6 Jet aligned star formation in the halo

Figure 15: The spatial distribution of the bluest (U-V)<-0.75 (left panels) and the reddest (U-V)>2 (right panels) stars in Field 1 (top panels) and 2 (bottom panels) for stars with magnitudes above the 50% completeness limit (V<25 and U<25). The north is up and east is to the left.

The chains of bright blue compact objects and optical filaments, first noticed by Blanco et al. (1975), were recognized as young blue supergiants and OB associations by Graham & Price (1981) and Graham (1998). Our Field 1, situated $\sim$14 kpc away from the center of the galaxy and coinciding with the NE radio lobe, contains part of this "string of blue knots and filaments.''

In Fig. 15 we show the spatial distribution of the bluest (left) and the reddest (right) stars in both our fields. The red stars, belonging to the foreground Galactic population, are uniformly distributed in both fields. Field 2 has almost no very young blue stars, while the majority of the young population in Field 1 is aligned with the jet direction coming from the nucleus of the galaxy. The vertical structure to the right of the jet is located at the leftmost edge of a large H I cloud found by Schiminovich et al. (1994). In the northern part of this field, a CO molecular cloud was recently discovered by Charmandaris et al. (2000).

Mould et al. (2000), Graham (1998) and Fasset & Graham (2000) report the presence of collimated star formation in the halo of NGC 5128. They favor the scenario in which a past interaction between the radio jet (Morganti et al. 1999) and the H I cloud complex (Schiminovich et al. 1994) is responsible for the star formation.

Finding active star formation far out in the halo of an elliptical galaxy is unusual. Understanding the triggering mechanism and the origin of the star formation, as well as the nature of the underlying ionized gas (Graham 1998; Morganti et al. 1991,1992), is important for explaining the filamentary emission and jet-aligned structures observed in other galaxies (e.g., in NGC 1275: Lazareff et al. 1989; Sabra et al. 2000; in M87: Gavazzi et al. 2000; and in other more distant radio sources: Rees 1989, Best et al. 1996, De Breuck et al. 1999).

The alignment of the luminous blue stars with the radio axis of NGC 5128 (Fig. 15) may shed some light on the debated issue of the observed alignment of the radio and optical structures in high-zradio galaxies (Rees 1989; Best et al. 1996). One of the proposed alignment models is based on the assumption that the action of the jets tunneling through the galactic medium affects the gas via shock heating, thereby inducing vigorous star formation (Begelman & Cioffi 1989; Rees 1989; Daly 1990). In particular, as Rees (1989) discussed in detail, the stars form from the cool-phase gas (with $T\leq10^4$ K) in clouds massive or dense enough to be Jeans-unstable.

This mechanism may work well for massive high-z radio galaxies, in which most of the baryonic mass would still be gaseous. In the present-day giant ellipticals, on the other hand, usually little gas is left to support strong star formation. However, probably due to previous accretion of a gas-rich galaxy, NGC 5128 has substantial cool, dense clouds of molecular CO and atomic H I gas in its halo (Schiminovich et al. 1994; Charmandaris et al. 2000). We note that these clouds could form the clumpy ISM/IGM necessary for the "bursting bubble'' model proposed by Morganti et al. (1999).

We have shown clear evidence for blue stars as young as $10\times10^6$ yr (Fig. 10). We also see underlying emission, particularly in our U-band images, which probably comes from the [OII] 3727 emission line. High-resolution spectra with good signal-to-noise ratios are needed to decide whether this emission is due to gas being photoionized by the newly formed stars or being photoionized by particles coming directly from the nucleus.

Other low-redshift galaxies also show filamentary ionized gas structures in their halos. In the case of M 87 (Gavazzi et al. 2000), the filament of ionized gas in the NE part of the halo coincides with the Eastern radio-lobe, in much the same way as in the much closer NGC 5128. It would be interesting to search for young stars associated with that filament too.