profiles of rotating stars to obtain maps or images of the surface features on those stars has come to be known in stellar astronomy by the title “Doppler imaging”, although other titles have been used. The principles of Doppler imaging are fairly well known, but the extent of application has been less than might have been expected given the ready availability now of the electronic detectors, the large telescopes and the computing power needed to undertake this sort of imaging.

In recent years, Doppler imaging codes have been developed by a number of different groups and successfully applied to map the spot distribution on stellar surfaces. As more and more images are obtained we tend to see a pattern of high-latitude spots, particularly on RS CVn systems. These high-latitude spots were and still are greeted with scepticism as they are inferred from non-varying profile deformations. Here, I describe some of the problems of Doppler imaging and their likely impact on the resulting images.

We describe the extension of our Doppler imaging technique to the case of eclipsing binary systems. Using numerical experiments, we show that the simultaneous imaging of both components breaks the north-south ambiguity that occurs for single stars with high inclination. Characteristic map distortions can be used to correct errors in the orbital parameters of the binary.

We apply this technique to ER Vul – an active RS CVn type binary. Our preliminary results show the presence of the hot spots on the substellar sides of both components, perhaps due to the reflection effect.

We also discuss application of the same technique to contact binaries, where non-sphericity is significant.

In this paper, I will review the capabilities of magnetic imaging (also called Zeeman-Doppler imaging) to reconstruct spot distributions of surface fields from sets of rotationnally modulated Zeeman signatures in circularly polarised spectral lines. I will then outline a new method to measure small amplitude magnetic signals (typically 0.1% for cool active stars) with very high accuracy. Finally, I will present and comment new magnetic images reconstructed from data collected in 1993 December at the Anglo-Australian Telescope (AAT).

Modelling of magnetic Ap type stars has a long and distinguished history. The Oblique Rotator Model (ORM) – a dipole inside the star, its axis not aligned with the rotation axis – proposed by Babcock (1949a) provides a simple yet flexible enough paradigm for the modelling both of the magnetic and the spectral line variations of these stars. Deutsch (1958) developed a method to derive surface composition distributions from magnetic field measurements in conjunction with line strength variations but subsequent investigators concentrated either on the magnetic field or on the abundance distributions. Hardly ever was the question of consistency between field and composition mapping addressed – Landstreet (1988) constitutes the exception. In abundance mapping, Doppler imaging (Vogt et al. 1987) has meanwhile replaced most other approaches and is credited with fairly reliable results. But can one really carry out such mapping, as done by Hatzes (these proceedings) without accounting for the magnetic field and can these zero-field abundance maps and their relation to the magnetic configuration be compared to the predictions of diffusion theory? Did Landstreet ever have a real chance of disentangling magnetic and abundance effects using intensity (Stokes I) profiles only? What is the probability of obtaining spurious surface structure from intensity Doppler imaging of Ap stars?

Eclipse mapping is a technique to deduce spatial structure on very small angular scales in eclipsing cataclysmic variable stars (CVs). By analysing the eclipse light curve, information is obtained on the brightness structure of the accretion disk and of the compact mass-accreting object in these systems. This information would otherwise be well beyond the resolving power of any optical telescope. Since the development of the eclipse mapping technique by K. Horne, about one decade ago, it has now become an important tool in the study of CVs. Originally eclipse mapping was employed to construct brightness maps of accretion disks in broad spectral bands. Recently, maps of much higher spectral resolution have become available from which optical and UV spectra have been reconstructed in spatial detail across accretion disks. Such information is very important for our understanding of the physics of the accretion process.

In this paper I will describe the eclipse mapping technique and review recent results. In conjunction, I will briefly highlight other techniques related to the mapping of surface structure in CVs.

Stellar coronae cannot be spatially resolved, and yet stellar coronae are likely to exhibit a similar amount of structure as the solar corona. Currently, the only way to infer structure from spatially unresolved data is to look for rotational modulation of the X-ray emission or eclipses in the case of binary systems. I will discuss some of the observations obtained and critically review the methods used to infer structure from these data. Particular emphasis will be placed on the ill-conditioned nature of the inversion problem, that makes it next to impossible to infer the possibly three-dimensional structure of stellar coronae.

The concept of starspots or maculation is well established. There is indeed such a variety of corroborative evidence for those stars associated with ‘activity’ effects that we cannot imagine photospheric maculation not being a factor. The subject was reviewed by Eaton (1992), and a comprehensive background recently provided by Eker (1995), while relevant papers continue to be presented at more or less annual cool stars symposia. Even so, confidently testifying to more than the mere existence of starspots is non-trivial.

Throughout its eighteen years of operation, the International Ultraviolet Explorer has afforded us the opportunity to make many new discoveries regarding the nature of the outer atmospheres of magnetically active, late-type stars. High-resolution spectra obtained continuously throughout the rotational cycle of rapidly-rotating stars have been used to map the spatial structure of the magnetic active regions in stellar chromospheres. The RS CVn-type binary system AR Lacertae has been observed intensively at 3 epochs, with additional observations with less complete phase coverage was obtained at 4 other epochs. Taken together, we are able to probe the evolution of atmospheric structure of AR Lac over a 13-year timeline. I briefly review techniques of chromospheric emission line mapping, and I summarize the IUE results for AR Lac, V711 Tau, EI Eri, V1794 Cyg, and TY Pyx. I conclude with a description of chromospheric imaging programs (present and future) that make use of the Hubble Space Telescope.

The massive O and B-type stars will be among the first targets of the new generation of long baseline optical interferometers (such as the CHARA Array, a 400-m diameter distributed array of five 1-m telescopes). Many of these objects are binary stars for which joint astrometric and spectroscopic observations will provide masses and distances (e.g., 15 Mon; Gies et al. 1993), but there is also great interest in resolving disks of single stars. Early interferometric observations have already resolved the flattened circumstellar disks around some Be stars (Quirrenbach et al. 1994).

Until recently, all study of stellar surface structure, except for the sun, has been limited to indirect methods. This state of affairs is rapidly changing. With the introduction of interferometric techniques to optical astronomy, direct imaging of stellar surfaces is finally possible. Within a few years we will have images with sub-milliarcsecond resolution and 10 or more resolution elements across the stellar surface.

In this talk, I will describe the technique of optical interferometry and explain how it can be made to work through the earth's turbulent atmosphere. I will show some actual data and describe what can be expected in the near future.

The essence of high resolution interferometric imaging is Michelson interferometry and it has a long history. Yet the symposium programme describes it as the Last Frontier! Why was it not crossed decades ago? This paradox needs some explanation. I suspect that many participants here are tourists in interferometry, as I am in their fields, so it is interesting to look at the key developments since Michelson invented the field. It is, of course, a personal view.

The ESO Very Large Telescope (VLT) has the capability to coherently combine the light from four 8m telescopes and from smaller Auxiliary telescopes with baselines 4m and 200m (VLT Interferometer or VLTI). The resulting resolution approaches 0.5 milli-arcsec at visible wavelengths and 1 milli-arcsec in the near infra-red. We estimate that about 2000 nearby bright stars can be resolved with these baselines. The surface of about 400 stars, mainly K and M giants, can be mapped with a resolution of a dozen pixels or better across the stellar surface. This resolution permits detailed studies of the structure of stellar atmospheres, their hydrodynamics and magneto-hydrodynamics. VLTI can easily resolve starspots and very likely also convection cells on an active K-type giant at 10pc distance. A marginal detection of the largest starspots on a very active solar-type star at the same distance may be possible, but it appears very unlikely that the surface structure on an inactive solar-type star can be imaged.

We present computer simulations of stellar surfaces to study in detail the response of the interferometer to an extended, complex source. First results which simulate the end-to-end process in a simplified manner indicate that useful maps can be produced provided that a sufficient number of baselines are combined in Earth-rotation synthesis mode. We also give arguments that high spectral resolution will prove essential to constrain reconstructed maps.

We have obtained the first direct image of the surface of a star with HST in March 1995. Our near-ultraviolet images of the M2 Iab supergiant-star Betelgeuse (Alpha Orionis) taken with the COSTAR-corrected Faint Object Camera show about 10 resolution elements over the stellar disk. With direct resolution of Betelgeuse the following discoveries resulted: 1) The apparent size of the disk in the UV (≃ λ 2550) is larger by a factor of ≃2.1 than in the optical, demonstrating the extent of the supergiant atmosphere. 2) An unresolved hot spot (higher temperature than the disk by a few thousand degrees) dominates the spatial structure of the atmosphere. 3) The atmosphere at 2550 å extends to 125 milliarcsec (FWHM 108 mas) with a limb-darkening coefficient of about 1.0.

By VLBI astrometry, we show that the two RSCVn type binaries, UX Ari and σ2 CrB, have a preferred site of radio emission which is the intra-system region. It is known that radio emission from these stars is from the gyro-synchrotron process associated with large-scale magnetic fields. High gravity in the intra-system region might favor dense magnetic loops. Interactions in this region between loops attached to the surfaces of the two stellar components might produce reconnections required for electron acceleration.