9 Summary

Using the VLT with FORS1 and ISAAC, we have resolved stars in the halo and in the diffuse north-eastern shell of the closest giant elliptical galaxy, NGC 5128. With the U, V and $K_{\rm s}$-band photometry we probe different stellar populations. The U filter, in particular, is sensitive to the youngest stars. It has revealed the striking difference in the star formation history between the the inner halo field (Field 2) and outer halo field (Field 1), the latter coinciding with the diffuse shell that is presumably the signature of a recent merger.

In Field 1, stars as young as 10 Myr are present, while there are no stars younger than at least 40 Myr in Field 2. Thanks to the high quality of our UV photometry, we also detect the gap between the main sequence and the blue core-helium burning supergiants. The foreground and background contamination has been corrected using the Besançon Galaxy model and our own color-color diagram.

The recent star formation in Field 1 is approximately aligned with the direction of the jet coming from the center of the AGN. However, there is also a chain of young blue stars offset from the jet direction that is aligned with the edge of the molecular clouds. The presence of the atomic and molecular gas (Schiminovich et al. 1994; Charmandaris et al. 2000) and their association with the diffuse stellar shell is compatible with the dynamical scenario of phase-wrapping following the merger of a smaller gas-rich galaxy with NGC 5128 (Quinn 1984). In this scenario, the interaction of the material coming from the AGN in the center of the galaxy with the material left from the merger stimulated the star formation in the halo, $\sim$15 kpc away from the galactic centre. The metallicity of the newly formed stars gives an upper limit on the abundance of the gas of the accreted satellite with the value of Z=0.004, a value typical of the SMC.

While the UV CMDs probed the youngest stellar populations, the combination of optical and near-IR filters is most sensitive to old and intermediate-age populations. In our VK CMDs we observe a very wide and extended giant branch. The width indicates the presence of Population II stars with metallicities ranging from -2 to -0.7 dex. Unfortunately, our V-band images are not deep enough to detect more metal-rich stars, but the large number of stars detected on K-band images without a counterpart in V-band suggests an even redder and more metal-rich RGB population, in agreement with the previous HST studies (Soria et al. 1996; Harris et al. 1999; Harris & Harris 2000).

In contrast with the previous two studies, we find a very extended giant branch up to $M_{\rm bol}=-5$, which reveals the presence of the intermediate-age AGB population. Our extensive crowding experiments and theoretical predictions of the amount of blending (Renzini 1998) demonstrate that blending due to crowding is not significant enough to mimic this population in the outer field (Field 1). In the inner field, the extent and the properties of the AGB population will be assessed through the study of the long period variables that have been detected in both fields (Rejkuba et al. in preparation).

Acknowledgements

We thank Manuela Zoccali for help with DAOPHOT and ALLFRAME and Joel Vernet for help with IDL. We thank Francesco Ferraro who kindly provided the tables of VK globular cluster fiducial RGB sequences in electronic form. Felipe Barrientos kindly provided the U-V vs. V-K colors for galaxies of different morphological types and redshifts. MR acknowledges ESO studentship programme. TRB is grateful to the Australian Research Council and P. Universidad Católica for financial support and to the DITAC International Science & Technology Program. This work was supported by NASA Grant GO-07874.01-96A and by the Chilean Fondecyt No. 01990440 and 799884. Finally, we are grateful to the referee Mario Mateo for his helpful and thorough report.