SIMBAD references

We present a detailed analysis of the complex absorption apparent in the 2-6 keV X-ray spectrum of the bright nearby Seyfert galaxy NGC 4151. We first utilize the large bandpass and medium spectral resolution afforded by BeppoSAX data to construct a 1-100 keV spectral template, which assumes the absorption arises in both warm (i.e. partially photoionized) and cold gas present in the line of sight to the active nucleus of the source. Application of this spectral model to an ASCA `long-look' observation of NGC 4151 reveals a partial correlation between the underlying continuum flux and the ionization state of the warm absorber. Such a correlation is an intrinsic property of a warm absorber and argues strongly in favour of this interpretation for the complex absorbing column over alternative partial covering models.

The warm absorber in NGC 4151 has a column density of ∼2x10²³ cm ^–2 , with an ionization parameter in the range log(ξ) ~ 2.4-2.7. The inferred relatively low density (∼10⁵ cm^–3 ) for the warm gas implies an equilibration time-scale for the dominant ions of the same order as or longer than the time-scale of the continuum variability. It follows that the warm component will invariably be observed in a non-equilibrium ionization state. The warm absorber in NGC 4151 may be identified as a multi-temperature wind produced by evaporation from the inner edge of an obscuring torus as discussed in a recent paper by Krolik & Kriss. The unusually complex character of the absorption seen in NGC 4151 may then be explained in terms of a fortuitous line of sight which grazes the top edge of the obscuring torus so as to intercept a substantial column of both the warm and cold gas.

We also find that (i) the reported hardening of the spectrum of NGC 4151 as the continuum level falls may be simply due to the presence of an underlying (hard and relatively constant) Compton-reflection component, and (ii) the iron Kα line has a relatively narrow Gaussian profile and a line flux that remains constant over both short (days) and long (months to years) time-scales-a relativistically broadened iron Kα feature was not required in our modelling.