All Figures

	Figure 1: Colour-colour diagram for the NGC 5846 globular cluster system. The solid and dashed line mark the 15 Gyr and 2 Gyr isochrone respectively following the Bruzual & Charlot SSP model (Bruzual & Charlot 2000).
Open with DEXTER


In the text

	Figure 2: Comparison between the SSP model isochrones given by Bruzual & Charlot (2000) ( upper panel), by Vazdekis (1999) ( middle panel) and Maraston (2001) ( lower panel). Different symbols mark the 1 Gyr (solid circles), 3 Gyr (open squares), 5 Gyr (open stars) and 15 Gyr (open triangles) isochrones. The solid line at (V-K)=2.7 marks boundary between blue and red objects (see Sect. 2.4).
Open with DEXTER


In the text

	Figure 3: Least square fit of model isochrones given by Bruzual & Charlot (2000) ( left), Vazdekis (1999) ( center) and Maraston (2001) ( right). Various isochrones are marked with solid squares (1 Gyr), open stars (3 Gyr), solid circles (5 Gyr), solid triangles (10 Gyr) and asterisks (15 Gyr).
Open with DEXTER


In the text

	Figure 4: Cumulative age distribution in the globular cluster systems of NGC 5846, NGC 4365, NGC 7192, NGC 3115, NGC 4478 and M 87.
Open with DEXTER


In the text

Figure 5: Self-consistency test: age distributions of individual model systems compared to the complete model set. The result of the 2 test are shown as -contours. For this test we use the model of a purely 15 Gyr old population ( left), a  mixture ( middle) and a purely 3 Gyr old population ( right). As expected the result for the purely old population is much more uncertain than for the other two, allowing as best fitting model also mixed populations. The upper plots show the result for the BC00 model and the lower plots the VA99 models. The different spread of the isochrones (see Fig. 2) results in different uncertainties.

Open with DEXTER


In the text

	Figure 6: Comparison of colour-colour diagrams for the globular cluster systems and the HDF-S. Globular clusters are marked by filled circles and HDF-S objects by open squares. The solid lines mark the 1 Gyr and 15 Gyr isochrone respectively, as given by Bruzual & Charlot (2000). The box marks the colour range used for determining the cumulative age distribution.
Open with DEXTER


In the text

Figure 7: Colour-colour diagram for the HDF-S sample (K>21.5) before (triangles) and after (cross) applying shape parameter for selection. For selection the ellipticity (<0.2, left panel) and the star/galaxy classifier (>0.8, right panel) have been used. For final correction we choose a combination of photometric error, limiting magnitude and classifier. As in Fig. 6 the box indicates the colour limits.

Open with DEXTER


In the text

Figure 8: Cumulative age distribution in the HDF-S sample (BC00) normalised to 1 arcmin2 ( left: absolute, right: relative). The solid line marks the result if only a error cut of 0.15 mag for all filters is applied. Applying an additional ellipticity limit (<0.2) results in a age distribution indicated by the dashed line. Using a combination of error cut and star/galaxy classifier (>0.8) given by SExtractor (Bertin & Arnout 1996) most background galaxies can be rejected from the data set (dotted-dashed line).

Open with DEXTER


In the text

Figure 9: Cumulative age distribution in NGC 5846. The left panel shows the cumulative age distribution before (triangles, dotted line) and after (open squares, solid line) correction for contamination applying an error cut as well as a selection by star/galaxy classifier. In the right panel the distribution has been normalised to the total number of objects in the sample after correction (125 objects within the colour limits). The age dating was done following the SSP models by Bruzual & Charlot.

Open with DEXTER


In the text

	Figure 10: Cumulative age distribution (using BC00) in the globular cluster systems observed so far. The left panel shows the relative distribution without background correction. The correction applied in the right panel follows the procedure given in the previous section.
Open with DEXTER


In the text

	Figure 11: Cumulative age distribution for modeled samples, consisting of a different number of objects but assuming a constant mixture of old and intermediate age objects (50%:50%). The results are shown for the BC00 ( left) and VA99 ( right) SSP models respectively. Each panel shows the age distribution for two out of six different runs of modeling.
Open with DEXTER


In the text

	Figure 12: Cumulative age distribution (using BC00) derived for models of 10, 60 and 120 objects respectively. To show the importance of large number of objects we show the standard deviation of the counting rate. The models were created for a purely 15 Gyr old population (solid circles) and a 50%:50% mix () of 15 and 3 Gyr old objects.
Open with DEXTER


In the text

Figure 13: Examples of the reduced 2 test (see Sect. 3.2) for different sample sizes (10, 60, 120 and 200 objects). The age distribution for a mixture of 50 15 Gyr and 50 3 Gyr old clusters (see Fig. 11) with the complete set of models are compared. On top of each plot the number of old and intermediate objects within the "red'' population is given. We compare the results for six separate runs of modeling.

Open with DEXTER


In the text

	Figure 14: Cumulative age distribution in the globular cluster systems using BC00 ( left) and VA99 ( right) SSP model isochrones. The samples are corrected for background contamination following the procedure given in Sect. 4.1. The horizontal line marks the 50 level and the vertical line its intersection with the derived age distribution.
Open with DEXTER


In the text

Figure 15: We compare the 50-age calculated by a best fit of the cumulative age distribution for observed and simulated systems using BC00 ( left) and VA99 ( right) SSP model isochrones. The x-axis gives the amount of 15 Gyr old objects within the model and the label on the right hand side the age of the intermediate age sub-population. The larger spread in (V-I) for VA99 isochrones results in a less reliable result for smaller samples. All systems have been corrected for background contamination.

Open with DEXTER


In the text

Figure 16: Left panel: colour-colour diagram for globular clusters in 4 different galaxies. The model grid follows the Maraston (2001) SSP model isochrones (ages between 3 and 15 Gyr following the age-arrow). The metallicity ([Fe/H] between -2.25 and +0.35) increases in direction of the metallicity arrow. Right panel: comparison of index measurements with the models by Thomas et al. (2003) (age range: 3-15 Gyr, metallicity range:  -2.25-+0.67). Objects with spectroscopic ages below 5 Gyr are shown by solid circles.

Open with DEXTER


In the text

Figure 17: 2 test result for NGC 5846 ( upper) and NGC 4365 ( lower). The different levels represent the reduced 2 of the comparison between the age distribution in the observed systems and the various models. The correction for contamination following the procedure given in Sect. 4.1 has been applied. The left panels give the result following the BC00 models and the right panels following the VA99 models.

Open with DEXTER


In the text

	Figure 18: Result of the 2 test comparing the age distribution in NGC 5846 with simulated systems assuming the old population being 10 Gyr old. As clearly seen the best fitting model still contains a much smaller, but still significant number of intermediate age objects.
Open with DEXTER


In the text

	Figure 19: 2 test result for M 87 ( upper) and NGC 7192 ( lower). As in Fig. 17 the results for BC00 ( left panels) and VA99 ( right panels) are shown.
Open with DEXTER


In the text

Figure 20: Age resolution for SSP model isochrones by Bruzual & Charlot (Bruzual 2000) in (V-I) vs. (V-K) ( left panel) and (U-I) vs. (V-K) ( right panel) colour-colour diagrams. The age splitting in (U-I) vs. (V-K) plots is increased by a factor of 2 and will therefore result in a better age resolution. For direct comparison we chose an uniform scaling for both plots. The difference in (V-I) and (U-I), respectively, between a 15 Gyr model isochrone and the one for 13, 10, 7, and 5 Gyr (from bottom to top) is shown here.

Open with DEXTER


In the text