Right from the very early discovery of RR Lyrae towards the centre of the Galaxy (Baade 1951), perhaps the most common statement found in the literature addressing the Galactic bulge stellar populations is: The Milky Way bulge stellar population is predominantly old. The reasons for such a statement are indeed very well funded and are described in this section.
Without a doubt, deep photometric observations towards low extinction regions of the Bulge provided a defining view on the age of its stellar population thanks to the possibility of constructing colour-magnitude diagrams that reached the turn-off position - a useful indicator for the mean age of a given stellar population. However, the effects of differential reddening and, in particular, the uncertainty on the distance modulus of the stars towards a given line of sight does not allow the derivation of an absolute age estimation for the Bulge using the turn-off technique. To overcome these issues, Ortolani et al (1995) adopted a differential method, based on the position of red-clump stars in the luminosity function. The very small dependence on age and metallicity for the mean magnitude of the red-clump allowed Ortolani et al (1995) to match its position in the luminosity function to that of an old stellar population, the globular cluster NGC 6528, finding a remarkable agreement between the relative positions of the turn-off. This result provided strong support for the view of the Galactic Bulge being formed in a time scale shorter than 1 Gyr, thus making it as old as the globular cluster population.
Dedicated photometric studies spread across other regions of the Bulge, and based on similar techniques further strengthen the conclusion of a stellar population with a mean age of ∼ 10 Gyr, particularly setting a lower limit on ages higher than 5 Gyr (Zoccali et al 2003, Valenti et al 2013, Clarkson et al 2011). The major issue with these kind of studies is how to deal with the contamination of the foreground disc, in particular with the main sequence of the disc which lies right on top of the Bulge turn-off. Statistical decontamination methods, for example using disc control fields (Zoccali et al 2003, Valenti et al 2013), have been used to eliminate foreground stars to some extent. However, the contamination of foreground stars coupled with uncertainties in differential reddening and metallicity distribution effects lead to uncertainties on age determination via the analysis of color-magnitude diagrams that remain of the order of 2 Gyr.
Figure 2. Left panel: The metallicity of the microlensed dwarfs of the Bulge as a function of their ages taken from Bensby et al (2013a). Right panel: The color magnitude diagram for proper motion-selected Bulge objects from Clarkson et al (2008), using similar mean proper motion criteria to Kuijken and Rich (2002) but with a 6σ detection requirement imposed. A set of isochrones with different metallicities and ages is overplotted to the color magnitude diagram. An alpha-enhanced, solar-metallicity isochrone at 11 Gyr represents the median sequence well above the turn-off. Also shown in the figure are sequences at metallicity [Fe/H] = (-1.009, -0.226, +0.491) and ages (8, 10, 14) Gyr. [Left panel adapted from Fig. 15 in Bensby et al 2013, ©ESO reproduced with permission. Right panel adapted from Fig. 20 in Clarkson et al 2008, ©AAS reproduced with permission.]
Recently, Clarkson et al (2008) provided what is perhaps the cleanest, and thus most accurate, age determination of those studies based on the analysis of a colour-magnitude diagram (Fig. 2). Clarkson et al (208) were able to decontaminate the turn-off position of the Bulge population from that of the nearby disc by using a mean proper motion criteria as the one shown by Kuijken and Rich (2002) with a set of ACS WFC on HST observations taken over 123-orbit HST integrations in the SWEEPS field (l,b) = (1.25, -2.65). The study of Clarkson et al (2011) later concluded that a fraction of only up to a 3.5% of the Bulge stars can be younger than 5 Gyr. Therefore, this study is in agreement with most of other studies where the Bulge population is found to be dominantly old.
A large percentage of young stars is certainly not expected in a bulge that originates purely from early dissipation or merging processes. These events would take place at early times in the evolution of the galaxy and would occur rapidly. Although some intermediate-age stars could later be added to the bulge by diffusion from the inner disc, the bulk of the bulge stellar population would be old. On the other hand, a significant population of young stars could be found − and might be actually expected − in bulges formed via disc instabilities. So if the Milky Way bulge is certainly dominated by a component formed via disc instability, as seen by its structural properties, then where are those young stars? It is only very recently that a possible answer to this question has been brought to the table, and it was using a completely different approach: the high-resolution spectroscopy of Bulge microlensed stars. The microlensing event, which produces the brightness magnification of a dwarf star in the bulge due to the passing of a foreground lens star, provides a unique opportunity to obtain high resolution spectra of this otherwise unreachable target. Bensby et al (2013a) have collected enough micro-lensed dwarfs to investigate the overall metallicity distribution and also their age. While the metallicity distribution of the microlensed dwarfs has been found to be in good agreement with that of the bulge giant stars, their age distribution presented a significant number of young stars: nearly 22% of the micro-lensed dwarfs were found to be younger than 5 Gyr. As the number of analysed micro-lensed stars increases, these findings will be further confirmed or disproved with better statistics in order to refine the actual percentages of young and old stars found in the bulge. Particularly, notice the disagreement between the 22% of stars younger than 5 Gyr found in the microlensing sample and the corresponding 3.5% fraction of young stars given in the work of Clarkson et al (2011).