## 1. IntroductionThe extreme ultraviolet (EUV) emission of the quiet Sun forms a pattern of brighter and darker structures. The bright parts (referred to as the chromospheric or transition region network) surround darker cells with a diameter of approximately 30000 km. The brightness enhancements in the network result from concentrations of magnetic field, which in turn are produced by convection in the photospheric supergranules (e.g. Leighton 1959; Martin 1988). Several models describe the extension of the magnetic field from small areas between the cells of the photospheric supergranules into broad structures in the corona (Gabriel 1976; Athay 1982; Dowdy et al. 1986; Solanki & Steiner 1990). Intensity distributions of the quiet Sun have been investigated in older studies (Huber et al. 1974; Reeves 1976; Reeves et al. 1976) based on Skylab data and more recently using data from the Solar Ultraviolet Measurements of Emitted Radiation (SUMER) instrument (Lemaire et al. 1997; Bocchialini et al. 1997; Griffiths et al. 1999; Dammasch et al. 1999) and the Coronal Diagnostic Spectrometer (CDS) (Gallagher et al. 1998). All studies show that the quiet Sun radiance follows a distribution with a pronounced peak and a long tail extending to high intensities. The skewed form of the distribution precludes a fit with a single Gaussian. Reeves (1976) decomposed the radiance distribution obtained with Skylab data into a sum of two Gaussian distributions, which he interpreted as describing the network and the cell radiance. Gallagher et al. (1998) carried out a similar decomposition of CDS data. Griffiths et al. (1999) investigated SUMER observations of the quiet Sun transition region and found that the emission line intensities are consistent with a lognormal distribution. Which of these approaches - double Gaussian or lognormal - provides a better description can only be decided on the basis of a large amount of data, i.e. distributions of extremely high statistical significance. Distinguishing between a one-component (lognormal) and a two-component (double Gaussian) description is of potential interest, since it gives a hint on whether chromospheric and coronal heating distinguishes between network and cell interiors or not. In this paper we contribute to the statistical description of the quiet Sun in several ways. Firstly, by discussing and testing various possible models for the frequency distribution of the radiances. Secondly, we compare distributions obtained with two different instruments, SUMER and CDS onboard SOHO (Solar and Heliospheric Observatory). Thirdly, our statistics are based on more than two years (March 1996 to June 1998) of almost monthly observations and thus should give a representative sample of quiet Sun radiances. The extremely good statistics needed in order to distinguish between different functional representations of the distributions are thus achieved. Finally, we consider normalized distributions, which allows us to follow the change in the shape of the distribution as a function of formation temperature of the emitting ions. In the following (Sect. 2) we briefly describe the CDS and SUMER observational data. Sect. 3 gives the frequency distributions in the various wavelengths and their fits. In Sect. 4 we search for a relation between temperature and the parameters of the distribution function, and in Sect. 5 the results are summarized. © European Southern Observatory (ESO) 2000 Online publication: October 24, 2000 |