[en] In this paper, we describe a procedure for modelling strong lensing galaxy clusters with parametric methods, and to rank models quantitatively using the Bayesian evidence. We use a publicly available Markov chain Monte-Carlo (MCMC) sampler ('bayesys'), allowing us to avoid local minima in the likelihood functions. To illustrate the power of the MCMC technique, we simulate three clusters of galaxies, each composed of a cluster-scale halo and a set of perturbing galaxy-scale subhalos. We ray-trace three light beams through each model to produce a catalogue of multiple images, and then use the MCMC sampler to recover the model parameters in the three different lensing configurations. We find that, for typical Hubble Space Telescope (HST)-quality imaging data, the total mass in the Einstein radius is recovered with ∼1-5% error according to the considered lensing configuration. However, we find that the mass of the galaxies is strongly degenerated with the cluster mass when no multiple images appear in the cluster centre. The mass of the galaxies is generally recovered with a 20% error, largely due to the poorly constrained cut-off radius. Finally, we describe how to rank models quantitatively using the Bayesian evidence. We confirm the ability of strong lensing to constrain the mass profile in the central region of galaxy clusters in this way. Ultimately, such a method applied to strong lensing clusters with a very large number of multiple images may provide unique geometrical constraints on cosmology. The implementation of the MCMC sampler used in this paper has been done within the framework of the lenstool software package, which is publicly available

[en] We report the clear detection of the 2175 A dust absorption feature in the optical afterglow spectrum of the gamma-ray burst (GRB) GRB 070802 at a redshift of z = 2.45. This is the highest redshift for a detected 2175 A dust bump to date, and it is the first clear detection of the 2175 A bump in a GRB host galaxy, while several tens of optical afterglow spectra without the bump have been recorded in the past decade. The derived extinction curve gives A_V = 0.8-1.5 depending on the assumed intrinsic slope. Of the three local extinction laws, a Large Magellanic Cloud (LMC) type extinction gives the best fit to the extinction curve of the host of GRB 070802. Besides the 2175 A bump we find that the spectrum of GRB 070802 is characterized by unusually strong low-ionization metal lines and possibly a high metallicity for a GRB sightline ([Si/H] = -0.46 ± 0.38, [Zn/H] = -0.50 ± 0.68). In particular, the spectrum of GRB 070802 is unique for a GRB spectrum in that it shows clear C I absorption features, leading us to propose a correlation between the presence of the bump and C I. The gas-to-dust ratio for the host galaxy is found to be significantly lower than that of other GRB hosts with N(H I)/A_V = (2.4 ± 1.0) x 10²¹ cm^-2 mag^-1, which lies between typical Milky Way and LMC values. Our results are in agreement with the tentative conclusion reached by Gordon et al. that the shape of the extinction curve, in particular the presence of the bump, is affected by the UV flux density in the environment of the dust.