4.4.2 The Age of the Underlying Population
To detect underlying old populations it is essential to reach faint isophotal levels, and moreover to take into account the possible contamination from gas ionised by UV-photons leaking out from the central starburst. Thuan (1983) attempted to constrain the star formation histories by NIR aperture photometry and concluded that BCGs were old. However, this was based on central colours and old models, and the data do no longer allow for an unambiguous conclusion. Surface photometry or resolved star photometry would be the preferred method to investigate ages of BCGs. With modern detectors, it is feasible to make surface photometry of metal-poor BCGs at faint levels in the near IR. Bergvall and Östlin (1999) found a very clear signal in V - J of old stars formed on rather short timescales in the haloes of some luminous BCGs, while e.g. B - V remained inconclusive, illustrating the power of NIR observations.
Loose and Thuan (1986b) and Kunth et al. (1988) found B - R colours to redden with increasing radius, suggesting the presence of underlying old populations. Optical surface photometry has continued to unveil red haloes in most BCGs studied in detail (Papaderos et al. 1996, Doublier et al. 1997, 1999; Marlowe et al. 1999) and Telles and Terlevich (1997) found the underlying colours to be consistent with those of blue LSBGs and amorphous galaxies. However ages and star formation histories are not yet well constrained, since model predictions are degenerate. In many cases, the most influential parameter is the assumed shape of the star formation history which is what one wants to determine ultimately. Depending on how deep the halo is to be probed, and what pass band is used, different ages are obtained for the underlying population, indicating that composite stellar populations are present, or that nebular emission contributes to the colours, or both. For an example of colour profiles revealing a redder halo, see Fig. 8.
Nevertheless, all these studies demonstrate that the majority of BCGs are not young. Cases were no underlying populations have yet been found exist, but as more detailed studies are performed, old populations turn up in most young galaxy candidates, as e.g. in the previous young galaxy candidates Pox186 (Kunth et al. 1988; Doublier et al. 1999) and ESO 400-G43 (Bergvall and Jörsäter 1988; Bergvall and Östlin 1999).