Slope evolution of GRB correlations and cosmology

Gamma-ray bursts (GRBs) observed up to redshifts z > 9.4 can be used as possible probes to test cosmological models. Here we show how changes of the slope of the luminosity L-X*-break time T-a* correlation in GRB afterglows, hereafter the LT correlation, affect the determination of the cosmological parameters. With a simulated data set of 101 GRBs with a central value of the correlation slope that differs on the intrinsic one by a 5 sigma factor, we find an overestimated value of the matter density parameter, Omega(M), compared to the value obtained with Type Ia supernovae, while the Hubble constant, H-0, best-fitting value is still compatible in 1 sigma compared to other probes. We show that this compatibility of H-0 is due to the large intrinsic scatter associated with the simulated sample. Instead, if we consider a subsample of high-luminosity GRBs (High L), we find that the evaluation of both H-0 and Omega(M) is not more compatible in 1 sigma and Omega(M) is underestimated by 13 per cent. However, the High L sample choice reduces dramatically the intrinsic scatter of the correlation, thus possibly identifying this sample as the standard canonical 'GRBs' confirming previous results presented by Dainotti et al. Here, we consider the LT correlation as an example, but this reasoning can also be extended for all other GRB correlations. In the literature so far, GRB correlations are not corrected for redshift evolution and selection biases; therefore, we are not aware of their intrinsic slopes and consequently how far the use of the observed correlations can influence the derived 'best' cosmological settings. Therefore, we conclude that any approach that involves cosmology should take into consideration only intrinsic correlations and not the observed ones.