QSO clustering with the SDSS-IV eBOSS survey

1 Jun 2016, 16:50
20m
Room I

Room I

Speaker

Pauline Zarrouk (CEA)

Summary

The eBOSS multi-object spectrograph has undertaken a survey of Quasars (QSO) in the
almost unexplored redshift range 0.9 < z < 2.2. It will track both the Baryonic Acoustic
Oscillations (BAO) and the Redshift Space Distortions (RSD) of the 2-point correlation
function to extract cosmological information on the geometry of the universe and the nature
of dark energy. In particular, it will measure the growth rate of structures and allow for a
test of general relativity modifications and dark energy scenarios. Indeed, at the redshift
range of the eBOSS QSO sample, any deviation from general relativity predictions on
the growth rate of structures would start being a powerful discriminant between different
modified gravity models such as the Galileon. The eBOSS survey started 2 years ago and
preliminary results will be presented.
To perform this measurement, special care should be given to the RSD model. The most
popular model is the Gaussian Streaming model which convolutes a pairwise velocity probability
distribution function (PDF) assumed to be Gaussian with the real space correlation
function. In this talk, I will present one of the most recent RSD model based on Convolution
Lagrangian Perturbation Theory (CLPT) and its applicability for the QSO tracer.
CLPT provides predictions on velocity and real-space clustering statistics that need to be
tuned on N-body simulations. Moreover, the halo occupancy distribution of QSO in dark
matter halos which links the properties of galaxies with the ones of their hosted dark matter
halos can be investigated and we will examine at which scale this model is valid in the
redshift range of eBOSS. In addition, we resort to mock catalogues as a benchmark of our
analysis and more specifically to estimate the matrix describing the expected covariance
of our measurement.
Finally, recent studies started involving small scales where non-linear evolution has to be
taken into account. This will enlarge the range of scales and hence reduce the statistical
error. This task is not easy since different scale-dependent effects have to be considered, so
it complicates a lot the possibility to have an unique model which describes the distortions
in the clustering pattern at all scales. One way of improving RSD models would be to use
the N-body simulations to find the relevant quantities to be injected in the description of
the full infall velocity PDF such as local environment parameters. Going in that direction,
I will present an attempt to parametrize this PDF and then, using a specific streaming
model, we plan to quantify the difference with the Gaussian Streaming Model.

Primary author

Presentation materials