1. Modelling the Red Halos of Blue Compact Galaxies Zackrisson, E 1., Bergvall, N 1., Marquart, T 1., Mattsson, L 1. & Östlin, G. 2 1 Uppsala Astronomical Observatory, Sweden 2 Stockholm Observatory, Sweden Introduction to the Red Halo Problem Optical/near-IR broadband photometry of the faint halos surrounding Blue Compact Galaxies (BCGs) have revealed a very red spectral energy distribution (Bergvall & Östlin 2002, Bergvall et al. 2003), which cannot easily be reconciled with a metal-poor stellar population like that in the halo of the Milky Way. A similar problem has also noted for the halos detected around edge-on disk galaxies in stacked optical data from the Sloan Digital Sky Survey (Zibetti et al. 2004), suggesting that the red halo phenomenon may be common to several types of galaxies. What is the nature of these strange halos? Here, models of spectral evolution are used to test three different possible explanations for the red excess: High-metallicity stars, nebular emission and a stellar population with a very bottom-heavy initial mass function (IMF). A High Metallicity? Well… In Bergvall & Östlin (2002), high-metallicity stellar populations were suggested as an explanation for the red excess of BCG halos. In Fig. 1, we show that although this solution does reasonably well when confronted with the extended BCG halo data set of Bergvall et al. (2003), the metallicity of many of the halos has to be very high (solar or higher), which is a strange result given the low metallicity (~10% solar) of the gas in the central starburst of these objects. Also, this explanation fails to explain the halos of edge-on spirals (see dashed lines in Fig. 3b). In principle, the colours of BCG halos could be affected by nebular emission originating in an extended envelope of gas ionized by hot stars in the central starburst. Since photoionization models predict the spectrum of a complete Strömgren sphere to be very blue, this explanation for the red excess was dismissed by Bergvall & Östlin (2002). A problem with this argument is however that current halo observations do not probe the complete Strömgren sphere, but only a small part of the ionized region. For this reason, we have here used the photoionization model Cloudy (Ferland 1996) to predict colours for non-central lines-of-sight through spherical nebulae (Fig 2). Despite a substantial scatter, we find that even in this case, nebular emission produces colours much too blue to explain the red excess of BCG halos. A Bottom-Heavy Initial Mass Function? Yes?! In Fig. 3a, we show that stellar populations with a very bottom-heavy IMF (dN/dM  M -  with  =4.50, M=0.08—120 M solar ) can explain the BCG halo colours with low to intermediate stellar metallicities. Interestingly, a stellar population with the same IMF, age and metallicity also succeeds in explaining the halos of edge-on disks from the Sloan Digital Sky Survey (Fig 3b), suggesting that such strange halos may be common to galaxies of different types. References Bergvall, N & Östlin, G. 2004, A&A 347, 556 Bergvall, N., Marquart, T. Persson, C., Zackrisson, E.. & Östlin, G. 2003, In Multiwavelength Mapping of Galaxy Formation and Evolution, in press Ferland, G. J. 1996, HAZY, a brief introduction to Cloudy, University of Kentucky, Department of Physics and Astronomy Internal Report Fioc, M. & Rocca-Volmerange, B. 1999, astro-ph/9912179 Zibetti, S., White, S. D. M. & Brinkmann, J. 2004, MNRAS 347, 556 Fig 1. Colours of BCG halos (crosses) compared to the stellar population evolution (solid lines) predicted by PEGASE.2 (Fioc & Rocca- Volmerange 1999) for a Salpeter IMF at different metallicities. A star formation history typical of an early- type system (SFR  exp(-t/  ) with  =1 Gyr) has been assumed. Only the highest metallicities (red/black lines) provide a reasonable fit. Fig 2. Colours of BCG halos (crosses) compared to the colours predicted for various lines-of-sight through photoionized starburst envelopes (blue dots). Fig 3a. Colours of BCG halos (crosses) compared to PEGASE.2 predictions (solid lines) for the evolution of stellar populations with a bottom-heavy IMF. Fig 3b. Colours of the typical halo (cross) detected around edge-on disks in the SDSS, compared to the PEGASE.2 predictions for stellar populations with a bottom-heavy IMF (solid), as well as a Salpeter IMF (dashed). Nebular Emission? No!