|
|
|
Advances in the 1940s-early 1960s: more quantitative
work
In the 1940s-1960s several coronagraphs
were built and operated (
Pic-du-Midi France [Lyot 1930], Arosa Switzerland [Waldmeier
1938], Climax
Colorado [Roberts 1940], Sacramento Peak
New Mexico [Evans 1953],
Haleakala Hawaii [Rush 1959], for example). A nice history of
the early period is given by Bogdan. These
instruments were used to study continuum and line coronal
emission. In hindsight, Fe XIV data examined by Waldmeier
("Die Sonnenkorona", 1951) anticipated the much later
discovery of coronal holes from EUV disk observations (
Withbroe et al. 1971). Aspects of the behavior of the corona
as it varies with the solar cycle were quantified already from
eclipse and coronagraph data by 1953 (van de Hulst, 1953,
"The chromosphere and the corona", in "The
Sun", ed. G. P. Kuiper). The solar cyclic behavior of the
magnetic dipole lines in the corona was well documented in
Billings' 1966 book "A Guide to the Solar
Corona".
Over these two decades or so, several important conceptual
and theoretical strides were also taken:
- Biermann (1947),
Woolley & Allen (1948) and
Miyamoto (1949) realized that the statistical balance between
the radiating ions of different charges is dominated by two-body
collisions, namely ionization and (radiative) recombination by
electron impact. In this case, the electron density drops out of
the ionization equilibrium, which becomes just a function of
electron temperature. This contrasts with the local thermodynamic
equilibrium approximation, in which detailed balance between
opposite processes must occur, for example between three-body
recombination and collisional ionization by electron impact. LTE
leads to the Saha formula in which the electron density occurs
explicitly in the ionization balance.
- By measuring the elemental abundances spectroscopically in
the corona, the work by Woolley & Allen also marked an
important change in the understanding of the corona. Previously,
it had been accepted that abundances are subject to large
gravitational settling effects, because of the dominance of metal
lines seen in the chromspheric flash spectrum at eclipse, and the
dominance of H and He lines in the higher part. However, with
eclipse and coronagraphic data, it was clear from their
statistical equilibrium calculations and estimates of line
intensities, that the corona was "well mixed". Heavy
elements were abundant in the corona, perhaps as abundant as in
the lower atmosphere which was indeed mixed by convection.
- From Edlen's (1943) work until 1964, many attempts were
made to determine the temperature of the ions and electrons in
the corona, with increasing levels of sophistication. The
electron temperature enters the emission line spectrum primarily
by their effects on ionization and recombination rate
coefficients. The ion temperatures are seen (as upper limits) in
the widths of the emission lines (e.g., Evans 1963). Following a
period of concern, in which the line widths gave ion temperatures
consistently lower than electron temperatures, the two methods
were reconciled finally by recognition of the importance of
"doubly excited states" accessible at coronal thermal
energies, which can auto-ionize or recombine in the electron-ion
impact dynamics. The influence of "dielectronic
recombination" was thus recognized as the missing ingredient
by Burgess, following an earlier suggestion by Unsold. (
Burgess 1964,
Burgess & Seaton 1964).
- Although the new ionization equilibrium calculations
including this process brought electron temperatures into better
agreement with ion temperatures from line widths, at the same
time, radio data were revealing electron temperatures lower by a
factor of two or so (e.g.,
Kundu 1965). The radio data remain discrepant today (
Noci 2003). This is perhaps not surprising, given the very
different dependences of coronal line and radio continuum
emission on electron temperature and the obvious inhomogeneity of
the corona.
|
