Title: Generalized Chaplygin Gas Models tested with SNIa

Abstract: The so called Generalized Chaplygin Gas (GCG) with the equation of state $p = - \frac{A}{{\rho}^{\alpha}}$ was recently proposed as a candidate for dark energy in the Universe. In this paper we confront the GCG with SNIa data. Specifically we have tested the GCG cosmology in three different classes of models with (1) $\Omega_m= 0.3$, $\Omega_{Ch}= 0.7$; (2) $\Omega_m= 0.05$, $\Omega_{Ch}= 0.95$ and (3) $\Omega_m = 0$, $\Omega_{Ch} = 1$, as well as the model withouth any assumption on $\Omega_m$. The best fitted models are obtained by minimalizing the $\chi^2$ function and $\chi^2$ levels in the $(A_0, \alpha)$ plane. We supplemented our analysis with confidence intervals in the $(A_0, \alpha)$ plane, as well as one-dimensional probability distribution functions for models parameter. The general conclusion is that SNIa data strongly support the Chaplygin gas (with $\alpha = 1$). Extending our analysisby relaxing the flat prior lead to the result that even though the best fitted values of $\Omega_k$ are formally non-zero, still they are close to flat case. It should be viewed as an advantage of the GCG model since in similar analysisof $\Lambda$CDM model high negative value of $\Omega_{k}$ were found to be bestfitted to the data and independent inspiration from CMBR and extragalactic astronomy has been invoked to fix the curvature problem. Our results show clearly that in Generalized Chaplygin Gas cosmology distant $z >1$ supernovae should be brighter than in $\Lambda$CDM model.This prediction seems to be confirmed with new Riess high redshift SNIa sample. Moreover, we argue that with the future SNAP data it would be possible to differentiate between models with various value of $\alpha$ parameter and/or discriminated between GCG, Cardassian and $\Lambda$CDM models