The long-term spectral changes of eta Carinae: are they caused by a dissipating occulter as indicated by CMFGEN models?

Abstract

Eta Carinae ($\eta$\,Car) exhibits a unique set of P Cygni profiles with both broad and narrow components. Over many decades, the spectrum has changed -- there has been an increase in observed continuum fluxes and a decrease in FeII and HI emission line equivalent widths. The spectrum is evolving towards that of a P Cygni star such as P~Cygni itself and HDE~316285. The spectral evolution has been attributed to intrinsic variations such as a decrease in the mass-loss rate of the primary star or differential evolution in a latitudinal-dependent stellar wind. However intrinsic wind changes conflict with three observational results: the steady long-term bolometric luminosity; the repeating X-ray light curve over the binary period; and the constancy of the dust-scattered spectrum from the Homunculus. We extend previous work that showed a secular strengthening of P~Cygni absorptions by adding more orbital cycles to overcome temporary instabilities and by examining more atomic transitions. {\sc cmfgen} modeling of the primary wind shows that a time-decreasing mass-loss rate is not the best explanation for the observations. However, models with a `small' dissipating absorber in our line-of-site can explain both the increase in brightness and changes in the emission and P Cygni absorption profiles. If the spectral evolution is caused by the dissipating circumstellar medium, and not by intrinsic changes in the binary, the dynamical timescale to recover from the Great Eruption is much less than a century, different from previous suggestions.