Regions of enhanced radio brightness in the polar areas of the Sun were detected in 1974 at the Crimean Astrophysical Observatory of the Academy of Sciences of the U.S.S.R. The effective temperatures of these regions exceeded that of the undisturbed Sun by 1500 K and 2200 K at 8.2 and 13.5 mm wavelengths respectively. Radio brightness distributions of the Sun obtained with the 22-m radio telescope, at 8.2 and 13.5 mm with angular resolutions of 1!6 and 2!55 respectively are shown in Figure 1. Regions of enhanced radio brightness near the North Pole of the Sun are seen, beside the typical s-sources at middle and low latitudes.

This paper is a shortened version of a review of radio emission mechanisms for meter-λ radio bursts.

The centimeter -wave bursts are simple: they are characterized by a rapid rise in intensity and a slower decline. The burst-radiation is in general smooth, usually free of details in time and frequency and is a partially polarized broadband continuum. It starts almost immediately after an associated Hα flare and originates from a source of small angular size (< 1′ arc). The smooth continuum emission of centimeter-wave bursts has been accounted for by synchrotron radiation from fast electrons accelerated during flares. I shall not attempt to give here a full description of the observational characteristics of centimeter bursts, rather I shall concentrate on certain specific features relevant to the understanding of the generating mechanism of these bursts. More than 20 years ago, centimeter bursts were studied with a spatial resolution as good as ∼ 1′ arc. Even with this resolution the presence of small scale structures on scales of 1′–4′ were observed in cm-λ bursts sources. Since that time both spatial and time resolution have steadily improved; we now have a spatial resolution of ″ and time resolution of ∼ 10 ms. Such resolutions have revealed smaller scale structures of a few arc seconds and more complexity in burst sources. At the same time soft x-ray observations, particularly from Skylab, with resolution better than 5″ (e.g. Kahler et al. 1975; Vorpahl et al. 1975) have given new information about the flare phenomenon. Because of the close association of soft x-ray and cm-λ burst sources, these observations are particularly important for a proper understanding of the generating mechanism of cm-λ bursts, acceleration of electrons and the origin of solar flares.

We discuss the microwave emission from a flaring loop (Spicer 1977). In particular we examine the following question: What will be the characteristics of the radio emission at centimeter wavelengths from a small compact flaring loop (average plasma density ne ≃ 1010 cm−3, average magnetic field at the footpoint of the loop Bℓ ≃ 500 gauss and Bt ≃ 100 gauss at the top of the loop and length of the loop, L = 109 cm), when the mechanism which pumps magnetic energy into the plasma in the form of heating and/or electron acceleration satisfies the following conditions. a. The magnetic energy is released in a small volume, (the energy release volume (ERV)), compared to the volume of the loop, and the rate at which magnetic energy is transformed into plasma energy is faster than the energy losses from the same volume. This causes a local enhancement of the temperature by as much as one or two orders of magnitude above the coronal temperature. b). The bulk of the energy released goes into heating the plasma and heats primarily the electrons (Te > Ti). Using these two assumptions one can easily show (Brown, Melrose and Spicer 1979, Vlahos and Papadopoulos 1979) that the high energy electrons in the tail of the velocity distribution in the ERV will instantaneously run away from this volume, and the resulting charge imbalance between the ERV and its surroundings (which still have average coronal temperatures ~ 106 K), will drive a return current, with velocity VD. When VD reaches the value of the local sound speed Cs ≃ 107 cm/sec low frequency ion acoustic waves will be excited at the interface of the ERV and its surroundings. It has been shown that the heat flow along the magnetic field lines is greatly reduced due to the presence of ion-acoustic turbulence (cf. Manheimer 1977). The bulk of the electrons in the ERV have electron-wave collision times **τw << 10–100 sec, longer than the impulsive phase of the flare. But since τw ~ v3 for those electrons with velocity v > ve (see Rudakov and Korablev 1966) the electrons in the tail will not “see” the ion sound waves and will stream freely towards the chromosphere.

The microwave and hard-X-ray burst of 1970 March 01, 11:27 UT was found to originate from a common thermal plasma with a maximum temperature of 57 keV. The low coronal plasma with an average electron density of about 3.108cm−3 covered a projected area of 5.1018 cm2. In Fig. 1 the time profiles of the emission measure and the temperature are compared with the 10.5 GHz flux while Fig. 2 shows the reversible relationship between the hard X-ray emission measure and temperature during the impulsive phase. The arrows indicate the direction of increasing time. The dashed-dotted line, representing an adiabatic process with an index χ = 5/3, agrees well with the observations showing a compression followed by an expansion (Mätzler et al. 1978).

The spectra of a set of 77 microwave bursts of July 1971 to September 1974 as published in Solar Geophysical Data have been fitted with an idealized spectrum of triangular shape. The spectral index of the low frequency part α, the maximum flux density and the associated frequency have been determined. A cluster-analysis carried out on these parameters showed that the sample consists of only one class of events. In the histogram of the spectral index we found a most frequent value of 1.4. Such a low value cannot be explained by the mechanisms normally assumed to account for the low frequency attenuation, as gyroresonance absorption and the Razin effect.

During the radio observations of the Sun performed at λ = 21 cm on July 1st, 1974 using the Westerbork Radiotelescope (Bregman and Felli 1976; Chiuderi-Drago et al. 1977) four bursts were recorded. Two of them, which present some interesting peculiarity, that it would not be possible to observe with an instrument of lower resolving power, are discussed in this paper.

Observations of five impulsive microwave bursts were made during July 15–18, 1978 using the VLA at 4.9 GHz with temporal and spatial resolutions of 10 s and up to 1″ respectively. A series of 2-dimensional snapshot synthesis maps was made as a function of time, and compared with Hα photographs and magnetograms. Each flare was characterized by a single source of constant size, located on the magnetic neutral line between the Hα kernels.

A preliminary account is presented of some characteristics of Fast Fine Structure, that is fine structure with time scales in the milli second range, and of the flare events in which they appear. Fast Fine Structure is found to be a fairly common phenomenon that could, but not necessarily does, appear in flare events of any importance occurring at any longitude. It may occur with any type of microwave burst (except most likely “Rise” and “Fall”). It appears more often in the more important events and it may last for only a few seconds up to many minutes. Fast Fine Structure may have any degree of circular polarization, apparently independent of its longitude position.

The use of increased sensitivity and temporal resolution at mm-wavelengths has revealed a number of novel features associated with solar activity (Kaufmann et al., 1979a). Of particular interest are the ultra-fast time structures superimposed on impulsive bursts. They are distinguished by time scales and repetition rates much shorter than the impulsive times scales. For weak bursts at λ = 13mm (i.e., few s.f.u.) the superimposed spikes are separated in time by several seconds. For moderate fluxes (i.e., 100 s.f.u.) the repetition rate of superimposed spikes is ~ 10–15 per second (i.e., time interval between two repeated spikes ~ 80 msec). For bursts larger than 500 s.f.u. the predicted times between repeated ultra-fast structures are shorter than 10 msec. Various examples have been studied with a time resolution better than 10 msec. (Kaufmann et al., 1979b). The impulsive flux level increases with the repetition rate of the ultra-fast component, following a nearly linear relationship. The simplest interpretation suggests that the injected spikes are associated with the bursting source function and are quasi-quantized in energy (Kaufmann et al., 1979b).

Several classes of “normal” stellar objects exhibit microwave phenomena, predominantly at cm wavelengths. These include RS CVn binaries, Algol-type binaries, M supergiants, and UV Ceti-type flare stars. The emission mechanisms are generally either synchrotron or gryosynchrotron radiation, with the latter being predominant in the numerous RS CVn radio emitters which sometimes exhibit variable circular polarization. “Events” have time scales varying from minutes to days.

Our success or failure in understanding solar bursts is largely determined by the kinds of instrument with which we observe them. For example, although the basic feature of Type II bursts, a slow drift from high to low frequencies, was recognized and correctly interpreted from radiometer measurements at a few frequencies (Payne-Scott et al., 1947) other properties of Type II bursts such as fundamental-harmonic structure, split bands and herringbone structure can only be recognized on dynamic spectrograms. For this review I have chosen to group together the observations made with a particular type of instrument. I have also tended to emphasize what remains to be done rather than what has already been done. Unfortunately, with so many topics to choose from my treatment can only be cursory; and in attempting to select the interesting topics I have inevitably been biased toward the work I know best - that is, the work of the group to which I belong.

Results on the space-time evolution of type III burst sources are summarized (Raoult and Pick, 1979) and observations on the temporal characteristics of the circular polarization are presented. It is shown that type III burst sources can be resolved into narrow components. Complex or very large sources may be explained by the occurrence of electron beams which propagate along different diverging paths. Propagating effects in a scattering corona are not required to explain the observations. The origin of the fundamental-harmonic pairs is briefly discussed.

Herringbone Type II bursts were first described by Roberts (1959). These bursts are distinguished by their unusual structure - rib-like features (herringbones) extending on either side of a Type II backbone.

(Paper to be submitted to The Astrophysical Journal)

A first solar radar experiment in microwaves has been carried out with the 300 m dish in Arecibo. *The goal was to scatter the radio waves at 2380 MHz on coronal Langmuir waves in the range from 170 to 270 MHz. The nonlinear interaction of these two wave modes produce a radio signal in backward direction at the beat frequencies

Receiption was at 2600 MHz with a bandwidth of 100 MHz. The transmitter of 250 kW power was pulsed in periods between 0.5 and 1.5 s. The received flux was Fourier analysed and searched for the transmitter period.

(paper submitted to Astronomy and Astrophysics)

We propose a model for type I emission based on scattering of Langmuir waves by ion acoustic waves, which are associated with the evolution of the associated active region on the Sun.

During noise storms narrow-band, short-lived bursts make up an important part of the activity. Frequently the bursts show different characteristic arrangements in the frequency-time plane (Elgaröy, 1977). In the present investigation the frequency of occurrence of single bursts and sequences of bursts has been determined. The work started out as a search for statistical evidence for the occurrence of echoes of type I bursts. As progress was made, more attention was devoted to the general question of what was the frequency distribution of events containing different numbers of bursts and the conclusions that could be drawn from the observed distribution.

In the first part we present the analysis of digitally recorded type-I bursts. The reduction procedure is described and some typical examples and results are presented.

In the second part we propose a model for type-I emission, compatible with low magnetic field strengths.

Incoherent langmuir waves, spontaneously emitted by trapped fast electrons are converted to transverse radiation by coalescence with low frequency whistlers. Recurrent upper hybrid instability in strong m.h.d. pulses explains the bursts in this model.

The interpretation of ground based observations of solar “plasma frequency” radio bursts has been hampered in the past by an insufficient knowledge of coronal scattering by density inhomogeneities close to the sun. Calculations based on measurements of the angular broadening of natural radio sources, and Woo's 1975 measurement of the angular broadening of the telemetry carrier of Helios I near occultation (Woo, 1978), indicate that plasma frequency solar bursts should undergo considerable scattering, at least near the maximum of the sunspot cycle. The calculated displacements of the apparent positions of the bursts are about equal to the observed displacements which have been attributed to the bursts occurring in dense streamers. In order to obtain more scattering data close to the sun, interferometer measurements of the angular broadening of spacecraft signals are planned, and the important contribution which could be made with large dishes is discussed.

Complementary bursts (C.B's) have been observed in the decametric range during noise storms and/or type IV activity. These bursts essentially consist of two components, each component having a duration ~ 1 second. The first component shows weak emission or emission gap over a certain frequency range. The second component is observed after a certain delay. If the bursts are assumed to be generated at the fundamental, and if the radiation corresponding to the gap propagates through an electron density irregularity located close to the source along the line of sight, whose cross-section is less than the linear extent of the source, then almost all properties of the C.B.'s can be explained. High sensitivity, and high frequency and time resolution spectra of type IV bursts at 137 MHz revealed new microscopic spectral features displaying “wave-like” and “fork-like” shapes.