Skip to main content

Full text of "Evolution of a Magnetic Flux Tube in a Sunspot Penumbra"

See other formats

Evolution of a Magnetic Flux Tube in a Sunspot 

K. Jahn 1 ' 2 , R. Schlichenmaier 3 , H.U. Schmidt 1 

1 Max Planck Institut fur Astrophysik, 

Karl Schwarzschild Str.l, D-85748 Garching, Germany 

2 Astronomical Observatory of the Warsaw University, 

Al. Ujazdowskie 4, 00-478 Warsaw, Poland, (e-mail: 

3 Max Planck Institut fur Extraterrestrische Physik, 
Karl Schwarzschild Str.l, D-85748 Garching, Germany. 

(Received ) 

The motion of an individual magnetic flux tube inside the penumbra of a sunspot is studied 
numerically. Here, we present preliminary results. The thin flux tube approximation together 
with a simplified radiative heat exchange with the surroundings is used to study the evolution of 
a flux tube embedded into a background given by a global magneto-static sunspot model. The 
investigation is undertaken in order to verify the conjecture that convection in sunspot penumbrae 
occurs by an interchange of magnetic flux tubes. The code being developed can be used to study 
dynamic aspects of filamentary structure in the penumbra: the temporal and spatial fluctuations 
of the temperature and the magnetic field, the motion of bright penumbral grains, or the Evershed 
effect. Here we present the evolution of a wave formed by the tube whose fragment emerges in 
the penumbral photosphere and migrates towards the umbra. The properties of this wave show 
qualitative features of the observed bright penumbral grains with corresponding upward velocity 
and its correlation with brightness and the inclination of the magnetic field, and also of the 
Evershed effect. 


High resolution observations of filamentary sunspot penumbrae give some de- 
tailed hints at a particular type of convection that seems to occur there (see an 
extensive discussion in Jahn & Schmidt, 1994). In summary, the inclination of the 
magnetic field exhibits substantial azimuthal fluctuations which preserve their ra- 
dial coherence and are well correlated with the brightness of filaments (Title et al., 
1993): the dark ones are permeated by an almost horizontal field whereas the bright 
ones contain a more vertical field. The velocity field shows similar correlations: the 
horizontal Evershed effect occurs in the dark regions and upward flow in brighter 
ones. Additionally, an inward motion of bright penumbral grains towards the um- 
bra is observed (Muller, 1992), which may be associated with the inward part of a 
balanced two-way migration of the magnetic flux. Those features observed in the 

pcnumbral photosphere concur with the concept of an interchange convection in 
the penumbra which appears to have a deep structure (Schmidt 1987, Jahn 1989, 
Solanki & Schmidt 1993). This hypothetical convection would be accomplished by 
a non-local exchange of azimuthally narrow but radially elongated magnetic flux 
tubes that extend many scale heights below the photosphere (Schmidt 1991, Jahn 
1992). It would set in when the total flux and with it the inclination of the field lines 
at the edge of a magnetic flux concentration exceeds some critical value. Rucklidgc 
et al. (1995) studied an abrupt development of the penumbra with a simple model 
based on this concept. Some anticipated properties of interchanging flux tubes have 
been used to approximate the average structure of a deep penumbra by a one di- 
mensional stratification of the matter, and to construct a tripartite magneto-static 
model for a sunspot (Jahn & Schmidt 1994). Now we intend to construct a 2-D 
model for the penumbra based on a more detailed study of the dynamics of the 
interchange convection. This might be achieved in an iterative manner by following 
the non-linear time-dependent evolution of a single magnetic flux tube embedded 
into a background which asymptotically assumes the thermodynamic and magnetic 
properties determined by the properties of the flux tube itself averaged over all 
phases of its motion. 

In this progress report we present preliminary results of a study of the migration 
of a flux tube embedded into the fixed pregiven penumbra of the tripartite model, 
and compare its dynamic properties with the observed evolution of bright pcnumbral 

FIGURE 1. Parameters of the tripartite sunspot model (cf. Jahn & Schmidt 1994) 


The structure of the magnetic flux element has been approximated by a f-D 
thin flux tube, and the Lagrangian formalism has been adapted to describe the 2-D 
dynamics of the tube. The equations governing the time evolution of the tube are 
basically the same as those used by Moreno- Insertis (1986) except that we allow for 
the radiative heat exchange between the tube and the surroundings and calculate 
the rate of the entropy change due to radiation. This is made in a similar manner 
as by Montesinos & Thomas (1993) who distinguished between large optical depths 
(r > 2/3), where the entropy change is proportional to the temperature difference 
between the tube and the surroundings (according to Spiegel's (1957) analysis), and 
small optical depths (r < 2/3), where this change is proportional to the difference 
in the local values of the Planck function (Kalkofen & Ulmschneider 1977). 

Another difference is related to the 2-D nature of the force-free magnetic field 
of the background in which our tube is embedded. Although the stratification of the 
gas in the tripartite model is one dimensional, the magnetic pressure has a lateral 
gradient, so that the distribution of the total pressure (gas plus magnetic) is two- 
dimensional. This leads to additional terms in the equation of motion and in the 
condition of mechanical equilibrium of the tube with the surroundings. Finally, the 
set of equations describing the tube's evolution is completed by the same equation 
of state as that used in the tripartite model, i.e. with partial ionization effects 
included for a gas of solar chemical composition. 

All parameters of the particular tripartite model used as background in the 
present simulations are given in Fig. 1 (see also Fig.l in Jahn & Schmidt, 1994). 


A thin tube with a magnetic flux equal to 10 15 Mx is originally placed on the 
inner side of the magnetopause of the tripartite model. The corresponding diameter 
of the tube at the penumbral photosphere is equal to about 10 km. Such a tube is 
much too small to represent directly any of the fine-structure elements observed in 
the penumbra, but our simplified treatment of the radiative transfer does not allow 
to consider much larger flux tubes. Otherwise, we would have to take into account 
an internal radiative diffusion. The initial stratification along the tube is identical 
to that in the penumbra. The end points of the tube have been placed far from the 
photosphere of the penumbra: the lower one at a depth of 15 Mm and the upper 
one on a sphere 25 Mm away from the center of the spot. For the time being both 
points are kept at fixed positions on the magnetopause during the evolution of the 

When the tube is slightly perturbed locally, its near photospheric portion at 
depths less than about 500 km, starts to oscillate across the magnetopause, be- 
cause of two competing forces: the buoyancy which expels the tube from the more 
dense stratification of the quiet sun, and the curvature force acting in the oppo- 
site direction. But each time the tube intrudes into the quiet sun it heats up by 
radiative exchange which is still efficient at those depth in a nonlinear fashion, so 
that the amplitude of the oscillation grows with each cycle. Eventually, the tube 
gains a buoyancy large enough to overcome the curvature force, then emerges in the 
penumbral photosphere and migrates further towards the umbra. 

Figure 2 shows the geometry of the outer part of the penumbra and the shape 
of the tube at sample times differing form each other by five minutes. The arrows 

8 10 12 14 


FIGURE 2. Migration of a flux tube in the penumbra plotted with a time step of 5 
minutes. Dashed line corresponds to the magnctopause, and arrows show the flow 
of the matter. Insertions show the time history of the three points crossing the 
photosphere: the emerging (continuous line) and sinking one (dotted line). Shown 
are: fluctuations of the density, temperature, and the magnetic field relatively to 
the corresponding background values, the inclination of the tube to the vertical di- 
rection, vertical component of the velocity of plasma in the tube, and the distance 
of the points from the spot's center 

along the tube indicate the velocity of the matter. The length of arrows corre- 
sponding to the Alfven and sound speed at the photosphere are also marked in the 
figure. Initially the tube crosses the penumbral photosphere only once, but as it 
proceeds inwards its part laying above the photosphere near the edge of the spot 
sinks, so that the tube assumes a sine-wave shape with three intersections with the 
photosphere. The time history of different parameters of the two inner intersection 
points arc plotted in the insertions, where continuous and dotted lines correspond to 
the ascending and the descending intersection respectively. The third intersection 
retracted after about 10 minutes and never intruded further than 500 km into the 
penumbra. For a short time two more intersections appeared near the penumbral 

boundary, because the tube continuously oscillates across the magnetopause in this 
region. The last shape of the tube shown in the Fig. 2 may suggest that most prob- 
ably there will soon be generated a second wave crossing the surface and traveling 
inward. However, problems with our oversimplified lower boundary condition do 
not allow to continue the run. Therefore, we concentrate at present on the features 
of the first wave which traverses about 2.5 Mm in the penumbra during the sim- 
ulation. It forms a flattened arch above the photosphere with a typical height of 
about 50 km which has a length that slowly increases with time from about 300 to 
800 km (see insertions in Fig. 2). 

The flow of the gas in the arch is directed opposite to the migration of the tube. 
It assumes speeds close to sonic, or even supersonic in the descending intersection. 
It should be noted that our code does not yet allow a proper treatment of shocks. 
This streaming along the arch has a nature similar to a siphon flow (Meyer & 
Schmidt, 1968). The velocity of this flow in our model, of about 7 km/s, is larger 
than the observed maximum velocity of the Evershed effect, i.e. 6 km/s, reached at 
the outer penumbra boundary (see e.g. Mullcr, 1992). 

Properties of both intersection points show some qualitative similarities to those 
observed in penumbral fine-structure, i.e. flow velocities, magnetic field directions, 
and their correlations with brightness. Thus, the ascending intersection has a more 
vertical magnetic field than the mean penumbral field in the model. The matter 
flows upward there and is hotter than the ambient medium. That concurs with 
the upward velocity of the bright matter measured in penumbrae (Beckers 1981, 
Degenhardt & Wiehr 1991). The other intersection is more horizontal, but shows a 
downflow with a substantial vertical component of the velocity as the gas is cooler 
than the surrounding background in the photosphere of the penumbra. 

It is tempting to identify the ascending intersection of the arch with a bright 
penumbral grain, despite the fact that it does not possess all the quantitative fea- 
tures of the grains. It moves inward with a speed of about 2 km/s, the descending 
one somewhat slower. These values are by a factor of four larger than the ob- 
served velocities of the penumbral grains (Muller, 1992). The large velocity may 
be ascribed to the mechanism of heating of the flux tube in our present model. It 
is achieved by the radiative exchange whenever the tube intrudes into the quiet 
sun. The radiative time scale increases abruptly with depth below the photosphere 
due to increasing opacity. Thus, the tube can acquire the energy efficiently only 
over a limited range of depths, so that the buoyancy can be increased in a short 
fragment of the tube only, which then rises thereby forming a relatively short and 
fast-moving wave. One might expect that more efficient heating at larger depths 
might be achieved by complex magneto-convective processes. This then would make 
the tube buoyant over larger lengths, and thereby would induce longer and slower 


We have developed a code that allows investigations of the motion of a thin flux 
tube embedded in a pregiven magnetic field and background stratification extending 
over many scale heights in the solar convection zone and atmosphere. Preliminary 
simulations presented here show the effect of radiative heating of the flux tube 
immediately outside the magnetopause occurring over a relatively small range of 
depths. It induces a wave-like inward migration of the tube, and the properties 

of the migrating wave correspond in general to the observed features of the fine 
structure in the penumbra, although the particular values differ by a factor of a few 
from the observed quantities. Probably, a better agreement can be achieved with a 
more efficient heating of the tube at larger depths. That requires, however, a model 
for the heat exchange across an inclined magnetopause in convectively unstable 
stratification, like that of Schmidt et al. (1986). Also a correct treatment of shocks 
should be included in the code, because supersonic velocities along the tube appear 
in the photospheric layers. 

Acknowledgments. We acknowledge the support of the NATO Advanced 
Study Institute. KJ acknowledges the hospitality and the support from the Max 
Planck Institut fur Astrophysik in Garching. RS likes to thank Prof. G. Hacrcndel 
for providing the possibility to work on this subject. This work was also supported 
by the KBN grant 2 P304 004 007. 


Beckers J.M, 1981, in " The Sun as a Star", S. Jordan (ed.), NASA-SP-450, Wash- 
ington, p. 11 

Degenhardt D., Wiehr E., 1991, A&A 252, 821 
Jahn K., 1989, A&A 222, 264 

Jahn K., 1992, in "Sunspots: Theory and observations", J.H. Thomas & N.O. Weiss 

(eds.), Kluwer, Dordrecht, p. 139 
Jahn K., Schmidt H.U., 1994, A&A 290, 295 
Kalkofen W., Ulmschneider P., 1977, A&A 57, 193 

Meyer F., Schmidt H.U., 1968, Zeitschrift f. Angew. Mathematik u. Mechanik 48, 

Montcsinos B., Thomas J.H., 1993, ApJ 402, 314 
Moreno-Insertis F., 1986, A&A 166, 291 

Mullcr R., 1992, in "Sunspots: Theory and observations", J.H. Thomas & N.O. 

Weiss (eds.), Kluwer, Dordrecht, p. 175 
Rucklidge A.M., Schmidt H.U., Weiss N.O., 1995, MNRAS 273, 491 
Schmidt H.U., 1987, in "The role of fine-scale magnetic field on the structure of the 

solar atmosphere", E.H. Schroter, M. Vazquez, A. A. Wyller (eds.), Cambridge 

University Press, p. 2 19 
Schmidt, H.U.: 1991, Geophys. Astrophys. Fluid Dyn. 62, 249 
Schmidt H.U., Spruit H.C., Weiss N.O., 1986, A&A 158, 351 
Solanki S.K., Schmidt H.U., 1993, A&A 267, 287 
Spiegel E.A., 1957, ApJ 126, 202 

Title A.M., Frank Z.A., Shine R.A. et al., 1993, ApJ 403, 780