Early-type galaxy stellar populations in the near-infrared

Stellar population synthesis (SPS) models are key to interpreting the spectral energy distributions of galaxies. While different models typically agree at optical wavelengths, these same models give vastly different predictions in the near-infrared, mainly due to different treatments of late-stage stellar evolution, in particular the Thermally Pulsing Asymptotic Giant Branch (TP-AGB) phase. This thesis tests three popular SPS models, with a focus on evaluating the differing treatments of this phase. To this end, we obtain high quality near infrared spectroscopy of a sample of galaxies with well-known spectral properties. These galaxies were selected from the ATLAS 3D sample to span a large range of ages, but other physical properties such as velocity dispersion do not vary significantly within the sample, allowing age-dependent phases such as the TP-AGB to be studied. We measure the strengths of the CN index at 1.1 μm and the CO index at 2.29 μm and compare these with the predictions of the Bruzual and Charlot (2003), Maraston and Strömbäck (2011), and Conroy et al. (2009); Conroy and Gunn (2010) SPS models. We also perform full spectral fitting of these same models to the data. We find that none of the models can accurately predict the strengths of either the CN or CO indices. Furthermore, we find that the age estimates obtained by fitting single stellar populations (SSPs) to the near infrared data do not correlate with those obtained from the optical or with those obtained by each other. The inclusion of multiple stellar populations alleviates some issues found with SSP fitting, however all models are found to have certain fundamental inaccuracies in the near-infrared regime.