Spectropolarimetric analysis of an active region filament. II. Evidence of the limitations of a single-component model

Aims. Our aim is to demonstrate the limitations of using a single-component model to study the magnetic field of an active region filament. To do this, we analyzed the polarimetric signals of the He I 10830 angstrom multiplet, which were acquired with the Infrared spectrograph GRIS of the GREGOR telescope (Tenerife, Spain). Methods. After a first analysis of the general properties of the filament using HAZEL under the assumption of a single-component model atmosphere, in this second part we focus our attention on the observed Stokes profiles and the signatures that cannot be explained with this model. Results. We have found an optically thick filament whose blue and red components have the same sign in the linear polarization as an indication of radiative transfer effects. Moreover, the circular polarization signals inside the filament show strong magnetic field gradients. We also show that even a filament with such high absorption still shows signatures of the circular polarization that is generated by the magnetic field below the filament. The reason is that the absorption of the spectral line decays very quickly toward the wings, just where the circular polarization has a larger amplitude. In order to separate the two contributions, we explore the possibility of a two-component model, but the Inference becomes impossible to overcome because very many solutions are compatible with the observations.