During the last decades the formation of massive stars in galaxies of different Hubble-Types has been studied in great detail. For spirals this process is considered to be almost exclusively concentrated to the disk as the efficiency with which gas is converted into stars is highest there.
On the other hand, the existence of a young and massive OB-star population in the halo of the Milky Way and other spiral galaxies, such as NGC253, is widely accepted.
However, as these stars are detected at large distances (~15kpc) above the star forming disk, their creation and origin are rather poorly constrained.
In order to clarify the origin of this young stellar component in the halo different scenarios have been proposed, such as ejection from the disk as a consequence of supernova (SN) explosions, ejection from stellar clusters as a result of gravitational encounters, and star formation in the halo itself
(Keenan 1992, Ferguson 2002).
Estimates of the distance a star could travel at a given speed through the halo during its lifetime, reveals that in situ star formation in the halo is the most likely scenario for a large fraction of the studied sample of halo stars.
If proper motions of these stars are reasonably low, it should be possible in the future to detect their stellar cradles, faint gaseous envelopes in the vicinity of these stars, and thus to further strengthen extraplanar star formation (ESF).
Although the mechanism that triggers ESF still remains unclear, especially in view of low gas densities, we have increasing evidence that star formation occurs at rather unusual sites, such as in a galaxy halo.
In the following, we investigate the possibility of ESF and its triggering mechanism by analyzing first VLT-data of compact extraplanar gas clouds with embedded stellar sources located in the disk-halo interface of the edge-on galaxy NGC55.
Extraplanar HII-regions in NGC55
There are fairly compact and isolated objects visible on Halpha-imaging data for NGC55
(Ferguson et al. 1996) which are located at distances of up to 1.5kpc above the disk of this galaxy. From a morphological point of view, these objects appear very much like a small scale disk HII-region with embedded clusters of massive stars formed recently.
Interestingly, similar regions are also discernable in other well known edge-on galaxies, such as NGC891, NGC3628, or NGC5775.
What makes the barred spiral galaxy NGC55, a member of the Sculptor Group, an ideal target to study ESF is its proximity of only 1.6Mpc and its high inclination of i=80
o. Moreover, this galaxy reveals violent ongoing star formation in the center and at least two prominent extraplanar HII-region (EHR) candidates. All these features can be seen very nicely from
Fig. 6.1 where the VLT Halpha-image obtained with UT1+FORS1 is presented.
A huge curved filament of gas and dust, anchored in the disk, is protruding out of the image plane, apparently pointing towards R.A.(2000): 00h15m07s and Dec.(2000): -39d12m00s. Of particular interest are the two objects marked by arrows, whose magnifications are shown in
Fig. 6.2.
EHR_1 in the north has a diameter of 17pc and is located 0.8kpc above the disk, whereas EHR_2 in the southern halo reveals a projected distance of 1.5kpc and spans 22pc in diameter.
If effects along the line of sight are negligible, EHR_1 is located within an expanding oxygen-bright SN shell that was detected 20 years ago by
Graham & Lawrie (1982).
It would be interesting to learn from similar observations in other galaxies if these extraplanar regions predominantly occur at points where such shells, created by OB stars and SNe, intersect and the gas is piled
up. At least the compressed gas at R.A.(2000): 00h15m07s and Dec.(2000): -39d11m00s is in favour of this idea.
However, the most important immediate result is provided by optical multi-object-spectroscopy (MOS) and concerns the detection of spatially concentrated continuum emission, which originates within the more diffuse body of the extraplanar objects
(Fig. 6.3).
The morphology of the continuum and the nebular emission-line distribution is direct evidence for stellar sources responsible for the excitation of these regions. Correspondingly, the flux-calibrated and background-subtracted spectra, integrated along their total spatial extent are very similar to low excitation HII-regions
(Fig.
6.4a
,
b
,
c
,
d
)
.
For EHR_1 in the northern halo, continuum emission is very weak and can hardly be seen in the presented spectrum. However, it is clearly visible in Fig. 6.3. Figs. 6.4c and 6.4d also show the spectrum of EHR_2 (southern halo) which reveals a continuum much more prominent than that found in EHR_1.
A comparison with CLOUDY model simulations reveals that the ionization mechanism of these compact objects is most likely photoionization by late OB stars (O9.5 to B0). Further analysis of diagnostic diagrams unambiguously confirms the HII-region character.
Constraining the origin of the EHRs
The existence of HII-regions in the halo raises immediately the question whether these objects originated from the prominent extraplanar gas of this galaxy or have just been expelled from the disk into the halo.
Ejection from the disk is quickly ruled out by hydrodynamical considerations regarding the enormous drag, the gas phase of these regions encountered on its way out of the disk into the halo.
Even a small amount of interstellar matter sitting along the path would lead to a separation of cloud and embedded stars due to the enormous difference in impact parameters cloud vs. cloud as compared to stars vs. cloud.
Therefore, we conclude that these objects must have formed within the halo.
In addition, knowledge of the gas phase abundances also helps to distinguish between different creation mechanisms of the extraplanar ionized regions. Rather low metallicities compared to the disk abundances would indicate that these regions have formed from almost pristine local halo material. Relatively high abundances would restrict the origin of the clouds to material processed in star-forming regions of the disk.
We therefore determined the element abundances of both EHRs and compared them to those measured in the disk using two independent methods (R23 and the nebular abundance tool (NAT), see
Tüllmann et al. 2003 for details).
The results are shown in
Table 1.
Element abundances for the HII-region and the EHRs as calculated by NAT for gas temperatures of 11500K. Values in brackets were derived with the empirical R23-calibration. Solar abundances are compiled from the most recent data including
Christensen-Dalsgaard (1998) and
Grevesse & Sauval (1998). The average metallicity Z/Z_solar has been calculated from oxygen abundances as this element is the most abundant and efficient coolant.
A comparison between the average metal abundance of the central disk HII-region of NGC55 (45% Z_solar) and both EHRs reveals substantially lower [O/H] abundances of about 10% Z_solar and thus independently also supports the ESF scenario.
With metal abundances derived this way, we can visualise for the first time the strong differences in the metal content along the minor axis of this galaxy.
Fig. 6.5 plots oxygen abundance as a function of z, the distance along the minor axis of NGC55. The open symbol represents the averaged oxygen abundance for the disk, whereas the error bar represents data published by other authors. The data-point labelled "HR" has been slightly shifted along z to separate error bars. The one named "DIG" represents a special component of the ISM at intermediate z-distance (see below). From this it is obvious that the gas phase of oxygen is less abundant in the halo by about a factor of 4.
In order to reach a better coverage of the oxygen abundance along the minor axis,
Fig. 6.5 also
displays the metal abundance of the Diffuse Ionized Gas (DIG). This gas phase is pushed into the halo of a galaxy by multiple SNe where it is visible as a diffuse extended Halpha-emitting gaseous layer surrounding the disk. The ionization of the DIG is maintained most likely by photoionization from stars located in the star forming disk below. As the DIG is no longer involved in star formation, it is expected to be also a good tracer of the metal content of the halo gas. The interested reader is referred to
Dettmar (1992) for a comprehensive review.
The global picture
Two important questions directly emanate from the ESF hypothesis: (a) how did gas reach the halo in a quantity to cool, collapse, and form neutral, dense clouds from which new stars were born, and (b) what triggered the collapse to actually form those stars?
The most simple and natural explanation is to assume that clustered SNe during an early burst of star formation pushed a significant amount of ionized material into the halo.
After star formation stopped the extraplanar gas had time to cool, collapse, and form dense molecular clouds. These molecular gas clouds, out of which EHRs have formed, can survive and collapse only in the period between two successive bursts of star formation.
Since both EHRs are located above the central part of NGC55, it appears likely, that their formation was triggered by star formation activity in the disk below. In this global picture star formation in the disk could stimulate as well as terminate the creation of EHRs.
Future work will test the ESF scenario for a larger sample of galaxies, investigate initial formation conditions for EHRs, and check if the central stars can separate within their lifetime far enough from their cradles and contribute to the observed stellar halo population.
