Reading Report of PDS 456 Soft X-ray Observation

Main Idea

This paper (Reeves et al. 2020) discussed new XMM-Newton observations in 2018 and 2019 of the luminous quasar PDS 456. The new observation shows significant soft X-ray absorption features along with Fe K absorption lines that already observed in previous works. The lines in the new observation could be well explained by absorption from ultra-fast outflow. The separate spectral fitting of two components, soft X-ray absorber and iron K absorber, yield same outflow velocity of \(\sim -0.25c\), consistent with previous studies of this source. The soft X-ray component shows strong variability while the iron K absorber is relative stable within the \(\sim 20\) day period of the two 2019 observations, indicating the two absorption features originate from diferent distances. The author gave theoretical interpretations of the observed features. This work is observational paper which could be rated as good. The conclusion of ultra fast outflow is strong, and calls for further investigation of this source with more advanced instruments like XRISM and ATHENA.

Motivation

PDS 456 is a luminous quasar with redshift \(z = 0.184\). Previous observations reveals the existence of iron K-shell absorption line profile with outflow velocity of \(\sim 0.25c\). It has been claimed that high-velocity iron K winds might be common in nearby AGNs due to numerous observations (e.g. APM 08279+5255, PG 1211+143). However, the detection of blue-shifted absorption lines in soft X-ray band is relatively rare. For the source PDS 456, soft X-ray outflow with lower ionization has been confirmed with the archival XMM-Newton observations. Soft X-ray fast winds has been reported in a few other sources, such as IRAS 13224-3809 and 1H 0707-495. There were also examples of co-existing soft X-ray and Fe K winds with consistent velocity, e.g. PG 1211+143 with \(v \sim -0.06c\) and IRAS 13224-3809 with \(v \sim 0.24c\). Despite the increasing observations, the correlation between iron K wind and soft X-ray wind was still poorly understood. This paper is motivated by three XMM-Newton observations, one in 2018 and two in 2019. The latter two observations were carried out with an interval \(\sim 20\) days. This provides a chance to examine the variability timescale of the outflow wind, providing constraints on the location, size and density. Furthermore, these new observations have resolved both the soft X-ray component and the iron K absorber, enabling detailed investigation of the exact relation between these two components.

Data and Methods

The 2018 XMM-Newton observation was made on Sep.9, with 86 ks exposure, along with HST and NuSTAR observations. Observation 2019a was on Sep. 2, with 83 ks exposure, simultaneous with HST observation. 2019b was made on Sep 24., about 20 days after 2019a. This observation has 94 ks exposure and conducted simultaneously with NuSTAR. The 2018 observation is in the bright state, with energy flux \(F= 9.4\times 10^{-12}\) ergs cm, \(^{-2}s^{-1}\) at 2-10 keV. The soft X-ray spectrum of 2018 observation is featureless, while the hard X-ray spectrum extends to \(> 50\)keV according to NuSTAR observation. Being about 5 times fainter than 2018 observation, both 2019 observations demonstrate Fe K absorption profile at 7-10 keV and soft X-ray absorption features, while 2019b is more significant. Interestingly, Swift monitoring shows that 2019a is almost at the state with minimum flux.

The author mainly fitted the 2019b spectrum to investigate the properties of soft X-ray absorber. They first used gaussian-like profile to locate the three most significant absorption lines, then fitted it with blue-shifted photoionization model (XSTAR) with solar abundance. In addition to the fast outflow, they also tried fitting with models of low velocity shift, which was badly fitted and thus ruled out the latter assumption. In the broad-band fitting, the author also used XSTAR to identify the iron K absorption features – the 9 keV trough and the 10.2 keV trough. The former one is probably due too Fe XXVI Ly\(\alpha\) absorption line, with velocity shift consistent with the soft X-ray component; the latter requires a second faster wind component, with outflow velocity \(v \sim 0.36c\).

Results and Conclusion

Soft X-ray Absorber Spectral Fitting

At soft X-ray band, the fit with a power-law model with galactic absorption is poor, and there are absorption profiles in addition to the continuum. With three Gaussian profile, the fit was significantly improved. With the a priori knowledge that outflow velocity might be \(-0.25-0.3c\), the Gaussian absorption lines could be identified as OIII Ly\(\alpha\) (18.97 \(\AA\) ), Ne IX He\(\alpha\) (13.45 \(\AA\) ) and Ne X Ly\(\alpha\) (12.13 \(\AA\) ). All the three lines give a common velocity shift of \(v/c = -0.257\), in agreement with the previous studies. The fitting using photoionization model XSTAR further improves the fitting, with the expense of 3 extra free parameters. The fitting yield a ionization parameter \(\log\xi = 3.4\), column density \(N_H = 2.3\times 10^21\)cm\(^{-3}\), and a covering factor \(f_{\mathrm{cov}} = 0.92^{+0.08}_{-0.32}\). In order to confirm the outflow explanation, the author further fitted the soft X-ray spectrum with absorber of zero or small velocity shift. The EW upperlimit of lines, constrained by the non-detection of the accompanying lines, is significantly weaker than observation, thus the zero-blueshifit scenario could be ruled out.

Fe K Line Fitting

The Fe K absorption line at 9 keV restframe could be identified with Fe XXVI Ly\(\alpha\), with velocity shift \(-0.25\pm 0.02c\), in perfect agreement with the soft X-ray absorber. The absorption trough at \(\sim 10.2\) keV cannot be explained by the same outflow, thus might calling for an additional absorption component with \(v= 0.36c\). Compared with soft X-ray wind, the Fe K absorber exhibit higher ionization parameter \(\log \xi = 5.0\), almost two orders of manitude larger than the soft X-ray wind. Its column density is also larger (\(N_H = 7\times 10^{23}\) cm\(^{-3}\)).

Interpretation

The 2018 observation has a 2-10 keV flux 4 times stronger than 2019b. By fitting the 2018a observation with same photoionization model, ionization parameter \(\log \xi\) is 4 times stronger than 2019b, which corresponds to the flux variability. Meanwhile, the 2019a observation has lower ionization factor due to the low flux, and the best-fit column density is higher. By another fit with neutral partial covering absorber, the zero-blueshift scenario was ruled out.

In this work, the author gave self-consistent estimations of the location and size of the soft X-ray absorber with observations 2019a and 2019b. Both observations result in \(\sim\) parsec distance from central BH estimated by the ionization luminosity and covering fraction. Interestingly, the assumption of 20-day variation timescale gave an estimate of absorber size \(\Delta R \sim 10^{15}-10^{16}\)cm, also agrees with the covering fraction. The soft X-ray absorber is believed to locate outside the high-ionization Fe K absorber, and possibly originate from the ultra-fast outflow. The mass outflow rate for a uniform soft X-ray wind exceeds the upper limit set by accretion rate, a strong implication of its clumpy nature, explaining the fast variability of column density and ionization parameter.

Scientific Significance

This work contributes to the understanding of the correlation between absorbers in soft X-ray band and iron K band. With spectral fitting, the author built a clear physical picture of relative location of the two components, and gave interpretation of the observed features during three different times.

This work not only enhances the understanding of the source PDS 456, but also have profound impact on larger topics. The discovery of the clumpy absorber provides observational evidence for the hypothesis that X-ray wind originate from the post-shock interstellar medium interacting with the ultra-fast outflow. The estimated outflow rate and filling fraction also hints on the possible role in energy feedback to the galaxy that the soft X-ray wind might play.

Key Factor Leading to Conclusion

The observations in different state of PDS 456 made it possible to estimate the properties and locations of the absorption components. The authors have published several papers discussing about PDS 456. Along with the three observations mainly introduced in this paper, previous multi-wavelength observations of this source, and other objects exhibiting similar absorption features are also crucial in getting the conclusion.

Possibilities of Improvement

The size \(\Delta R\) is estimated based on the covering fraction \(f_{\mathrm{cov}} > 0.6\) for 2019b, that the soft X-ray wind has comparable size with corona. However, I doubt whether this transverse scale could be directly applied in estimation of the number density along the line of sight. If the outflow is intermittent, line of sight depth may vary and dependent on the morphology of the wind. Though a spherical approximation suffice for a rough estimate, it could be improved regarding the light curve monitored by Swift.

Besides, PDS 456 is in the PV source list of XRISM. With unprecedented spectral resolution and sensitivity, there is no doubt that the new observation will provide evidences supporting or against the conclusions in this work.