The VIMOS Public Extragalactic Redshift Survey (VIPERS)

Aims. Building on the two-point correlation function analyses of the VIMOS Public Extragalactic Redshift Survey (VIPERS), we investigate the higher-order correlation properties of the same galaxy samples to test the hierarchical scaling hypothesis at z 1 and the dependence on galaxy luminosity, stellar mass, and redshift. With this work we also aim to assess possible deviations from the linearity of galaxy bias independently from a previously performed analysis of our survey. Methods. We have measured the count probability distribution function in spherical cells of varying radii (3 R 10 h -1 Mpc), deriving 8g (the galaxy rms at 8 h -1 Mpc), the volume-averaged two-, three-, and four-point correlation functions and the normalized skewness S 3g and kurtosis S 4g for different volume-limited subsamples, covering the following ranges: -19.5 M B (z = 1.1) -5 log(h) -21.0 in absolute magnitude, 9.0 log(M * /M h -2 ) 11.0 in stellar mass, and 0.5 z < 1.1 in redshift. Results. We have performed the first measurement of high-order correlation functions at z 1 in a spectroscopic redshift survey. Our main results are the following. 1) The hierarchical scaling between the volume-averaged two-and three-point and two-and four-point correlation functions holds throughout the whole range of scale and redshift we could test. 2) We do not find a significant dependence of S 3g on luminosity (below z = 0.9 the value of S 3g decreases with luminosity, but only at 1-level). 3) We do not detect a significant dependence of S 3g and S 4g on scale, except beyond z 0.9, where S 3g and S 4g have higher values on large scales (R 10 h -1 Mpc): this increase is mainly due to one of the two CFHTLS Wide Fields observed by VIPERS and can be explained as a consequence of sample variance, consistently with our analysis of mock catalogs. 4) We do not detect a significant evolution of S 3g and S 4g with redshift (apart from the increase of their values with scale in the last redshift bin). 5) 8g increases with luminosity, but does not show significant evolution with redshift. As a consequence, the linear bias factor b = 8g / 8m , where 8m is the rms of matter at a scale of 8 h -1 Mpc, increases with redshift, in agreement with the independent analysis of VIPERS and of other surveys such as the VIMOS-VLT Deep Survey (VVDS). We measure the lowest bias b = 1.47 0.18 for galaxies with M B (z = 1.1) -5 log(h) -19.5 in the first redshift bin (0.5 z < 0.7) and the highest bias b = 2.12 0.28 for galaxies with M B (z = 1.1) -5 log(h) -21.0 in the last redshift bin (0.9 z < 1.1). 6) We quantify deviations from the linear bias by means of the Taylor expansion parameter b 2 . We obtain b 2 = -0.20 0.49 for 0.5 z < 0.7 and b 2 = -0.24 0.35 for 0.7 z < 0.9, while for the redshift range 0.9 z < 1.1 we find b 2 = +0.78 0.82. These results are compatible with a null non-linear bias term, but taking into account another analysis for VIPERS and the analysis of other surveys, we argue that there is evidence for a small but non-zero non-linear bias term.