Aims. We present 3.65 × 3.34 angular-resolution IRAM Plateau de Bure Interferometer (PdBI) observations of the CS J = 2-1 line toward the Horsehead Photodissociation Region (PDR), complemented with IRAM-30m single-dish observations of several rotational lines of CS, C 34 S and HCS + . We analyse the CS and HCS + photochemistry, excitation and radiative transfer to obtain their abundances and the physical conditions prevailing in the cloud edge. Since the CS abundance scales to that of sulfur, we determine the gas phase sulfur abundance in the PDR, an interesting intermediate medium between translucent clouds (where sulfur remains in the gas phase) and dark clouds (where large depletions have been invoked). Methods. A nonlocal non-LTE radiative transfer code including dust and cosmic background illumination adapted to the Horsehead geometry has been developed to carefuly analyse the CS, C 34 S, HCS + and C 18 O rotational line emission. We use this model to consistently link the line observations with photochemical models to determine the CS/HCS + /S/S + structure of the PDR. Results. Densities of n(H 2 ) (0.5−1.0) × 10 5 cm −3 are required to reproduce the CS and C 34 S J = 2-1 and 3-2 line emission. CS J = 5-4 lines show narrower line widths than the CS low-J lines and require higher density gas components not resolved by the ∼10 IRAM-30m beam. These values are larger than previous estimates based in CO observations. We found χ(CS) = (7 ± 3) × 10 −9 and χ(HCS + ) = (4 ± 2) × 10 −11 as the averaged abundances in the PDR. According to photochemical models, the gas phase sulfur abundance required to reproduce these values is S/H = (3.5 ± 1.5) × 10 −6 , only a factor < ∼ 4 less abundant than the solar sulfur elemental abundance. Since only lower limits to the gas temperature are constrained, even lower sulfur depletion values are possible if the gas is significantly warmer. Conclusions. The combination of CS, C 34 S and HCS + observations together with the inclusion of the most recent CS collisional and chemical rates in our models implies that sulfur depletion invoked to account for CS and HCS + abundances is much smaller than in previous studies.Key words. astrochemistry -ISM: clouds -ISM: molecules -ISM: individual objects: Horsehead nebula -radio lines: ISMradiative transfer IntroductionSulfur is an abundant element (the solar photosphere abundance is S/H = 1.38 × 10 −5 ; Asplund et al. 2005), which remains undepleted in diffuse interstellar gas (e.g. Howk et al. 2006) and HII regions (e.g. Martín-Hernández et al. 2002;García-Rojas et al. 2006, and references therein) but it is historically assumed to deplete on grains in higher density molecular clouds by factors as large as ∼10 3 (Tieftrunk et al. 1994). This conclusion is simply reached by adding up the observed gas phase abundances of S-bearing molecules in well known dark clouds such as TMC1 (e.g. Irvine et al. 1985;Ohishi & Kaifu 1998). As Appendix A is only available in electronic form at http://www.edpsciences.org sulfur is easily ionized (ioniz...

The development of radiation hydrodynamical methods that are able to follow gas dynamics and radiative transfer (RT) self‐consistently is key to the solution of many problems in numerical astrophysics. Such fluid flows are highly complex, rarely allowing even for approximate analytical solutions against which numerical codes can be tested. An alternative validation procedure is to compare different methods against each other on common problems, in order to assess the robustness of the results and establish a range of validity for the methods. Previously, we presented such a comparison for a set of pure RT tests (i.e. for fixed, non‐evolving density fields). This is the second paper of the Cosmological Radiative Transfer Comparison Project, in which we compare nine independent RT codes directly coupled to gas dynamics on three relatively simple astrophysical hydrodynamics problems: (i) the expansion of an H ii region in a uniform medium, (ii) an ionization front in a 1/r2 density profile with a flat core and (iii) the photoevaporation of a uniform dense clump. Results show a broad agreement between the different methods and no big failures, indicating that the participating codes have reached a certain level of maturity and reliability. However, many details still do differ, and virtually every code has showed some shortcomings and has disagreed, in one respect or another, with the majority of the results. This underscores the fact that no method is universal and all require careful testing of the particular features which are most relevant to the specific problem at hand.