The Helium II Transverse Proximity Effect
Reionization of helium is the final phase transition of the Intergalactic Medium (IGM). While hydrogen was reionized by stars at redshift z > 6, HeII reionization requires UV photons with energies >54.4eV which could only be supplied in sufficient quantities by quasars at much later cosmic times. HeII reionization therefore finished only around z=2.7.
The gas in the IGM is usually too thin to be observed in emission but it can be studied in detail via absorption spectroscopy, in particular along sightlines towards bright quasars. The so called Lyman-α forest – the ensemble of absorption lines from many absorbers – encodes extensive information about the density and ionization structure of the IGM along the line of sight.
Quasars are powerful sources of ionizing photons and can cause a substantial impact on the ionization state of the intergalactic medium on Megaparsec scales. The sketch above illustrates the showcase example of the so called HeII transverse proximity effect along the Q0302-003 sightline (Heap et al. 2000, Jakobsen et al 2003). The HST/STIS Far UV spectrum of the background quasar displays saturated Gunn-Peterson absorption everywhere, except close to the background quasar and close to a foreground quasar located at z=3.05 and separated by just 6 arcminutes from the background sightline. The interpretation in both cases is that the intense UV radiation from the nearby quasar causes enhanced IGM ionization and therefore relatively high HeII Lyα forest transmission in these regions, leading to the line of sight and transverse proximity effects..
The transverse proximity effect is an ideal constellation to study the emission properties of the foreground quasar. Since photons from the foreground quasar require time to reach the background sightline, the IGM absorption along the background sightline probes the quasar emission at some point in the past. For the case shown above, this time delay is about 10 Myr. Since we observe the quasar in an active phase along the direct sight and the transmission spike at the background sightline, the quasar has to shine for at least 10 Myr. The transverse proximity effect can as well constrain the emission geometry of quasars. It is assumed that quasars are obscured towards some directions. However, the transmission spike along the background sightline is a clear indication that this quasar does indeed emit towards Earth and towards the background sightline.
In our group, we have expanded the sample of helium quasars with science-grade HST/COS FUV spectra to 25 sightlines and re-reduced the archival data in a homogeneous way (Worseck et al. 2016). This allows for the first time a statistical study of HeII reionization. To make proper use of the sample of FUV spectra, I was in charge of a dedicated optical foreground quasar survey around all helium sightlines. This survey was split up in a wide survey, run on 3.5m class telescopes to covers large separations from the background quasar and a deep survey conducted on 8m class telescopes to reach quasars as faint as r < 24 mag. In total, I discovered 130 new quasars of which 20 should have a substantial impact on the HeII background sightlines. The figure above shows the HST/COS Far UV spectrum of the background quasar SDSS J1253+6817 together with foreground quasars discovered within my survey (orange) and additional objects from the SDSS survey (Schmidt et al. 2017).
By stacking the FUV spectra on the positions of the foreground quasars I derived average transmission profile in the vicinity of the foreground quasars. In these stacks, the Transverse Proximity Effect is clearly visible which proves that this effect is not limited to the prototype object but indeed a common phenomenon. In addition, I could constrain the quasar lifetime. For this, I included only objects above a certain separation from the sightline into the stacks. The enhanced HeII transmission in the vicinity of the foreground quasars persisted in the stack, meaning these quasars were old enough that their radiation could reach the background sightlines. This sets a statistical and purely geometrical lower limit on the quasar lifetime of 25 Myr (Schmidt et al. 2017).
However, among the four foreground quasars with the highest HeII photoionization rates, three did not show any impact on the background sightline, letting us speculate that these might be either very young or highly obscured. I therefore developed a detailed numerical model of the HeII proximity effect that incorporates light travel time effects, finite quasar ages and quasar obscuration (Schmidt et al. 2018). It is based on outputs from one of the largest cosmological hydrodynamical simulations (Nyx, L100_N4096, Almgren et al. 2013, Lukić et al. 2015) and post-processed with a semi-numerical, fluctuating HeII UV background (Davies et al 2017) to which the radiation of single, isolated foreground quasars with variable quasar age and obscuration was added. An example of this is shown in the above figure. The quasar has similar luminosity to the foreground quasar in the Q0302-003 prototype case and the background sightline is placed at a similar separation. A strong impact on the HeII Lyα transmission in the IGM (top panel) and a clear signature in the HeII spectrum along the background sightline (bottom panel) are visible. I have assumed that the quasar illuminates only half of the sky and is obscured towards other directions, leading to the bi-conical shape. The exact shape and structure of the proximity region therefore depends on the orientation of the quasar, the amount of obscuration and its age. Clicking on the plot leads to a video that illustrates the expansion of the quasar proximity region with time, but for simplicity does not include obscuration effects.
Using purpose developed Bayesian statistics, I have compared the synthetic HeII transmission spectra computed from the models to our observations and inferred joint constraints on the age and obscuration of six foreground quasars (Schmidt et al. 2018). The results indicates a bimodality of the quasar emission properties: One quasar is old (22 Myr) and nearly unobscured (<35%) while three others are either young (<10 Myr) or highly obscured (>70%). This apparent bimodality is so far not understood and requires further investigation and in particular a larger sample.