Extraplanar DIG (eDIG) in halos of spiral galaxies is most likely correlated with star forming regions in the galactic plane. In current models this interstellar gas phase is blown out into the halo of a galaxy by supernova (SN) explosions. It extends up to 1-2 kpc above the star forming regions in the galactic disk and is disposed in filaments, shells, or bubbles. After roughly yrs there is cooling and condensing of the formerly ionized matter and high velocity clouds (HVC) are created.
Fig. 1.1: Schematic overview of a large scale convection of ionized matter. SNe in the galactic disk create chimney structures through which ionized gas can reach the halo. Cooling and condensing of the ionized gas phase leads after roughly yrs to an infall of neutral matter onto the disk (Figure taken from Norman & Ikeuchi 1989).
These HVCs can fall back onto the disk so that this large scale convection may be responsible for the metallicity distribution in direction. Due to the detection of diffuse ionized emission at high galactic latitudes a very important question for the understanding of the dynamics of the ISM in external galaxies arises, namely: what are the ionization/excitation mechanisms of the eDIG?
Due to the fact that for the DIG
up to now only few spectral observations in a quite restricted
wavelength range exist, the parameter space of theoretical
models describing the ionization of the DIG is restricted by the line
ratios of few bright emission lines only. Consequently it is
presently not clear what the importance of different ionization
mechanisms such as photoionization by the interstellar radiation
field, shocks, or possibly cooling of coronal gas might be.
However, estimates for the DIG in the Milky Way show that
its power requirement is equal to 100% of the power provided by
supernovae or to 16% of the ionizing radiation of OB stars.
(Reynolds, 1993). This makes OB stars the most likely source of ionization.
Model calculations by Domgörgen & Mathis (1994)
showed that the line ratios, as they are found for the DIG in our galaxy,
can in principle be explained by low-density, low-excitation
photoionization models. However, there are strong constraints which have to be fulfilled if photoionization is indeed the dominant ionization mechanism. One example is the strength of the HeI recombination line at 5876Å. This line is (roughly speaking) a good tracer of the hardness of the stellar radiation field and allows to check if the adopted stellar atmosphere needs to be adjusted in order to reproduce the observed line ratios.
Reynolds & Tufte (1995) searched for this HeI
recombination line at two different line-of-sights. For both directions
an upper limit below the value predicted by model simulations was found and still constitute a serious challenge for photoionization codes.
However, it is not clear how representative the selected directions are for the DIG in general. Observations of extraplanar DIG (eDIG) in external edge-on galaxies could contribute much to our understanding of this gas phase as the DIG-surface brightness is usually higher compared to that of the MW-DIG. In addition, the Doppler shift would help to avoid critical night-sky lines. Indeed, the first detection of DIG in external galaxies dates back more than 10 years when Rand et al. (1990) and Dettmar (1990) reported on a 2kpc detection in the halo of NGC891.
Today, the record holder still is NGC5775 with an extended DIG halo of about 10kpc (Rand 2000,Tüllmann et al. 2000).
While the He discussion demonstrates how the interstellar radiation
field in halos of galaxies can be probed by deep emission line spectra,
it has not yet been possible to address other important physical
quantities for DIG in a quantitative way, namely the electron temperature
and metallicity. To obtain a reliable excitation temperature very faint
diagnostic emission lines such as [NII] 5755Å or
[OIII] 4363Å have to be measured. This is not in reach of
4m class telescopes and therefore only upper limits for the temperature
exist Rand (1997).
Once the temperature could be fixed by using one of these lines it is possible
to obtain reasonable abundances for this gas (see
below
).