Keele Research Repository

We present an analysis of the 2013 Suzaku campaign on the nearby luminous quasar PDS 456, covering a total duration of ~1 Ms and a net exposure of 455 ks. During these observations, the X-ray flux was suppressed by a factor of >10 in the soft X-ray band when compared to other epochs. We investigated the broadband continuum by constructing a spectral energy distribution, making use of the optical/UV photometry and hard X-ray spectra from the later XMM-Newton/NuSTAR campaign in 2014. The high energy part of this low flux state cannot be accounted for by self-consistent accretion disc and corona models without attenuation by absorbing gas, which partially covers a substantial fraction of the line of sight towards the X-ray source. Two absorption layers are required, of column density $\log (N_{\rm{H,low}}/{\rm cm^{-2}})=22.3\pm0.1$ and $\log (N_{\rm{H,high}}/{\rm cm^{-2}})=23.2\pm0.1$, with average covering factors of ~80% (with typical 5% variations) and 60% ($\pm$10-15%), respectively. In these observations PDS 456 displays significant short term X-ray spectral variability, on timescales of ~100 ks, which can be accounted for by variable covering of the absorbing gas. The partial covering absorber prefers an outflow velocity of $v_{\rm pc} = 0.25^{+0.01}_{-0.05}c$ at the >99.9% confidence level over the case where $v_{\rm pc}=0$. This is consistent with the velocity of the highly ionised outflow responsible for the blueshifted iron K absorption profile. We therefore suggest that the partial covering clouds could be the denser, or clumpy part of an inhomogeneous accretion disc wind. Finally we estimate the size-scale of the X-ray source from its variability. The radial extent of the X-ray emitter is found to be of the order ~15-20 $R_{\rm g}$, although the hard X-ray (>2 keV) emission may originate from a more compact or patchy corona of hot electrons, which is ~6-8 $R_{\rm g}$ in size.