Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 348, L37-L40 (1999)

Previous Section Next Section Title Page Table of Contents

2. Observational results

2.1. The observations

The observations on which we base our analysis have been described in detail by Rüedi et al. (1998). Here, we recall only their most important characteristics.

The slit was placed in the heliocentric E-W direction through the center of a sunspot located on the limbward side of active region NOAA 7958 on April 21 1996 ([FORMULA]). The full Stokes vectors of the extremely magnetically sensitive Ti [FORMULA] lines at 2.2310 µm ([FORMULA]) and 2.2211 µm ([FORMULA]) were observed using the NIM (Near Infrared Magnetograph, Rabin 1994). These lines are also very temperature sensitive and are present only in cool regions such as sunspot umbrae (Hall 1974, Rüedi et al. 1995). Similar observations of this sunspot were obtained a few hours earlier in the Fe I 1.5648 and 1.5652 µm lines using the same apparatus and at approximately the same slit location.

Fig. 1 shows the most important parameters obtained from the analysis of the Ti [FORMULA] observations (carried out using an improved version of the inversion code of Solanki et al. 1992): the magnetic field strength, B, the inclination angle of the magnetic vector to the vertical, [FORMULA], and the Doppler velocity. In most cases two magnetic components, with very different characteristics were necessary to satisfactorily reproduce the Ti [FORMULA] -line profiles. The triangles represent the component with larger magnetic field strength, which is fairly vertical and shows no significant Doppler shift, while the diamonds represent the weaker field component, which is closer to the horizontal and shows a strong signature of the Evershed effect, even at the moderate spatial resolution of these observations of about 3". Each of the two distinct cool magnetic components shows surprisingly homogeneous parameter values (a cut through the same sunspot in the perpendicular direction gave similar results, Rüedi et al. 1998). Due to the weakness of the signals produced by the magnetic component with low field strength, we can only trust it to be horizontal to within roughly [FORMULA]. The parameters of the strong magnetic component are more accurate.

[FIGURE] Fig. 1. Parameters obtained from the inversion of NIM spectra. The spectrograph slit, oriented in the heliocentric E-W direction, runs through the centre of the umbra. The x-axis gives the projected position along the slit. The triangles and diamonds respectively represent the strong and weak magnetic field component observed in the Ti [FORMULA] line at 2.2 µm. The "+"-signs represent the values obtained from the Fe I lines at 1.56 µm. Top: the magnetic field strength, B, Middle: the inclination angle to the solar surface normal, [FORMULA] (zenith angle), Bottom: the Doppler shift along the line of sight. The zero velocity level was set by requiring that the more vertical umbral field component is unshifted. More details on the Ti [FORMULA] observations are given by Rüedi et al. (1998).

2.2. Comparison with other observations

Visible lines, as well as the Fe I lines at 1.56 µm, invariably gives a much smoother picture of the magnetic structure of sunspots at similar spatial resolution: a continuous decrease of the field strength from the sunspot centre towards its boundary, with no sudden jumps in the field strength and inclination angle. Also, they do not suggest the presence of 2 distinct magnetic components within the sunspot.

Observations carried out earlier in the day in the same sunspot, with the Fe I lines at 1.56 µm confirm this standard picture (as shown by the "+"-signs in Fig. 1). Hence the strange behaviour exhibited by the Ti [FORMULA] lines is not due to a peculiarity of this sunspot or problems with the observations. We show here that it is rather due to the unique combination of the strong sensitivity of the Ti [FORMULA] lines to cool features and their large Zeeman sensitivity.

A low field strength and a low temperature are both drawbacks to unambiguously detecting an unresolved magnetic component using standard visible lines, and even using the Fe I lines at 1.56 µm. Let us illustrate this by assuming that the penumbra is composed of 2 magnetic components whose properties are taken to be compatible with those derived from the Ti [FORMULA] lines: a hotter plasma with strong magnetic field which is fairly vertical and a cooler component with almost horizontal, weak magnetic field. Each component covers half of the resolution element. They have the following parameters: [FORMULA] K, [FORMULA] G, [FORMULA] and [FORMULA] K, [FORMULA] G, [FORMULA]. The Stokes V profiles due to each component are plotted in Fig. 2 for the Ti [FORMULA] 2.2310 µm and Fe I 1.5648 µm lines (dotted (1) and dashed (2) lines). They have been broadened by a macroturbulent velocity of 2 km s-1. The solid lines denote the composite profiles.

[FIGURE] Fig. 2. Composite Stokes V profiles (solid lines) of the Ti [FORMULA] 2.2310 µm and Fe I 1.5648 µm lines resulting in the presence of two components (dotted and dashed lines), each covering half of the resolution element and having the parameters given in the text.

The Ti [FORMULA] 2.2310 µm line unambiguously shows the presence of 2 components, with the cool low-field-strength component dominating the profile shape. In contrast, the Fe I 1.5648 µm line profile is only slightly affected by this component. Lines in the visible are found to behave in a manner similar to the 1.56 µm lines. Consequently, in order to be detected using visible lines or the 1.56 µm lines, a sufficiently cool, weak-field component needs to cover most of the resolution element. These calculations demonstrate that the anomalous results obtained with the Ti [FORMULA] lines are compatible with the Fe I 1.56 µm lines and with earlier observations.

It is widely accepted that the Evershed effect is concentrated into dark channels of almost horizontal field (e.g. Title et al. 1992, Hofmann et al. 1994). Our Ti [FORMULA] observations show that these channels are indeed cool and that the field-strength in these channels is very low and nearly constant throughout the sunspot penumbra, at least wherever the gas is sufficiently cool to harbour a significant amount of neutral Ti. This suggests that sunspots harbour two quite distinct cool components, with an additional hotter component in the penumbra (revealed by the 1.56 µm lines) having stronger and more vertical magnetic field than the cool penumbral component. Since the spatial resolution of our observations was limited by the moderate seeing conditions, we cannot rule out that the cool horizontal fields seen in Ti [FORMULA] are restricted to the outer penumbra. For the same reason we cannot decide if the more vertical cool component is restricted to the umbra or also extends into the penumbra (cf. Lites et al. 1993).

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: July 26, 1999