new solar model consistent with these data within their errors; to have determined empirically the internal rotation as a function of depth and latitude; to have obtained a quantitative theoretical understanding of the physical processes giving rise to this differential rotation; to have detected the elusive large-scale circulation flows in the deep convection zone; to have mapped the three-dimensional structure of surface active regions and sunspots; and (if they penetrate to the surface with observable amplitude) to have made the initial detection of g-modes generated in the deep interior. This ambitious set of goals is made realistic because of the extraordinary observational and theoretical progress in helioseismology made in the '80s, coupled with realistic expectations for observational and theoretical capabilities to be developed in the '90s.
The cycle of solar magnetic activity is believed to arise in the interior by a combination of differential rotation and cyclonic convection acting on magnetic fields. More accurate and higher resolution measurements of surface magnetic fields are required to understand how magnetic flux evolves and is dissipated from the surface as part of the solar cycle. Also needed are accurate measurements of mass flows both on small and large scales to gain a better understanding of the role of magnetoconvection. Such observations, combined with some of the helioseismic results described above, should go a long way toward reaching the goal of an accurate model of the solar dynamo. A realistic goal for the '90s is one which has been sought for decades: finally to obtain a real physical understanding of the origin and nature of magnetic activity in the Sun and stars.
Observations have shown that the solar surface layers - from the chromosphere to the corona - are permeated, heated and controlled by the magnetic field rooted in the photosphere. The physics of this region is complicated by large density and magnetic variations and violent mass motions. A major observational problem is the difficulty in making accurate physical measurements in the face of spatially blurred observations. Physical quantities deduced from such blurred measurements may apply to an average within the measured volume of the quantity, but because of extreme nonlinearities they may apply to no physically realizable state at all. It is little wonder that important problems such as heating of the upper solar atmosphere or storage of magnetic energy and its violent release in flares have not been solved. The frontier in this research is very much controlled by how small a volume can be measured accurately. While space offers the most certain route to improvement, the development of adaptive optics promises significant benefits using ground-based telescopes. A realistic goal for the '90s is to obtain a clear physical understanding of the interaction of magnetic fields and convective motions at and immediately beneath the surface, and specifically to understand the surprising shredding of the field into spatially intermittent "flux knots," which appears fundamental both to the evolution of surface magnetic fields and to their consequences for atmospheric heating.
New measurement techniques will also be needed to characterize accurately the physics of the lower atmosphere. Particularly important will be observations of the magnetic field as a vector varying with height and time, along with corresponding vector mass motion measurements. The technology to make such measurements is under active development at several observatories. Advantages of diverse spectral regions, from the extreme ultraviolet through the infrared to mm wavelengths, are being exploited. Extremely important is the capability to relate physical conditions measured in the surface layers (through visible and infrared data) to conditions in the overlying heated chromosphere (through UV, extreme UV, and mm data); the data must have adequate spatial and temporal resolution to isolate physically near-homogeneous regions and to establish cause-and-effect relations between phenomena at the various levels. A reasonable goal for the end of the decade is to make substantial progress in developing and testing a specific physical description of the magnetohydrodynamic processes giving rise to atmospheric heating in these layers.
Our understanding of the corona and heliosphere was revolutionized by space observations during the last three decades. While much was learned, we still do not have a good understanding of what compels the Sun to produce a hot, X-ray emitting corona. Evidently most other stars also have coronas. Observations have demonstrated that magnetic fields play a controlling role in the morphology and large-scale dynamics of the solar corona. There is also evidence, but not proof, that magnetic processes are responsible for supplying the energy to heat the corona and to produce violent events such as flares and coronal mass ejections. The key to further progress lies in obtaining improved observations on all accessible spatial and temporal scales, but particularly on the smallest size scales. Progress in X-ray and gamma-ray imaging promises to allow