Determining $H_0$ using a model-independent method

Abstract: By using type Ia supernovae (SNIa) to provide the luminosity distance (LD) directly, which depends on the value of the Hubble constant $H_0= 100 h\; {\rm km\; s^{-1}\; Mpc^{-1}}$, and the angular diameter distance from galaxy clusters or baryon acoustic oscillations (BAOs) to give the derived LD according to the distance duality relation, we propose a model-independent method to determine $h$ from the fact that different observations should give the same LD at a given redshift. Combining the Sloan Digital Sky Survey II (SDSS-II) SNIa from the MLCS2k2 light curve fit and galaxy cluster data, we find that at the $1\sigma$ confidence level (CL), $h=0.5867\pm0.0303$ for the sample of the elliptical $\beta$ model for galaxy clusters, and $h=0.6199\pm0.0293$ for that of the spherical $\beta$ model. The former is smaller than the values from other observations, whereas the latter is consistent with the Planck result at the $2\sigma$ CL and agrees very well with the value reconstructed directly from the $H(z)$ data. With the SDSS-II SNIa and BAO measurements, a tighter constraint, $h=0.6683\pm0.0221$, is obtained. For comparison, we also consider the Union 2.1 SNIa from the SALT2 light curve fitting. The results from the Union 2.1 SNIa are slightly larger than those from the SDSS-II SNIa, and the Union 2.1 SNIa + BAOs give the tightest value. We find that the values from SNIa + BAOs are quite consistent with those from the Planck and the BAOs, as well as the local measurement from Cepheids and very-low-redshift SNIa.