Astron. Astrophys. 355, 333-346 (2000)

6. Summary

We began by comparing OH and HCO⁺ profiles taken over an approximately five year interval. Profiles are generally stable, but may exhibit small differences. Fig. 3 summarizes the results over time, showing profiles of the HCO⁺ optical depth and optical depth difference, along with the spectrum of the expected variance. Variations of the sort observed in the strong lines toward B0528+134 and B2200+420 are largely integral-conserving and placed near the line center, and could in principle be explained solely as a slight velocity shift. Some of the variations seen toward B0355+508 are different in character - non-conserving, narrow and off-center. In general, the differences seen in HCO⁺ are not obviously mimicked in OH (and vice versa ). Overall, things seem to change at the level of 1%-2% only. Longer time baselines are required.

Between species, OH and HCO⁺ profiles viewed at 0.14-0.24 km s^-1 resolution are disarmingly similar. There is no signature in the kinematics of macroscopic shocks - broad wings off narrow cores, profile doubling, - or of gross chemical differentiation. OH and HCO⁺ lines have pretty much the same linewidths in the same components, while showing occasional small differences such as toward B1730-130 where the red wing of the HCO⁺ line is missing in OH. Unfortunately the OH absorption profile does not bear the same direct optical depth - column density relationship applicable to HCO⁺, because the OH excitation temperature is not negligible. So profile differences could in principle be explained by cloud substructure independent of the kinematics. The general absence of OH-HCO⁺ optical depth profile differences should be useful in constraining internal cloud structure.

We found a new component of molecular absorption due to distributed gas, seen over the same range where the optical depth is appreciable in H I. Toward B0355+508 at b = -1.6^o the HCO⁺ absorption extends continuously over more then 40 km s^-1, sampling gas out to R = 1.6 R₀. This behaviour can be straightforwardly explained by current estimates of the molecular gas fraction at/beyond the Solar Circle, if the abundance of HCO⁺ has its typically observed value wherever forms in noticeable quantities. Using a model of the H- and C⁺-CO conversion based entirely on physical mechanisms detailed in the literature, we showed that modern values of the various physical parameters predict that formation can occur at quite low densities, which accounts for the existence of gas which readily absorbs in the HCO⁺ lines but emits little in mm-wave species (this gas does however appear in OH emission, at least toward B0355+508). We then pursued the consequences of this ubiquitous HCO⁺ gas and showed through modelling that the observed behaviour of N(CO) N() can be explained solely by allowing HCO⁺ to recombine to CO in otherwise standard diffuse cloud models.

Finally, we considered the nature of the cloud substructure which might be responsible for the ubiquitous profile variations seen in so many high-resolution line absorption experiments, including those exemplified in Fig. 3. For the diffuse but partially molecular gas, the weakness of HCO⁺ emission is a powerful constraint: using it we were able to show very straightforwardly that it is impossible to construct a successful model of the usual kind in which small, dense clumps are supposed to produce frequent optical depth fluctuations while remaining unobtrusive in emission. Instead, we conclude that it is necessary to incorporate small-scale chemical and/or other inhomogeneities - a fractal geometry? - and that these will give the added latitude necessary to reproduce the rapidly-growing body of emission and absorption measurements of molecules in the diffuse interstellar medium.

© European Southern Observatory (ESO) 2000

Online publication: March 17, 2000
helpdesk.link@springer.de