Are quasars reliable standard candles?

In this paper we address the question whether the non-linear relation between the X-ray and UV emission of quasars can be used to derive their distances. In previous works of our group, we demonstrated that such a relation does not show any redshift evolution in its slope. The derived distances are in agreement with the standard flat $\Lambda$CDM model up to $z$$\sim$1.5, but they show significant deviations at higher redshifts. Yet, several authors suggested that this discrepancy is due to inconsistencies between the low- and high-redshift sources within the parent sample, or to a redshift evolution of the relation. Here, we discuss these issues through a quantitative comparison with supernova-derived distances in the common redshift range, complemented by simulations showing that all the claimed inconsistencies would naturally arise from any limitation of the cosmological model adopted for the data analysis, that is, from our ignorance of the true cosmology. We argue that the reliability of the method can only be based on a cosmology-independent evaluation of the hypothesis of non-evolution of the X-ray to UV relation at $z$>1.5, subsequent to a careful check of the sample selection and of the flux measurements for possible redshift-dependent systematic effects. Since we do not conceive any physical reason for a sudden change of the normalization of the relation at $z$>1.5, and we can exclude any severe systematic effect in the data selection and flux measurements, we conclude that the application of the X-ray to UV relation to cosmology is well motivated. To further strengthen this point, we need to achieve a better understanding of the physical process behind the observed relation and/or an independent observational proof possibly confirming the discrepancy with $\Lambda$CDM found with quasars, such as future supernova measurements at $z$$\sim$2 or higher.