The near- and mid-infrared continuum emission of Seyfert nuclei: Constraints on the models of obscuring tori

For an extended sample of Seyfert galaxies we compile from the literature the infrared fluxes in the four IRAS bands, the ground-based small-beam (similar to 5 "-10 ") fluxes in the standard Q, N, M, L (or L') bands, and the nuclear (nonstellar) estimated fluxes in the JHK bands. We estimate nuclear fluxes in the L band by applying a correction for stellar light. From the statistical study of the infrared colors and luminosities, we derive the typical spectral energy distributions (SEDs) of Seyfert 1 and 2 nuclei and the typical differences in luminosities between the two types of objects in the mid-and near-infrared spectral ranges. The analyses of colors and luminosities agree with the fact that in general Seyfert 2 nuclei become increasingly fainter with respect to Seyfert 1 nuclei as we go toward shorter infrared wavelengths. This behavior is consistent with growing anisotropy of a dusty torus emission toward shorter wavelengths, but the degree of anisotropy is low (the radiation appears to be substantially isotropic at lambda greater than or similar to 25 mu m). For Seyfert 2 galaxies having Compton-thin obscuring structures at hard X-ray energies, we find correlations between the absorbing hydrogen columns and some infrared colors and luminosities. The observational data appear to severely challenge many models of dusty tori, which hardly account for the shapes of the SEDs and the degree of anisotropy observed in Seyfert galaxies. In particular, at variance with some earlier claims, very thick and compact tori are basically inconsistent with these observational constraints. The most successful models, though having problems in accounting for several details, can fit the major infrared observational data of both Seyfert 1 and Seyfert 2 nuclei with tori that extend up to several hundred parsecs and have fairly low optical thickness.