Preprint typeset using LATEX style emulateapj v. 5/2/11 CALIBRATING STELLAR POPULATION MODELS WITH MAGELLANIC CLOUD STAR CLUSTERS

Stellar population models are commonly calculated using star clusters as calibrators for those evolu-tionary stages that depend on free parameters. However, discrepancies exist among different models, even if similar sets of calibration clusters are used. With the aim of understanding these discrepancies, and of improving the calibration procedure, we consider a set of 43 Magellanic Cloud (MC) clusters taking age and photometric information from the literature. We carefully assign ages to each cluster based on up-to-date determinations ensuring that these are as homogeneous as possible. To cope with statistical fluctuations, we stack the clusters in five age bins deriving for each of them integrated luminosities and colors. We find that clusters become abruptly red in optical and optical-IR colors as they age from ∼ 0.6 to ∼1 Gyr, which we interpret as due to the development of a well-populated thermally pulsing asymptotic giant branch (TP-AGB). We argue that other studies missed this de-tection because of coarser age binnings. Maraston (2005) and Girardi et al. (2010) models predict the presence of a populated TP-AGB at ∼0.6 Gyr, with a correspondingly very red integrated color, at variance with the data; Bruzual & Charlot (2003) and Conroy et al. (2009) models run within the error bars at all ages. The discrepancy between the synthetic colors of Maraston (2005) models and the average colors of MC clusters results from the now obsolete age scale adopted. Finally, our finding that the TP-AGB phase appears to develop between ∼ 0.6 − 1 Gyr is dependent on the adopted age scale for the clusters and may have important implications for stellar evolution