LARGE VELOCITY DISPERSION HI IN THE GALACTIC HALO

The Leiden/Dwingeloo Survey (LDS) offers new possibilities for analysi ng Galactic HI with an outstanding sensitivity. The survey data have b een carefully corrected for side-lobe contamination of the antenna and for baseline effects. At present this survey is the most reliable dat abase for analysis of faint, large-scale HI features. Together with th e longstanding dispute whether the Galactic halo is hot or cold, this motivated our investigations which are described in this paper. We hav e improved the stray-radiation correction procedure significantly by i ncluding reflections from the ground. HI-gas with an unusually large v elocity dispersion (LVD) is revealed when these enhanced LDS Data are massively integrated. Gaussian decomposition of more than 250 integrat ed profiles for b > 20 degrees yields a complete set of 8500 HI-compon ents representing the north galactic sky on 10 degrees x 10 degrees fi elds. LVD components were found in every direction of the sky having a characteristic dispersion of greater than or equal to 60 kms(-1) and column densities of greater than or equal to 1.4.10(19) cm(-2). We do not detect the HI-gas which is associated with the ''Lockman-Layer'' ( sigma similar to 35 kms(-1)) and conclude therefore that his analysis was biased by instrumental effects. Correction of this bias in the Hel l Survey data set makes the ''Lockman-Layer'' disappear and does show the LVD component reported here. The LVD gas exhibits minimal sub-rota tion and extends several kpc into the Galactic halo. Its scale height is calculated to be greater than or equal to 2 kpc and a non-uniform d istribution with respect to the distance rom the Galactic center is fo und. The distribution of the LVD gas is presented. Theoretical spectra are calculated from a simple model of the LVD halo and compared to th e real data. The LVD gas seems to be a very sensitive indicator of vio lent disk phenomena. Two possible Galactic chimneys and a peculiar loc al velocity field are identified, Implications for the modelling of a Galactic halo with various components are discussed. Strict applicatio n of the principle, that the turbulent gas pressure plus magnetic and cosmic ray pressure equals the gravitational pull, leads to a stable h alo, extending up to a = 3.3 kpc.