1994A&A...287..601P

We made a systematic separation of both the neutral phases using the atlases of 21-cm profiles of Heiles & Habing (1974) and Colomb et al. (1980), complemented with other data. First, we fitted the emission of the warm neutral medium (WNM) by means of a broad Gaussian curve (velocity dispersion σ∼10-14 km/s). No additional broad components were required. We derived maps of the column densities N_WH and the radial velocities V_W of the WNM. Its overall distribution appears to be very inhomogeneous with a large hole in the range b ≥+50deg. However, if the hole is excluded, the mean latitude-profiles admit a rough cosec |b|-fit common to both hemispheres. A kinematical analysis of V_W for the range 10deg ≤|b|≤ 40deg indicates a mean differential rotation with a small nodal deviation. At |b|> 50deg V_W is negative, with larger values and discontinuities in the north. On the mean, σ increases for |b| decreasing, as is expected from differential rotation. From a statistical study of the peaks of the residual profiles we derived some characteristics of the cold neutral medium (CNM). The latter is generally characterized by a single component of σ∼2-6 km/s. Additionally we derived the sky-distribution of the column densities N_CH and the radial velocities V_c of the CNM within bins of 1.2deg sec b x 1deg in l, b. The bulk of the CNM is localized at |b|≤30deg with a clumpy distribution and V_c predominantly positive. At |b|> 45deg the gas is very much fainter. No plane parallel stratification is apparent. There are notorious asymmetries between both hemispheres regarding the distributions of both N_CH and V_c. On the mean N_CH is about 30% larger in the south. Furthermore, we focused on the characteristics of Lindblad's feature A of cool gas by considering the narrow ridge of local H I, which appears in the b-V contour maps at fixed l (e.g. Schober 1976). The ridge appears to be the main component of the CNM. For |b|< 25deg the kinematics of feature A is consistent with an expansion. Olano's (1982) shell models fit the mean velocities for most longitudes. However, the data are not sensitive to the shape adopted for the shell. We suggest the occurrence of rolling motions within it. The large molecular cloud complexes of the Gould belt are well correlated in position and velocity with the ridge of cold local H I. In addition, a significant percentage of the molecular clouds at high |b| appears to be associated to tht H I-ridge. We suggest a scenario for the formation and evolution of the Gould belt system of stars and gas on the basis of an explosive event within a shingle of cold dense gas tilted to the galactic plane. The scenario appears to be consistent with the results found for both the neutral phases, as well as with Danly's (1989) optical and UV observations of interstellar cool gas in the lower halo.