Theories for the origin of the gaseous galactic halo must be able to
explain the
support and ionization of the gas. Two competing models for the support of the
gas are the `galactic fountain model'
(Shapiro and Field
1976;
Bregman 1980;
also see
Chevalier and Oegerle
1979
and Habe and Ikeuchi
1980)
and the cosmic ray supported halo models
(Hartquist, Pettini
and Tallant 1984;
Chevalier and
Fransson 1984).
In the galactic fountain model it is noted that gas will flow into the
halo as a consequence
of supernova explosions which heat and elevate the pressure of gaseous
regions in the
galactic disk. The regions of elevated pressure may breakout from the
galactic plane
and provide an injection of gas into the halo where the gas may cool and
return to
the disk in a flow pattern resembling a fountain. Calculations of the
hydrodynamics
of the breakout process produced by correlated supernovae in OB
associations are found in
Mac Low and McCray
(1988)
while the physical effects of falling clouds on
the galactic disk are considered by
Tenorico-Tagle et
al. (1987).
A process that will
occur on a more gradual basis is `galactic convection'. Here it is
recognized from far ultraviolet absorption line studies of OVI
(Jenkins 1978)
and soft diffuse X-ray studies
(McCammon et
al. 1983)
that there exists in the galactic disk a hot (T
0.3
to 2 x 106 °K) low density (n
10-3
cm-3) phase of the interstellar medium which
may fill much of the volume of the galactic disk. Hot (106
°K) gas in the solar
region of the galaxy is buoyant. Its thermal scale height is about 7
kpc. The gas will
therefore tend to flow outward away from the galactic plane into the
halo where it
may cool and return to the disk as cooler clouds in a convective like
flow. The spatial
characteristics observed for the cool gas as traced through its HI 21 cm
emission
illustrated in Figure 3 is
suggestive of a fountain like or convective like flow. In the
cosmic ray supported halo models it is assumed that the gas is supported
at its large
z distances by the pressure of outwardly streaming cosmic rays.
For the ionization of the gas, the possibilities for the production of
the highly
ionized species (SiIV , CIV and NV) include electron collisional
ionization in 0.8 to
3 x 105 °K gas and photoionization by hot white dwarf
stars
(Dupree and Raymond
1983),
by normal Population I stars, and by the extragalactic EUV background. A
number of recent calculations have concentrated on determining the
production of the highly ionized gas by photoionization
(Hartquist, Pettini
and Tallant 1984;
Chevalier and
Fransson 1984;
Fransson and
Chevalier 1985;
Bregman and
Harrington 1986).
The predictions of
Chevalier and
Fransson (1984)
are shown in Figure 5d. From this
work it appears possible to understand the observed amounts of SiIV and CIV and
in particular the sudden rise in N(ion)|sin b| near z = 1 kpc
from photoionization by
the EUV background. However, the various calculations have difficulty
producing the observed amount of NV
(Savage and Massa 1985,
1987).
NV is an important ion since
among those ions accessible to the IUE, it requires the greatest amount
of energy
for its production (77 eV). Most hot stars containing He have strong
He+ edges at
54 eV. Therefore, the only stellar sources that might be capable of
converting NIV
into NV are the very hot hydrogen white dwarfs. In order to explain the
observed NV,
Edgar and Chevalier
(1986)
calculated the amount of SiIV, CIV, NV and OVI
produced in cooling gas in a galactic fountain flow. For a galactic
fountain mass
flow rate of 4 M
yr-1 on each side of the galactic plane, they predict the column
densities perpendicular to the galactic plane listed in
Table 3. We see that in the
cooling fountain flow enough NV is produced to explain the observations. Their
calculation also predicts CIV and OIII] emission line strengths which
are compatible with the recent diffuse background measurements of
Martin and Bowyer
(1987).
The theory predicts a large OVI column density [N(OVI) = 5.8-6.0 x
1014 cm-2] while the measurements of
Jenkins (1978)
yield a column density of approximately 3 x 1013
cm-2 out to a |z| of about 1 kpc. If the theory is
correct such a large difference
will require the presence of substantial quantities of OVI in the halo
for |z| > 1 kpc.
Future measures with the proposed Lyman spacecraft will provide a
crucial test of the cooling fountain gas theory.
| Ion | Predicted a N(cm-2) | Observed b N(cm-2) |
| SiIV | (3.3-6.4) x 1012 | 2 x 1013
|
| CIV | (4.3-7.9) x 1013 | 1 x 1014
|
| NV | (2.8-3.6) x 1013 | 3 x 1013
|
| OVI | (5.8-6.0) x 1014 | > 3 x 1013 |
aPredicted values assume a fountain mass flow rate of 4
M
b Observations are from
Savage and Massa
(1987)
for SiIV, CIV and NV and from
Jenkins (1978)
for OVI. The value for OVI is listed as a lower limit because the
measures only extend to |z|
| ||
Savage and Massa (1987) found that the explanation for the support and ionization of halo gas requires a blending of the ideas from the galactic fountain models and the photoionized halo models. In this situation SiIV would mostly be produced by photoionization while NV would mostly be produced by collisional ionization in cooling fountain gas. CIV therefore, probably represents an intermediate situation with important contributions from collisional ionization and photoionization.
If a galactic fountain flow actually exists, it seems reasonable to interpret the inflowing neutral gas at large |b| as gas associated with that flow. This interpretation is consistent with the abundance studies of Caldwell (1979) for HD93521 and Albert (1983) for TiII toward many stars. Hot (106 °K) upflowing disk gas would be expected to be devoid of grains because of rapid destruction during the heating of that gas. When this hot gas cools and returns to the plane, it is unlikely that conditions will be favorable for grain formation. Therefore, the returning gas should have nearly solar abundances in the gas phase as is observed. In a galactic fountain flow the relative amounts of gas residing in high ionization states (e.g. SiIV, CIV and NV) versus gas in low ionization states (e.g. SiII, CII and NII) will be influenced by the cooling rates of the gas through the various temperature regimes. The observation that the column density of weakly ionized gas exceeds that of the highly ionized gas by about a factor of ten seems compatible with the fact that gas rapidly cools from about 3 x 105 to 104 °K.
The gaseous galactic halo is a region of our galaxy that is now ideally suited for innovative observational and/or theoretical studies. In a brief period of time (about 10 years) the subject of the gaseous halo has moved from one where even the very existence of such a region was doubted by many to one which is now considered of vital importance for the overall regulation of the interstellar medium of the galactic disk (e.g. see Cox 1981 and Cowie 1987).