## 6. Discussion and conclusionsWe have modelled the facular, quiet-sun and spot intensities of the Sun using Kurucz' ATLAS9 spectral synthesis program. The model for the facular atmosphere is a modified version of model P of Fontenla et al. (1993). The model atmospheres for the quiet Sun and the spots were standard Kurucz radiative equilibrium model atmospheres. Using this much more realistic approach, we confirm our previous conjecture (Papers I and II) that a 3-component model can adequately fit the spectral irradiance variations in the UV, in the visible as measured by VIRGO and the total irradiance variations as measured by ACRIM. We also check how well our facular model performs when compared to measurements resolving the solar disc. Our calculated facular contrasts as a function of limb angle fit into the general picture of the measurements, though we tend to find larger contrasts than indicated by most of the measurements. However, the scatter between measurements by different groups is large and the absolute contrast values are highly dependent on the resolution of the measurements, the magnetic filling factor of the observed region, and on the technique employed to derive the contrasts (see Solanki 1994for a more detailed discussion). We find a steeper increase of the contrast very close to the limb, compared to a number of measurements, mainly due to the assumption of a plane-parallel atmosphere. The exact contrast behaviour very close to the limb, however, is not going to be crucially important when modelling the disk-integrated irradiance variations as the flux contribution beyond is very small. One of the main new features of the modelling described in this paper is that we can reconstruct the spectral dependence of the facular contrast. The calculated colour dependence of the contrast at various limb angles matches the available data very well, suggesting that our approach is not only adequate in a global sense (i.e. in reproducing spectral irradiance variations), but also locally. The inclusion of the influence of spectral lines plays an important role in reproducing these data. The other significant new result of this paper is that we estimate the influence of line blanketing on irradiance variations in different wavelength ranges. Our model reproduces the observations of Mitchell & Livingston (1991), who estimate the change in line blanketing over the solar cycle from observed spectra, relatively well. Finally, we find that the line blanketing is the major contributor to the total irradiance variations on solar-cycle time scales. The reason is that the sunspot and facular contributions to continuum variations cancel each other out to a large degree, whereas the line-blanketing is changed by sunspots by a much smaller amount than by faculae. © European Southern Observatory (ESO) 1999 Online publication: April 19, 1999 |