Key points are not available for this paper at this time.
We present ROSAT observations of NGC 4051 from 1991 November 15 (28-ks exposure) and 1991 December 12 (4.7-ks exposure). During these observations the source spectrum is well described empirically by a simple power law with energy index ∼ 1.5 and an absorption edge with optical depth ∼ 0.8 at ∼ 0.85 keV or, more physically, by a warm absorber with an underlying power law of energy index 1.19, similar to that seen at higher energies, an ionization parameter ξ ∼ 300 and a column of warm gas Nwarm ∼ 8 × 1022 cm−2. During the November observation the source flux varies by a factor of ∼ 10, and there is some evidence that ξ may increase with increasing flux. However, the correspondence between luminosity and ξ is not simple, and there is a suggestion that ξ may lag behind luminosity by a few thousand seconds. If the lag represents a cooling time-scale for the warm absorber, we can crudely estimate the gas density as ∼ 107 cm−3, and hence a thickness of ∼ 5 × 1015 cm for the warm gas cloud. The December observation is not well fitted by a simple warm absorber, and the fit is much improved by the addition of an emission line at 0.63±0.07 keV, i.e. at the energy of O VIII Ly α (0.65 keV). We note that Nwarm as deduced from the November observation is noticeably less than the value deduced by Pounds et al. (∼ 3.7 × 1023 cm−2) from observations one year prior to ours, and there is an indication that the value deduced from the December observation is also different from that of the November observation, demonstrating the presence of large-scale inhomogeneities and motions in the warm gas in the nuclei of active galaxies.
McHardy et al. (1995) studied this question.