Proceedings of the Workshop
"The Magellanic Clouds and Other Dwarf Galaxies"
of the Bonn/Bochum-Graduiertenkolleg

Chemodynamical Evolution of Dwarf Galaxies

Andreas Rieschick and Gerhard Hensler

Institut für Theoretische Physik und Astrophysik, Universität Kiel

Received 28th August 1998
Abstract. Main criteria for comparing observations of dwarf galaxies with evolutionary simulations are the metal abundances and adundance ratios, especially those for oxygen and N/O. Models of chemical evolution usually start with a high N/O ratio at low O abundance and reproduce the observed N/O-O peculiarities by the application of multiple starbursts. Their galactic winds are invoked to reduce O selectively. Our chemodynamical models of dwarf galaxies, however, demonstrate that strong evaporation of clouds in hot supernova-expelled gas leads to an almost perfect mixing of the interstellar medium and its element abundances. Hot, metal-enriched gas expands over the dwarf galaxies and condensates preferably in the outer parts. Gas clouds falling back into the inner regions therefore show almost homogeneous element abundances. First results indicate that by taking into consideration both dynamical and chemical evolution processes in a self-consistent way, we can explain the observed N/O-O values without the assumption of artificial starbursts but by using recent stellar yields which provide secondary nitrogen production also in massive stars. Here we present some features of the dynamical evolution of a simulated 109 Msun dwarf irregular galaxy.

1. Introduction: The Chemodynamical Treatment

For systems and sites of low potential energy we know from empirical studies and theoretical investigations that the ISM is on average held in balance by counteracting processes like heating and cooling, turbulence and dissipation (Burkert & Hensler 1989; Hensler et al. 1998b). Since these processes are non-linearly coupled, the effect of neglecting one of them will alter the evolution completely. A number of studies of self-regulated SF exist with particular attention to the influences of stellar radiation, supernova explosions, and the evaporation/condensation balance between the two chemically and dynamically distinct gas phases, the cloudy medium (CM) and the hot intercloud medium (ICM) (Franco & Cox 1983; Ikeuchi et al. 1984; McKee 1989; Bertoldi & McKee 1995; Köppen et al. 1995,1998). The evolution of DGs is self-regulated and determined by large-scale outflows (Hensler et al. 1993,1998a).

To approach global models of galaxy evolution which yield the structural differences and details, adequate treatment of the dynamics of stellar and gaseous components is essential. At least the following processes have to be taken into account: SN, SF, heating, cooling, stellar mass loss, condensation and evaporation. This includes the treatment of the multi-phase character of the ISM as well as the star-gas interactions and phase transitions. Since gas and stars evolve dynamically, and because several processes both depend on their metallicities and also influence the element abundances in each component, these models are called chemodynamical (cd).

It must be emphasized, however, that the number of free parameters in the cd scheme is not large but actually smaller than in multi-zone models. The allowed ranges of parameter values are strongly constrained, either because they are theoretically evaluated (like e.g. evaporation and condensation), empirically determined (like e.g. stellar winds), or because they force self-regulation in a way that is independent of the parameterization. Because of limited space we refer the interested reader to more comprehensive descriptions of the cd treatment and to different applications (non-rotating galaxies: Theis et al. 1992; Hensler et al. 1993,1998a; vertical settling of the galactic disk: Burkert et al. 1992; disk galaxies: Samland & Hensler 1996; the MWG: Samland et al. 1997 (SHT97); Samland 1998; dwarf galaxies: Hensler & Rieschick 1998).

2. The Dynamical Evolution

The models calculated with our chemodynamical evolution code CoDEx show a collapse into a system with a thick gas disk being in rotational equilibrium. Timescale and strength of the collapse strongly depend on the initial model parameters. These are mainly the total mass of the system, the total angular momentum and in this case especially the amount of dark matter. During the collapse phase the star formation is concentrated to the center of the galaxy. Phases of gas infall (by gravitational forces) and SN driven outflow alternate for about 1.0 to 3.0 Gyr. Though the models show strong star formation, real star bursts cannot be produced intrinsically with our 2d chemodynamical models, other sources such as gas infall or gravitational disturbances would be required.

At the end of the collapse phase an equilibrium is reached, the further evolution continues in a rather smooth way. Large star forming regions and regions with low activity can be found all over the thick disk. Spatial self-propagation of star formation leading to a kind of star forming waves moving through the galactic disk can be found. This feature depends also strongly on the initial model parameters. See Rieschick & Hensler (1998) for details.

Figure 1 presents some snapshots of the evolution of a simulated 109 Msun dwarf irregular galaxy. Mass densities of high mass stars and low mass stars are shown for six points in evolution time. The left edge is the rotational axis of the 2d cylinder symmetrical model. The horizontal line in the center separating HMS and LMS densities represents the galactic plane. The HMS density is directly correlated with the current star formation activity due to the short stellar lifetimes, while the LMS density includes the accumulated mass of various star forming periods. Therefore the LMS density increases very smoothly with time, the HMS density in the contrary allows to identify the individual star forming regions. The whole evolution will be available as online movie (Rieschick 1998).

[Click here to see Fig. 1!]

3. Mixing Processes

The low gravitational potential of low-mass galaxies leads to vehement expansion of hot SN II gas, which is unable to cool by radiation on a short timescale due to its low density. Actually also the imbedded cloud matter is evaporated and carried outside the galaxy. Afterwards the blown away metal-enriched gas condenses and falls back into the galaxy as clouds again. This cycle of evaporation, outflow, condensation and infall leads to an almost perfect mixing of the ISM (except of the very inner parts of the galaxy with the strongest SN and PN activity). Therefore the often used assumption of selective outflow of O is incompatible with our models.

Since N and O are polluting the different phases of the ISM separately, their simultaneous existence in H II regions, which represents the ionized CM, can only result from phase transitions between CM and ICM. Kobulnicky (1998) reports the non-detection of any sizable O, N, and He anomalies from H II regions in the vicinity of young star clusters in SBDGs with one exception, NGC 5253, which reveals a central N overabundance. In NGC 1569, that has recently formed two super-star clusters, Kobulnicky (1998) finds constant O and N/O values over a radial extent of more than 400 pc with a scatter of N/O by only 0.2 dex, while self-enrichment tracks extent over one dex in N/O (Pilyugin 1992) and dispersal distances are much less. The same fact holds in I Zw 18 (Izotov 1998).

4. Conclusion

Our cd models of dIrrs show that a strong mixing of the gas phases prevents a selective element depletion. The self-consistent cd models evolve with local variations but globally moderate SF activity and therefore typically for a large fraction of dIrrs.

The observed N and O abundances in the H II regions can be strikingly attained, if N is produced as a secondary element also in HMS. If the cd prescription is applied, important physical processes like large-scale coupling of different galactic regions by dynamical interactions as well as small-scale mixing effects between the gas phases are adequately taken into account, and this substantially fixes the element abundances. Abundances can serve as reliable diagnostic tools of galaxy evolution and provide a chance to deconvolve it in detail, if studies couple the above-mentioned gas processes with the dynamics of gas and stars as well as with their mutual interactions. Please refer to Hensler & Rieschick (1998) for more details about mixing processes in cd models and their influence on abundance ratios.

Acknowledgments. We gratefully acknowledge cooperations and discussions with J. Köppen, M. Samland and Ch. Theis. This work is supported by the Deutsche Forschungsgemeinschaft (DFG) under grant no. He 1487/5-3 (A.R.).

References


Links (back/forward) to:

First version: 29thAugust,1998
Last update: 28thSeptember,1998

Jochen M. Braun   &   Tom Richtler
 (E-Mail: jbraun|richtler@astro.uni-bonn.de)