Summary. (Also available in German and Spanish)
Subject of my Ph.D. thesis is the analysis and interpretation of the overall gas dynamics in the starburst galaxy NGC 1808, in comparison with other spiral galaxies in a similar evolutionary stage. Several ingredients of the interstellar medium have been investigated. The study includes line as well as continuum observations in various wavelength ranges, each providing insight into the physical conditions (and kinematics) of individual gas phases. In comparing and fitting together the results of these measurements -- together with other published observations -- a detailed picture of this relatively nearby galaxy and its environment has been derived. In the following we will summarize the results on particular topics and the outcoming global scenario.
The data consist of high-resolution observations in the 21-cm line of
neutral hydrogen (HI) and the Halpha line of ionized hydrogen, as well as
I-band continuum emission and, as a by-product of the HI line emission,
narrow-band 21-cm radio continuum maps.
The HI observations were carried out with three different configurations of
the Very Large Array (VLA) giving resolutions between 2 arcsec and 60 arcsec.
At the adopted distance of 10.9 Mpc for NGC 1808 one arcsecond corresponds
to 53 pc. The highest velocity resolution is 5.2 km/s.
An Halpha long-slit spectrum and two images (Halpha and I) were obtained
with the ESO New Technology Telescope (NTT) and have an angular resolution
of about 1 arcsec.
A group of nearby `active' (mostly starburst) galaxies has been selected for
comparison with NGC 1808. It consists of NGC 253, NGC 660, NGC 891, M 82,
NGC 3079, and NGC 4945. Apparently, all of them are in a similar evolutionary
stage, the so-called chimney-mode.
The following results have been obtained for NGC 1808:
---a) a warp of the outer HI gas layer,
---b) a bar of bright HII regions, whose kinematics are also visible in the HI velocity field,
---c) nuclear outflow of neutral gas into the halo of the galaxy,
---d) a fast-rotating nuclear ring of cold gas, and
---e) a complex central region consisting of various supernova remnants, very luminous HII regions, and an `active' nucleus.
a) The Warp. The large-scale HI distribution and kinematics in NGC 1808 can be modelled by a disk (inclination approx. 60 degrees) and warped outer spiral arms. The warp has been inferred from the shape of the global HI distribution and the velocity field, to which we applied a tilted-ring-model fit. This revealed a typical integral-sign warp starting beyond the disk, where the optical emission is fading. The outer spiral arms are seen more face-on than the disk, with a warp-angle of about 25 degrees. The HI extent (diameter < 12 arcmin) is about 1.5 times larger than the optical extent. The inclination-corrected rotation curve is flat in the outer regions indicating the presence of a dark halo. The total amount of HI gas in the disk is 7 x 10e8 solar masses and in the outer spirals about 10e9 solar masses. The total indicative mass has been estimated with 1.2 x 10e11 solar masses.
b) The Bar. High-resolution Halpha observations reveal the existence of a bar in the disk of NGC 1808 (PA = 155 degrees), about 15 degrees offset from the major axis. The bar has a length of about 6 kpc and rotates with a pattern speed of < 50 km/s/kpc. The bar-to-disk length ratio is about 0.3. The chain of HII regions confining the bar coincides with maxima in the HI gas distribution and further out connect to the outer spiral arms. The streaming of gas along the bar is clearly seen in the HI velocity field. The iso-velocity contours are nearly perpendicular to the leading side of the bar, where the gas is shocked, and very steep and elongated on the trailing side, where it is streaming inwards. The shock fronts can also be observed as broad dust lanes emerging near the nuclear region and curved along the leading sides of the bar. Inside corotation the gas accumulates at the inner Lindblad Resonances, which are found inside a radius of 1 kpc. The existence of two inner resonances might indicate the existence of a nuclear spiral. The outer Lindblad Resonance coincides with the outer spiral arms, if corotation is placed at about 5 kpc, which is significantly larger than the bar radius. The bar pattern speed is then reduced to 30 km/s/kpc. The locations of the inner and outer Lindblad Resonances have been obtained from the HI and Halpha rotation curves, respectively. An apparent inner spiral structure which dominates the disk seen on optical short exposures, does not show significant Halpha emission.
c) The Outflow. Radial dust filaments are seen in the northeastern part of the galaxy where they emerge from a well defined region of about 1.5 kpc into halo and are facing towards us. Their appearance suggests a cone like geometry with an opening angle of about 30 degrees and a scale height of at least 3 kpc. We determined a thickness of the filaments of about 100-200 pc. In the same area HI measurements have revealed a funnel-like region (opening angle about 50-60 degrees) of blue-shifted velocities, steepening towards the center as would be expected by the geometry of outflowing gas in a cone. Phillips (1993) found a bipolar outflow structure in Halpha with a similar opening angle as the dust cone (possible dust filaments in the southwestern part would be hidden by the emission in the disk). The different cone opening angles and flow velocities suggest that the neutral gas builds the outer cooled shell of a hot cavity where the ionized gas and the dust reside. The outflow velocity of the neutral gas is at least 150 km/s, whereas for the ionized gas Phillips determined much higher velocities of 400-700 km/s.
d) The Nuclear Ring. We have mapped the 21-cm absorption in the nuclear region of the starburst galaxy NGC 1808 using the VLA AnB-hybrid array with a resolution of approx. 5.2 arcsec (which corresponds to 275 pc). We find that the broad (velocity width approx. 380 km/s) absorption line seen at low spatial resolution is in fact a much narrower line which shifts its center velocity over the face of the source. The velocity center of the absorption changes from about +120 km/s relative to the systemic velocity in the NW to -145 km/s in the SE, following the rotation pattern in the outer regions of the galaxy. This is interpreted as a thick ring (torus) of cold, dense gas with a rotation velocity of about 250 km/s and radius 500 pc. From our data the ring is not expanding or infalling rapidly. We suggest that similar HI rings are also present in numerous other galaxies (e.g. all members of the comparison sample), for which the half-width of the unresolved, nuclear absorption line matches very closely the disk rotation velocity. Only in M 82 have the nuclear kinematics of the HI gas been resolved, also revealing a fast-rotating ring (Weliachev et al. 1984). The other galaxies of our sample show nuclear rings in other tracers. The gravitational mass required to explain this fast rotation in the center of NGC 1808 is a few times 10e9 solar masses. Apparently the distribution of mass goes as 1/r over at least the range from 500 pc to a few kpc.
e) The Nuclear Components. The nuclear region consists of several
groups of supernova remnants, numerous bright HII regions (`hot spots'), and
an `active' (starburst or Seyfert) nucleus, which is the strongest component
at all measured
frequencies. The central region is clearly dominated by starburst activity.
The star formation rate, SFR (M > 0.1 solar masses), has been estimated with
5.5 solar masses per year; the supernova (SN) rate is at least 0.3 SN per year.
Numerous compact components (< 50 pc) at 20-cm and 6-cm indicate
the sites of massive star formation. In between several `hot spots' are found
which have been ionized by the same short-lived, massive stars.
We propose a scenario for the evolution of the starburst galaxy NGC 1808 in which the above mentioned bar plays a crucial role. The formation of the bar has probably been stimulated by tidal interaction between NGC 1808 and its brightest companion NGC 1792. The warp of the outer gas layer is unlikely being caused by the interaction, as many warped galaxies are isolated, suggesting a self-maintaining mechanism (like e.g. a triaxial halo) for their origin. The barred potential in the disk clearly influences the gas dynamics of a galaxy. The accumulation of gas at the inner Lindblad Resonances (ILRs) is observed as a fast-rotating nuclear ring. The distribution of the `hot spots' also resembles a ring or nuclear spiral. The observed outflow of neutral hydrogen gas and dust from the central region into the halo of the NGC 1808 is then interpreted as a consequence of the central activity. The latter has been triggered and enhanced by the bar-induced mass transport into the nuclear region. We suggest that after the gas has been accumulated at the ILR(s) part of it reaches the central starburst region (maybe spiraling inward as indicated by the ILRs) where it will be heated and ejected approximately along the rotation axis of the galaxy due to the gas pressure of supernova remnants and winds. There are many indications that similar scenarios are also present in other far-infrared luminous galaxies like the ones discussed in the selected group.
Keywords. galaxies: individual (NGC 1808) - galaxies: active - galaxies: ISM - galaxies: kinematics and dynamics.
HI distribution of NGC 1808
HI mean velocity field of NGC 1808 (masked)
Bärbel Koribalski * Baerbel.Koribalski @ csiro.au