1998A&A...331...11D

Our purpose is to build a photoionization model for the Broad Line Region (BLR) of NGC 5548, based on observations performed during the first period of the AGN Watch series and on a compilation of EUVE, ROSAT, GINGA and OSSE observations. This provides us the line fluxes, their variations and those of the underlying continuum, as well as the overall spectral distribution. From these properties alone, without any assumption on the geometry of the BLR, we put into evidence three inconsistencies: an ``energy budget problem'', a ``line ratio problem'', and a ``line variation problem''. We discuss different possibilities to resolve these problems (a peculiar geometry, a large reddening, an incorrect estimate of the continuum flux irradiating the BLR, etc...). The most likely explanation is that, either a remote but still photoionized region, or a non radiatively heated region, contributes to the broad line spectrum. We favor the second alternative. Furthermore, a very dense and hot medium, for instance the atmosphere of the accretion disk, could account for the observed HeII 4686/1640 line ratio.