A review is given of the observational and theoretical background of global-scale velocity fields on the solar surface. A newly-developed method of reduction of the Mount Wilson velocity data is described, and the results from this new method are compared with the results of the old method. A preliminary analysis is made of the new results over a short time interval. Small-scale latitude irregularities in the differential rotation are shown to exist. Variations in time which occur in the rotation rate are broadly distributed in latitude and longitude. Although a non-solar (instrumental) cause cannot be found for these variations, such a cause cannot be ruled out at this time. Global-scale non-axisymmetric velocity field patterns intermediate between solar diameter and super-granular scale are shown to exist on the solar surface as predicted by theory.

High precision computer controlled tracings of bright Ca+-mottles were performed during 1974 and 1975 at the Locarno Observatory of Göttingen to study solar differential rotation and to search for giant cell circulation pattern. Details of the observing method and the results from observations during 1974 have been published very recently (Schröter and Wöhl, 1975). Our method consists of measuring the position of 5–15 bright Ca+-mottles with respect to the center of the solar disc every 10 to 15 min during 4 h every day. A computer controlled program determines first the center of the disc and immediately afterwards the right ascension and declination of a Ca+-mottle which was centered by the observer into a 1′-diameter spectrograph pinhole by watching a TV-monitor displaying a Ca+-K-line solar image. From a linear least squares fit of the observed positions the solar latitude and longitude were computed for the beginning and the end of the daily 4 h observation period. From this the components in latitude and longitude of the proper motions were derived which result from the differential rotation, possible giant cell circulation and the small scale random walk of these features.

From May 1972 to October 1973, daily measures were obtained of EUV coronal hole areas appearing at the central meridian. Autocorrelations of these coronal hole area time series provide synodic rotation periods which indicate an almost rigid rotation by such features for lag lengths as short as one rotation. The rotation periods of coronal holes at high latitudes best compare with inferred interplanetary field rotation periods.

We investigated the reality of the differences in rotation rates found from various features. We only used published observations including almost forgotten work of more than half a century ago.

The characteristics of the large-scale distribution of the solar magnetic fields on the basis of a series of solar magnetic synoptic charts covering more than 15 years of observations are given. The major part of our information concerns the morphology and only some results deal with the kinematics of the field distribution. Results of averaged solar magnetic field fluxes and polarity reversal studies as well as of preliminary investigation of the very-low angular resolution magnetic measurements are given. The regular zonal and sectoral distribution of photospheric background fields, the different role or visibility of structures in both polarities is discussed. The reflection of both main types of the longitudinal distribution of large-scale solar background magnetic fields (the 27-day, the 28–29-day successions, the ‘supergiant’ structures) in the interplanetary magnetic field distribution is also considered.

The observed properties of small-scale solar magnetic fields are reviewed. Most of the magnetic flux in the photosphere is in the form of strong fields of about 100–200 mT (1–2 kG), which have remarkably similar properties regardless of whether they occur in active or quiet regions. These fields are associated with strong atmospheric heating. Flux concentrations decay at a rate of about 107 Wb s-1, independent of the amount of flux in the decaying structure. The decay occurs by smaller flux fragments breaking loose from the larger ones, i.e. a transfer of magnetic flux from smaller to larger Fourier wave numbers, into the wave-number regime where ohmic diffusion becomes significant. This takes place in a time-scale much shorter than the length of the solar cycle.

The field amplification occurs mainly below the solar surface, since very little magnetic flux appears in diffuse form in the photosphere, and the life-time of the smallest flux elements is very short. The observations further suggest that most of the magnetic flux in quiet regions is supplied directly from below the solar surface rather than being the result of turbulent diffusion of active-region magnetic fields.

Coronal rotation is different in the northern and southern hemispheres of the Sun and varies with the cycle of solar activity. A sharp decrease in linear velocity by 0.4 km s−1 on the latitude and its relative constancy in the latitude interval 35° < φ < 50° are the most typical.

In the southern polar region an intensive stream of coronal gas directed oppositely to the solar rotation was detected during the time interval 1969–1972. In the greater part of the latitude range, on average, the corona rotates more slowly in comparison with the solar atmosphere at the photospheric level.

The rotation law of the Sun is often described by the Faye formula:

where ϕ is a heliographic latitude and a and b are constants. For a long time researchers have been studying the problem in what way the constants a and b change in time and in dependence on a phase of 11-year cycles.

The measurement of the magnitude of the limb effect was homogenized in time and a recurrent period of maxima of 27.8 days was found. A relation was found between the maximum values of the limb effect of the redshift, the boundaries of polarities of the interplanetary magnetic field, the characteristic large-scale distribution of the background magnetic fields and the complex of solar activity.

I am going to show two slides which should clearly prove that sunspots are drifting very regularly in the north–south direction on the solar surface.

The evolution of the velocity and magnetic fields associated with supergranulation has been investigated using the Sacramento Peak Observatory Diode Array Magnetograph. The observations consist of time sequences of simultaneous velocity, magnetic field, and chromospheric network measurements. From these data it appears that the supergranular velocity cells have lifetimes in excess of 30 h. Magnetic field motions associated with supergranulation were infrequent and seem to be accompanied by changes in the velocity field. More prevalent was the slow dissipation and diffusion of stationary flux points. These observations suggest that surface motions do not exhibit the detailed flux redistribution expected in the random-walk diffusion of magnetic fields. It is suggested that the surface motions are only the reflection of magnetic field-convective motion interactions which occur deeper in the convection zone.

The large-scale structure of the solar magnetic field during the past five sunspot cycles (representing by implication a much longer interval of time) has been investigated using the polarity (toward or away from the Sun) of the interplanetary magnetic field as inferred from polar geomagnetic observations. The polarity of the interplanetary magnetic field has previously been shown to be closely related to the polarity (into or out of the Sun) of the large-scale solar magnetic field. It appears that a solar structure with four sectors per rotation persisted through the past five sunspot cycles, with a synodic rotation period near 27.0 days, and a small relative westward drift during the first half of each sunspot cycle and a relative eastward drift during the second half of each cycle. Superposed on this four-sector structure there is another structure with inward field polarity, a width in solar longitude of about 100° and a synodic rotation period of about 28 to 29 days. This 28.5 day structure is usually most prominent during a few years near sunspot maximum. Some preliminary comparisons of these observed solar structures with theoretical considerations are given.

The data from magnetic field synoptic charts at Mt. Wilson for 16 years are separated into axisymmetric and nonaxisymmetric fields. The axisymmetric field derived simply by averaging over longitude corresponds to the general magnetic field and can be regarded as reflecting the poloidal (radial) field since bipolar magnetic fields which have been regarded as reflecting the toroidal field are cancelled out by the averaging. The evolution of pattern of the latitudinal distribution of this field shows a conspicuous appearance similar to the Butterfly Diagram of sunspots but having two branches of different polarity in each hemisphere. The two branches start from the middle latitudes, and one branch propagates towards the pole and the other toward the equator. This shows that the solar general magnetic field behaves like a quadrupole not a dipole as was previously believed. This feature is exactly what has been predicted by a numerical solar cycle model driven by the dynamo action of the global convection. Another axisymmetric field is also derived by averaging over longitude the absolute value of the magnetic field after subtracting the poloidal field. This field corresponds to the toroidal field since if this field is averaged over longitude it vanishes. The evolutionary pattern of the latitudinal distribution of this field shows a feature quite similar to the Butterfly Diagram of sunspots. These features of the two fields become conspicuous only after averaging over many rotations, e.g., over 27 rotations (2 yr): such a diagram averaged over a small number of rotations shows rather large noise.

Soft X-ray images of the inner corona obtained with the AS&E spectrograph telescope aboard Skylab revealed the presence of coronal bright points in far greater numbers than had previously been suspected. Bright points are associated with bipolar magnetic features with typical diameters of 1–2 × 104 km, mean lifetime of eight hours and magnetic flux 1019–1020 Mx. Several thousand bright points emerge over the solar surface per day, thereby bringing up more magnetic flux than is contributed by the larger active regions during the period of observation, May 1973 to February 1974. Bright points identified in X-ray photographs are seen as small (5–10″) emission features in ground-based Hα and Ca K spectroheliograms as well as in transition region lines observed in other Skylab instruments. Typical bright point temperatures are 1.5–2 × 106 K and typical densities are ∼ 5 × 109 cm−3.

It is difficult to prepare an invited paper on topics which are ABC for any specialist in Solar Physics. I considered my task as a systematization of a lot of information about cyclic and secular variations of solar activity rather than an account of new striking results in order to illustrate a certain concept. I beg your pardon in advance for possible defects in my report and for lack of time to quote all related papers.

In order to obtain new information on the manifestation of basic mechanisms of solar activity with the sunspot cycle development we have studied peculiarities of the latitude-time distribution of activation positions and suddenly disappearing filaments with respect to sunspot groups of cycles 18 to 20.