Since their discovery by Muller et al. (1963), HI clouds moving with velocities that can not be explained by differential Galactic rotation (often exceeding it by 100 km s-1) have been the target of numerous surveys. Many of these high-velocity clouds (HVCs) are located at intermediate and high Galactic latitudes (Giovanelli et al. 1973; Mathewson et al. 1977; Wakker & van Woerden 1991) and do not appear to have any connection with the gas in the Galactic disk. HVCs are often found in large HI complexes with angular sizes 10-90o. They cover from 10 to 37% of the sky, depending on the sensitivity of the studies (Murphy et al. 1995).
The origin of HVCs is still unclear, mainly because the distances to the individual complexes are in most cases unknown. They could be cold gas corresponding to the return flow in a Galactic fountain (e.g. Houck & Bregman 1990), or gas left over from the formation of the Galaxy. Some HVCs are probably belonging to the tidal gas streams torn from the Magellanic Clouds by the Milky-Way (Mathewson et al. 1974; Putman & Gibson 1999).
Several authors have already explored the possibility that HVCs are infalling primordial gas and have associated them with the Local Group (see Wakker & van Woerden 1997 for a thorough review). Recently Blitz et al. (1999) re-examined this hypothesis, and simulating the dynamical evolution of the Local Group galaxies, used the up-to-date HI maps of HVCs to show that the HVCs are consistent with a dynamical model of infall of the ISM onto the Local group. As such, they would represent the building blocks of our galaxies in the Local group and provide fuel for star formation in the disk of the Milky Way. In their model, HVCs contain altogether 1011 of neutral gas. From their stability analysis, they conclude that there is roughly 10 times more dark matter than luminous gas within each HVC, and that these could correspond to the mini-halos which are able to accumulate baryons, and can gather into filaments (e.g. Bond et al. 1988; Babul & Rees 1992; Kepner et al. 1997). HVCs would therefore be related to the hierarchical structure of the Universe (see e.g. Katz et al. 1996), and to the gas seen in absorption towards quasars (Lyman- forest and Lyman-limit lines). However, Giovanelli (1981) has pointed out that the velocity distribution of HVCs does not match that of the Local Group, but does favor an association with the Magellanic Stream, the most obvious tidal feature of the interaction between the Galaxy and the Magellanic Clouds. The discrepancy with the results from Blitz et al. (1999) comes from the fact that the latter authors have not considered all observed HVCs, but only a selection of them.
Maps of the brightest HVC complexes have revealed the existence of unresolved structure at 10 arcmin resolution, which was further resolved into high-density cloud cores at 1 arcmin resolution (Giovanelli & Haynes 1977; Wakker & Schwarz 1991; Wakker & van Woerden 1991). More generally, HVCs follow the fractal structure observed in the whole interstellar medium (Vogelaar & Wakker 1994). The HI column densities in these cores are estimated to be several times 1020 cm-2 and their temperatures are generally between 30 and 300 K. The central densities of individual clouds can reach 80 cm-3 , where Dkpc is their distance in kpc. Depending on the actual distance, which still remains poorly determined, those conditions make the HVC cores possible sites of star formation. HVCs have in fact been considered good candidates for the source of young Population I stars at large distances from the Galactic plane (see e.g. Sasselov 1993).
Attempts to measure the spin temperature of atomic hydrogen through 21cm absorption in front of background continuum sources have often only resulted in upper limits (Colgan et al. 1990; Mebold et al. 1991). A few detections have been reported (Payne et al. 1980; Wakker et al. 1991; Akeson & Blitz 1999), with inferred spin temperature as low as 36 K, but in general most HVCs must have spin temperatures larger than 200 K or be very clumpy.
Most of our current knowledge of the gaseous content of HVCs comes from HI observations. Efforts to search for molecular hydrogen using CO emission lines have been so far unsuccessful (Wakker et al. 1997), since the sub-solar metallicity and/or low density of HVCs makes direct CO emission line detection very difficult. However, optical absorption lines have shown that HVCs are not completely devoid of heavy elements (e.g. Robertson et al. 1991; Lu et al. 1994; Keenan et al. 1995; Wakker & van Woerden 1997). The lines detected are from SiII, CII, FeII, or CIV, but the strongest are from MgII (Savage et al. 1993; Bowen & Blades 1993; Sembach et al. 1995, 1998). Metallicity studies have been done in HVCs to determine their origin. If HVCs result from the Galactic fountain effect, their metallicity should be solar while if they were associated with the Magellanic or intergalactic Clouds, it could be even less than 0.1 solar. In fact, the determined abundances are around 0.1 solar, but with much uncertainties, because of saturated lines, or dust depletion (Sembach & Savage 1996). This average metallicity is compatible with a Local-Group infall model, since X-ray observations have revealed abundances of 0.1 solar in poor groups (Davis et al. 1996), and even 0.3 solar in intra-cluster gas (Renzini 1997).
If HVCs were local analogues of Lyman-limit absorbing clouds, background QSOs could enable us to detect molecules in absorption, a task which is considerably easier to achieve. Absorption has recently been reconfirmed as a very powerful tool in the millimetric range (Lucas & Liszt 1996 hereafter LL96; Combes & Wiklind 1996). A mm molecular detection would advance our knowledge of HVCs, their physical conditions, and their possible origin. A first attempt has been made in this domain by Akeson & Blitz (1999) with only negative results, using the BIMA and OVRO interferometers.
Here we report on HCO+(1-0) absorption line observations, made in the southern sky with the single dish 15m ESO-SEST telescope, and in the northern sky with the IRAM interferometer. The choice of HCO+ is justified because, due to its large dipole moment, it is generally not excited in diffuse media (the critical density for excitation is 107 cm-3); therefore confusion with emission is not a problem, contrary to the CO(1-0) line which requires an interferometer to resolve out the emission. Furthermore, the HCO+ absorption survey carried out by Lucas & Liszt (1994, 1996) has revealed more and wider absorption lines than in CO, suggesting that it might be a better tracer. In that study, the derived abundance HCO+/H2 was surprisingly large, of the order of 610-9, and sometimes even an order of magnitude larger. The details of the observations are presented in Sect. 2. Sect. 3 summarizes the results, which are then discussed in Sect. 4.
© European Southern Observatory (ESO) 2000
Online publication: May 3, 2000