H-ATLAS/GAMA: magnification bias tomography. Astrophysical constraints above ∼1 arcmin

An unambiguous manifestation of the magnification bias is the cross-correlation between two source samples with non-overlapping redshift distributions. In this work we measure and study the cross-correlation signal between a foreground sample of GAMA galaxies with spectroscopic redshifts in the range 0:2 < z < 0:8, and a background sample of HATLAS galaxies with photometric redshifts & 1:2. It constitutes a substantial improvement over the cross-correlation measurements made by Gonzalez-Nuevo et al. (2014) with updated catalogues and wider area (with S=N & 5 below 10 arcmin and reaching S=N _ 20 below 30 arcsec). The better statistics allow us to split the sample in different redshift bins and to perform a tomographic analysis (with S=N & 3 below 10 arcmin and reaching S=N _ 15 below 30 arcsec). Moreover, we implement a halo model to extract astrophysical information about the background galaxies and the detectors that are producing the lensing link between the foreground (lenses) and background (sources) samples. In the case of the sources, we find typical mass values in agreement with previous studies: a minimum halo mass to host a central galaxy, Mmin _ 1012:26M_, and a pivot halo mass to have at least one sub-halo satellite, M1 _ 1012:84M_. However, the lenses are massive galaxies or even galaxy groups/clusters, with minimum mass of Mlens min _ 1013:06M_. Above a mass of Mlens 1 _ 1014:57M_ they contain at least one additional satellite galaxy which contributes to the lensing effect. The tomographic analysis shows that, while Mlens 1 is almost redshift independent, there is a clear evolution of increase Mlens min with redshift in agreement with theoretical estimations. Finally, the halo modeling allows us to identify a strong lensing contribution to the cross-correlation for angular scales below 30 arcsec. This interpretation is supported by the results of basic but effective simulations