Astron. Astrophys. 363, 289-294 (2000)

1. Introduction

The solar photosphere can be treated as a transition region between thermal convection, which predominates below the surface and in the low photosphere (overshooting convection) and different kinds of oscillations, which completely define the gas dynamics in the upper layers.

In this context, it was studied in a large number of papers - for instance, among others, by Leighton et al. (1962), Evans (1964), Krat (1973), Canfield & Mehltretter (1973), Keil & Canfield (1978), Kneer et al. (1980), Durrant & Nesis (1981, 1982), Nesis et al. (1988), Komm et al. (1990, 1991a, 1991b), Karpinsky (1990), Hanslmeier et al. (1990, 1994), Balthasar et al. (1990), Kucera et al. (1995). Since the correlation analysis is the simplest method to extract information concerning the global behaviour of physical quantities in the photospheric medium, it was widely used in most these studies (e.g. Karpinsky 1990 and references therein).

In the paper of Espagnet et al. (1995) the height variation of the solar granulation was investigated using a 16-min time series of two- dimensional (2-D) multichannel subtractive double pass spectrograms in the 5690 line obtained at the Pic du Midi Observatory. Krieg et al. (1999) used Fabry-Perot interferometer data obtained at the Vacuum Tower Telescope at the Observatorio del Teide to analyze spectral scans of narrow-band images across the line. They found that granular intensity fluctuations occur in the deep photosphere only and disappear at heights of about 100 km. Contrary to Krieg et al., Espagnet et al. found an intensity pattern visible in the upper photosphere which is of turbulent origin.

On the other hand, multidimensional hydrodynamic (HD) simulations of solar granulation have reached a high level of realism. They reproduce a significant number of observables (Stein & Nordlund 1998, Gadun et al. 1999, Asplund et al. 2000, Georgobiani et al. 2000, Ploner et al. 2000) but they were never involved for a detailed investigation of the photospheric structure. Exceptions are the papers of Gadun et al. (1997 and 1999), who reproduced correlations between line parameters within 2-D model atmospheres.

In this paper we want to employ both the correlation analysis and 2-D model atmospheres to study the photospheric structure. We shall deal with three groups of correlations.

Into the first group we include the model (or theoretical) correlations. They describe correlations between selected quantities of 2-D model atmospheres. These correlations are derived directly from time-dependent 2-D models.

To test spectral observations for diagnostic purpose is another aim of this paper. For this reason, we have made an analysis of spectrograms obtained from 2-D models. They have been also studied in the context of correlation analysis. These correlations between simulated line parameters form the second group of correlative relationships, presented here. We shall call them the simulated correlations.

Finally, we test our simulations by comparing the model and simulated correlations with those provided by real spectral observations. We use the spectral observations made with high spatial and spectral resolution. These correlative dependencies will be called the observed correlations in the following. The third group of our correlations contains these quantities.

We show that the result of this comparison as well as the application of spectral line diagnostics are sensitive to the method providing an estimation of line formation depths and briefly analyse two criteria to calculate them.

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000
helpdesk.link@springer.de