Eberhart Jensen

The Prominence-Corona Transition Region in transverse magnetic fields

An emission measure analysis is performed for the Prominence-Corona Transition Region (PCTR) under the assumption that the cool matter of quiescent filaments is contained in long, thin magnetic flux loops imbedded in hot coronal cavity gas. Consequently, there is a transition region around each thread.Comparison of the model and observations implies that the temperature gradient is perpendicular to the magnetic lines of force in the lower part of the PCTR (T < 105 K). It is shown that in this layer the heating given by the divergence of the transverse conduction fails to account for the observed UV and EUV emission by several orders of magnitude. It is, therefore, suggested that the heating of these layers could be due to dissipation of Alfvén waves.In the high-temperature layers (T ≥ 105 K), where the plasma β ≥ 1, the temperature gradient is governed by radiative cooling balancing conductive heating from the surrounding hot coronal gas. Also in these outer layers the presence of magnetic fields reduces notably the thermal conduction relative to the ideal field-free case. Numerical modelling gives good agreement with observed DEM; the inferred value of the flux carried by Alfvén waves, as well as that of the damping length, greatly support the suggested form of heating. The model assumes that about 1/3 of the volume is occupied by threads and the rest by hot coronal cavity matter.The brightness of the EUV emission will depend on the angle between the thread structure and the line of sight, which may lead to a difference in brightness from observations at the limb and on the disk.

Kinematics of a loop prominence

The kinematics of a loop system has been studied from high resolution Ca ii K line spectra and Hα filtergrams recorded at Oslo Solar Observatory.Emission features are seen to fall at supersonic velocities from the top of the arches towards the chromosphere. Our data are in agreement with the assumption of matter falling freely along a dipole type magnetic field of maximum height 100–150 thousand km. There is a slight asymmetry between positive and negative line shifts which can be accounted for as a tilt of the individual loops relative to the plane of the sky of 5–10°. The planes of the loops are also inclined by a small angle of approximately 15°. It appears that matter starts from rest at the top of the loops. An observed tilt of some emission features in the K line spectra may be explained by a gradient in the line-of-sight velocity with height caused by the curvature of the dipole type loops.

The wilson effect and the transparency of sunspot models

Hydrostatic models of sunspot penumbra and umbra are evaluated using Bodes tables of monochromatic absorption coefficients andT-τ-relations given by Makita and Morimoto (1960) and by Zwaan (1965). These models are placed side by side to simulate a complete sunspot corresponding to an area of 480×10−6 of a hemisphere. Intensity profiles are evaluated for aspect angles up to 85° and compared to observations. The primary aim was to study the influence of spot transparency, which is closely related to the gas-pressure, on the Wilson-effect and on other changes in the intensity profile that appear close to the solar limb. The gas-pressures at the zero-level in the geometrical depth (z=0) corresponding to optical depth, τ=10−3, both in the umbra,P0u, and in the penumbra,P0p, appear as adjustable parameters.When curvature is taken into account, the Wilson-effect cannot be reproduced without depressing the zero-point in the geometrical scale in the umbra relative to the same layer in the photosphere. A depression of 400 km will give a reasonably good fit for the Wilson-effect providedP0u<P0P<P0Ph. The model we found to give the best fit is based on Makita and MorimotosT-τ-relations withP0P=3200 andP0u=800.We have here chosen an umbra pressure that gives a small limb-side intensity peak at the penumbra border, assuming that the bright points described by Bray and Loughhead (1964) may be interpreted in terms of a transparency effect.Other parameters measured by Wilson and Cannon (1968) are evaluated, and for some a good agreement was obtained, while for others only a qualitative effect in the same direction could be found.Surfaces along which the optical path is constant (isodiaphanous surfaces) are functions of aspect angle and well suited for visualizing the transparency in spots.It is shown how for a wide range of models the isodiaphanous surfaces get asymmetric close to the limb. This has consequences for the interpretation of the Evershed-effect. In fact, under certain conditions, a ‘masking’ effect may take place because the greater transparency in the penumbra will allow observations of a deep laying flow, which will not be visible through the more opaque photosphere. Due to the asymmetry this effect is different on the limb side and the center side.We also found the spot to show an apparent displacement away from the limb, which at a heliocentric distance of 85° amounts to about one second of arc. Intensity profiles in the near infrared at 8206 Å and at 16482 Å are evaluated, and the importance of observations in these spectral regions is emphasized.

OSLO SOLAR OBSERVATORY

Abstract : The report discusses the staff, instruments, research program, and publications of the Oslo solar observatory.

Filament oscillations as evidence for Alfvén waves

High resolution observations of quiescent filaments show oscillations that are strongly tied to their fine threads. It is shown that neither slow nor fast MHD modes may account for the observations, which rather are in accordance with Alfvén waves.

Velocity fields in quiescent prominences

Three quiescent prominences were observed in the Ca ii K-line and a fourth one also in the H-line at Oslo Solar Observatory, Harestua, and reduced by Rustad (1974) and by Engvold et al. (1980). These data are used to study the distribution of the line-of-sight velocity component, N(u0). It is pointed out that in a stationary and isotropic case, N(u0) should be a gaussian distribution. For each of the sets of measurements gaussians were therefore fitted by a least square procedure. The range in observed velocities varies considerably between the prominences. For the best observed prominence more than 70% of the kinetic energy is in the supersonic range. In the other cases none or only an insignificant part of the observations exceed the velocity of sound. Considerable deviations from gaussian distributions are apparent for the smallest velocities. This distortion shows up conspicuously in the slope of the energy spectrum, a parameter that may be used as a rough measure of spectral resolution.If it is assumed that we have to do with MHD-turbulence as described by Kraichnan (1965), a characteristic relationship should exist between velocity and eddy size. When supersonic velocities are present, compressibility effects may severely alter this relationship. The possibility of observational confirmation is discussed.If a turbulent velocity field is indeed present, the heat conductivity and other transport coefficients may be significantly altered as compared to the atomic values.

TURBULENT VELOCITY FIELDS IN QUIESCENT PROMINENCES

The velocity fields of prominences are derived from measurements of CaII K-line shifts in high resolution spectrograms. The turbulent character of the velocity field has been reported earlier (Jensen, 1982). That work was based on spectra of 4 large prominences observed at Oslo Solar Observatory. A much larger set of data is now beeing analysed. These spectrograms are recorded with the main spectrograph at the tower telescope at the Sacramento Peak Observatory. Details of observations are given in an earlier paper (Engvold, 1978). We shall here present and discuss the results obtained from nine quiescent prominences observed during 1973–74. The spectrograms have been analysed by means of the rapid scanning, computer controlled microphotometer of Institute of Theoretical Astrophysics, University of Oslo. The digitally recorded displacements within a given velocity-interval, giving the distribution of radial velocities, was denoted by N(u). In well-observed cases altogether 3–4000 profiles were obtained. Gaussian distributions were then fitted to the observed velocity-distributions by a least square procedure, giving a dependence of the form; N(u) ∝ exp (−(u−uo)2/α2). Here uo represent the combined rotational and bulk radial velocities, while α is a characteristic velocity parameter, describing the “turbulent exitation” in the prominence. In only one of the prominences investigated, did the velocity-distribution seem completely random and could not be represented by a Gaussian. In all the other cases the correlation-coefficient, r2, came out in excess of 0.9. This result indicates that the velocity-field, at least to a first approximation, may be assumed to be isotropic and stationary.