2. 2-D models and spectral observations
The 2-D models used here and the limitations of a 2-D approach were described by Gadun et al. (1999) in detail. The models have a computational box with sizes 3360 km1960 km in horizontal and vertical directions (the atmospheric layers occupy about 700 km). The spatial step is 28 km.
2-D models normally produce larger model oscillations than 3-D simulations. Therefore in these simulations special boundary conditions at the top and at the bottom of the computational domain were chosen such that the oscillations were kept as weak as possible. These conditions were described by Gadun et al. (1999). The total evolution time of the simulated granulation which was taken into consideration in our analysis is about . The time step between the studied models is 30 s.
The spectral observations were described in detail by Kucera et al. (1995) and Hanslmeier et al. (2000). They were carried out with the Vacuum Tower Telescope (Observatorio del Teide, Tenerife) using its echelle spectrograph. Five Fe I spectral lines have been observed strictly simultaneously with exposure time of 0.3 s: 5434.543 Å, 6494.994 Å, 5576.508 Å, 6301.508 Å, and 6302.499 Å. Their main characteristics are given in Kucera et al. (1995). The slit width was 0.28" and the subsequent spectral profiles are separated by to 0.17" in the spatial direction. The dispersions of the spectra were between 3.3 and 4.0 mÅ/px. Only the most quiet region covering 30arcsec on the Sun was used for the computation of the continuum intensity and the spectral line characteristics (the residual line core intensity and the Doppler shifts). This region was without supergranular network activity and was situated close to the disk center ().
© European Southern Observatory (ESO) 2000
Online publication: December 5, 2000