Impact of lensing magnification on the analysis of galaxy clustering in redshift space

¹ Aix Marseille Univ, CNRS, CNES, LAM, Marseille, France
e-mail: breton@ice.csic.es
² Institute of Space Sciences (ICE, CSIC), Campus UAB, Carrer de Can Magrans, s/n, 08193 Barcelona, Spain
³ Institut d’Estudis Espacials de Catalunya (IEEC), Carrer Gran Capità 2-4, 08193 Barcelona, Spain

Received: 20 October 2021
Accepted: 8 March 2022

Abstract

We study the impact of lensing magnification on the observed three-dimensional galaxy clustering in redshift space. We used the RayGal suite of N-body simulations, from which we extracted samples of dark matter particles and haloes in the redshift regime of interest for future large redshift surveys. Several magnitude-limited samples were built that reproduce various levels of magnification bias ranging from s = 0 to s = 1.2, where s is the logarithmic slope of the cumulative magnitude number counts, in three redshift intervals within 1 < z < 1.95. We studied the two-point correlation function multipole moments in the different cases in the same way as would be applied to real data, and investigated how well the growth rate of structure parameter could be recovered. In the analysis, we used an hybrid model that combines non-linear redshift-space distortions and linear curved-sky lensing magnification. We find that the growth rate is underestimated when magnification bias is not accounted for in the modelling. This bias becomes non-negligible for z > 1.3 and can reach 10% at z ≃ 1.8, depending on the properties of the target sample. In our data, adding the lensing linear correction allowed us to recover an unbiased estimate of the growth rate in most cases when the correction was small, even when the fiducial cosmology was different from that of the data. For larger corrections (high redshifts, low bias, and high s value), we find that the weak-lensing limit has to be treated with caution as it may no longer be a good approximation. Our results also show the importance of knowing s in advance instead of letting this parameter free with flat priors because in this case, the error bars increase significantly.