Astron. Astrophys. 356, 301-307 (2000)
1. Introduction
The extrapolation of solar magnetic fields from measurements of the
line-of-sight component using the
force-free assumption
![[EQUATION]](img2.gif)
or equivalently
![[EQUATION]](img3.gif)
holds fairly well in the corona, where the plasma
is much smaller than unity. Note that
modern magnetographs (THEMIS: Mein & Rayrole 1985, Rayrole 1992;
LEST: Stenflo 1985) will provide all of the three components of
, at least at one level in the
photosphere, and therefore from the set of Eqs. (2) written
as
![[EQUATION]](img6.gif)
will enable us to determine only
from the vertical component of the
density current vector (3d) if (1)
holds , but not from (3a) or (3b) since the vertical derivatives
and
cannot be easily estimated from
observations. On the opposite, system (3) can be used to extrapolate
to the lower corona (Wu et al. 1985,
1990; Cuperman et al. 1989, 1990), but lots of difficulties arise
because of the ill-posedness of the problem (Amari et al. 1997). Other
methods, existing or presently under development, are based on
iterations or MHD codes (Amari & Démoulin 1992; Amari et
al. 1997; Démoulin et al. 1997; Mc Clymont et al. 1997) and
they are expected to produce stable solutions.
Nevertheless, at the photosphere
holds, and the force-free balance (1a) must be replaced by the
magnetostatic balance (4a) on the horizontal
![[EQUATION]](img14.gif)
as far as inertial terms of acceleration
J m
and advection
J m
are negligible with respect to the pressure force
J m
and the magnetic force
J m-3),
estimated on the basis of typical values
m s-1;
T;
m;
s;
Pa;
kg m-3).
This is a starting assumption, since we know that velocity and
magnetic field are well correlated in sunspots (Berton 1986).
Moreover, there is serious observational evidence that magnetic forces
are important up to 400 km above the photosphere and become negligible
beyond (Metcalf et al. 1995).
If horizontal pressure gradients are by some means measured, one
may think to solve Eqs. (4) with standard boundary conditions
( known at the photosphere). In a
first step brightness fluctuations I currently observed across
sunspots may be related to horizontal temperature gradients through
the blackbody assumption of radiation transfer
![[EQUATION]](img27.gif)
where h and k denote respectively Planck and
Boltzmann constants, and the absolute temperature T is
connected to the thermal pressure p by the perfect gas law
![[EQUATION]](img28.gif)
where the density is assumed to
vary only with height
( m2 s-2 K-1
in the photosphere). Actually, the relative brightness
is measured at the photospheric
level
![[EQUATION]](img32.gif)
with respect to the continuum value
far from the line centre.
With this background, the relationship between I, p
and T will be discussed in Sect. 2, the calculation
algorithms will be explained with demonstrations of the well-posedness
in Sect. 3, and prospects will be proposed in the concluding
Sect. 4.
© European Southern Observatory (ESO) 2000
Online publication: March 28, 2000
helpdesk.link@springer.de  |