Cosmic-rays, gas, and dust in nearby anti-centre clouds

Abstract

Aims. We have explored the capabilities of dust extinction and γ rays to probe the properties of the interstellar medium in the nearby anti-centre region. In particular, we aim at quantifying the variations of the dust properties per gas nucleon across the different gas phases and different clouds. The comparison of dust extinction and emission properties with other physical quantities of large grains (emission spectral index β, dust colour temperature Tdust, total-to-selective extinction factor RV) helps the theoretical modelling of grains as they evolve from diffuse to dense cloud environments. Methods. We have jointly modelled the γ-ray intensity, recorded between 0.4 and 100 GeV with the Fermi Large Area Telescope (LAT), and the stellar reddening, E(B − V), inferred from Pan-STARRS and 2MASS photometry, as a combination of HI-bright, CO-bright, and ionised gas components. The complementary information from dust reddening and γ rays is used to reveal the gas not seen, or poorly traced, by HI, free-free, and 12CO emissions, namely (i) the opaque HI and diffuse H2 present in the dark neutral medium (DNM) at the atomic-molecular transition, and (ii) the dense H2 to be added where 12CO lines saturate (COsat). We compare the total gas column densities, NH, derived from the γ rays and stellar reddening with those inferred from a similar, previously published analysis of γ rays and of the optical depth of the thermal dust emission, τ353, at 353 GHz. We can therefore compare environmental variations in specific dust reddening, E(B − V)∕NH, and in dust emission opacity (dust optical depth per gas nucleon), τ353∕NH. Results. The gas column densities obtained when combining γ rays with either dust reddening or dust emission compare reasonably well in the atomic and DNM gas phases and over most of the CO-bright phase, but we find localised differences in the dense media (COsat component) due to differences in the two dust tracers. Over the whole anti-centre region, we find an average E(B − V)∕NH ratio of (2.02 ± 0.48) ×10−22 mag cm2, with maximum local variations of about ± 30% at variance with the two to six fold coincident increase seen in emission opacity as the gas column density increases. We show how the specific reddening and opacity vary with the colour temperature and spectral index of the thermal emission of the large grains. Additionally, we find a better agreement between the XCO = N(H2)/WCO conversion factors derived with dust reddening or with γ rays than with those inferred from dust emission, especially towards clouds with large τ353 optical depths. The comparison confirms that the high XCO values found with dust emission are biased by the significant rise in emission opacity inside molecular clouds. Conclusions. In the diffuse medium, we find only small variations in specific reddening, E(B − V)∕NH, compatible with the dispersion in the RV factor reported by other studies. This implies a relatively uniform dust-to-gas mass ratio in the diffuse parts of the anti-centre clouds. The small amplitude of the E(B − V)∕NH variations with increasing NH column density confirms that the large opacity τ353∕NH rise seen towards dense CO clouds is primarily due to changes in dust emissivity. The environmental changes are qualitatively compatible with model predictions based on mantle accretion on the grains and the formation of grain aggregates.