Mapping Quasar Light Echos in 3D with Lyα Forest Tomography
My studies of the HeII transverse proximity effect have led to interesting new results about the emission history and emission geometry of quasars, however in the end, I was left with more open questions than answers and clearly needed more objects to get a clearer picture about quasar properties. Unfortunately, it is difficult to substantially expand the current sample and investigate the HeII transverse proximity effect of a lot more quasars. The main reason is that HeII transparent sightlines are particularly rare. As soon as there is too much neutral hydrogen in between the background quasar and Earth, the HeII Lyα forest becomes unobservable. Therefore, there are not very many additional targets in the sky that could be observed with the (aging) HST/COS spectrograph. The next generation of space-based Far UV telescopes is currently just on the drawing board and it will certainly take over a decade before these instruments become available. The solution therefore is to focus on the hydrogen proximity effect. The HI proximity effect is much weaker then the HeII counterpart, mostly since at intermediate redshifts (z~3) hydrogen in the intergalactic medium is already highly ionized and the mean transmission in the HI Lyα forest high. A foreground quasar can therefore enhance the transmission only by a relatively low amount. On the other hand, the HI Lyα forest can be observed with powerful optical telescopes from the ground, which compensates for the relative weakness of the HI proximity effect.
The technique I will employ for this project is called 'Lyα forest Tomography' and was pioneered by the CLAMATO collaboration (Lee et al. 2017). The basic idea is to take spectra of a large number of background sources in a relatively small field. The high density of background sources (~1000 deg2) allows to interpolate between sightlines and to derive a three-dimensional map of the intergalactic medium on Megaparsec scales. To reach a sufficient background sightline density, one can not rely on quasars as background sources alone but has to include faint galaxies as well. Still, one has to push current-generation muti-object spectrographs to their limits and acquire spectra of extremely faint sources (e.g. r=24.5 mag). However, the information gained by this approach are worth the effort and offer a completely new understanding of the three dimensional structure of the intergalactic medium.
The primary application by Lee et al. (2017) is to map the cosmic density structure and find e.g protoclusters or cosmic voids. However, the method is equally well capable to map the light echos of luminous quasars. This allows to map for individual quasars the full three-dimensional structure of the proximity region and gain detailed insights into their emission history (lifetime, age, flickering, ..) and emission geometry (orientation, obscuration, beaming).
These information will allow to test e.g. AGN unification models and might shed light on AGN triggering and activity cycles as well as the processes by which supermassive Black holes assemble their mass. Since quasars are such powerful energy sources, these properties are also highly relevant in a broader context beyond AGN physics, in particular galaxy formation (feedback processes) and the the ionization state of the CGM and IGM (topology of Helium reionization and metal absorbers).
The figure above shows a simulation of the HI proximity effect. I have assumed a hyperluminous quasar at redshift z=3.2 that shines for 25 Myr and illuminates half of the sky. The transmission enhancement in the IGM is clearly visible, as well as its bi-conical shape. The finite age of the quasar limits the extent of the proximity zone to higher redshifts (right) and confines it to a parabolic shaped region which expands with time. Clicking on the figure leads to a video that visualizes the time evolution of the proximity region. The lower panel depicts a computed Lyα forest spectrum along the dashed sightline. The IGM transmission is clearly enhanced in regions that are illuminated by the foreground quasar. However, the stochastic IGM absorption will make it difficult to reconstruct which parts of the sightline are illuminated and observing only one background sightline will certainly not give enough information to reconstruct quasar age, obscuration and orientation. The general idea of Lyα forest tomography is however to have many sightlines that probe the proximity region, e.g. 30 sightlines with an average separation of 4 Mpc, (red dotted lines). Combining the information of all of them in a statistical manner will allow to determine the properties of the proximity region and to derive a map of the quasar light echo.
This project requires, ahead of the spectroscopic observation, dedicated pre-imaging for selection of the background sightline targets. It is therefore imminent to choose the right observing strategy and in particular the optimal redshift of the targeted foreground quasar. At high redshift, the average IGM transmission is lower and the proximity effect therefore stronger but on the other hand the number of available background galaxies drops dramatically. It is therefore important to find the best compromise between good sampling of the proximity region by background sightlines and strength of the transmission enhancement.
To find the optimal strategy, I currently work on models of the HI proximity effect. These are, very similar to my previous work on the HeII proximity effect, based on outputs of large cosmological hydrodynamical simulations (Nyx, L100_N4096, Almgren et al. 2013, Lukić et al. 2015) which I postprocess with a photoionization model of the the foreground quasar. In addition, I develop the necessary statistical tools to analyze the data and constrain quasar parameters in a fully Bayesian way. This is a challenging task, given the complexity of the problem and the size of the parameter space. It will require a large number of numerical simulations that can then be compared to the data. However, substantial progress has already been made and the optimal survey strategy in terms of foreground quasar redshift, required spectral resolution, sightline density, S/N and so on will soon be determined. It is then time to actually collect data.