Evidence for a precessing accretion disk in the nucleus of NGC 1097

We present new spectroscopic observations of the LINER (and now Seyfert 1) nucleus of NGC 1097, and discuss the evolution of its broad, double-peaked Balmer lines. When originally discovered in 1991, the red peak of the double-peaked Hα line was stronger than the blue, while by 1994 the Hα profile had become almost symmetric and the integrated line flux had decreased to half its original value. Our new spectrum, taken in 1996, shows that the broad, double-peaked lines have returned to almost their original strengths, the profiles of Hβ and Hα are identical to within errors, and the broad-line emitting region is unreddened. However, the profile of the Balmer lines is now such that the blue peak is stronger than the red, opposite to the asymmetry observed in 1991. Various models are considered for the observed behavior, all assuming that the emission lines originate in an accretion disk. We present a refined version of the precessing, planar, elliptical accretion-ring model proposed by Storchi-Bergmann et al. and Eracleous et al. This model provides an acceptable fit to the line profiles. We also consider the possibility that the line-profile evolution results from a precessing warp in the disk, induced by irradiation from the center, and show that the range of radii and precession timescales expected in this model are consistent with the observations. The sudden appearance of the "disklike" broad-line profiles in NGC 1097 could have resulted from the formation of a new accretion disk due, for example, to the tidal disruption of a star or the illumination of a preexisting disk by a transient ionizing source at the center of the disk.