We present population synthesis models for the calcium II triplet (CaT), currently the most popular metallicity indicator, based on high-resolution empirical spectral energy distributions (SEDs). Our new CaT models, based on empirical SEDs, show a linear correlation below [Fe/H] ~ -0.5, but the linear relation breaks down in the metal-rich regime by converging to the same equivalent width. This relation shows good agreement with the observed CaT of globular clusters (GCs) in NGC 1407 and the Milky Way. However, a model based on theoretical SEDs does not show this feature of the CaT and fails to reproduce observed GCs in the metal-rich regime. This linear relation may cause inaccurate metallicity determination for metal-rich stellar populations. We have also conrmed that the effect of horizontal- branch stars on the CaT is almost negligible in models based on both empirical and theoretical SEDs. Our new empirical model may explain the difference between the color distributions and CaT distributions of GCs in various early-type galaxies. Based on our model, we claim that the CaT is not a good metallicity indicator for simple stellar populations in the metal-rich regime.

ABSTRACT
1. CONSTRUCTION OF CALCIUM II TRIPLET MODEL
2. IMPLICATIONS OF NEW CAT-METALLICITY RELA-TIONS
REFERENCES

• Chul Chung(Department of Astronomy & Center for Galaxy Evolution Research, Yonsei University)
• Suk-Jin Yoon(Department of Astronomy & Center for Galaxy Evolution Research, Yonsei University)
• Sang-Yoon Lee(Department of Astronomy & Center for Galaxy Evolution Research, Yonsei University)
• Young-Wook Lee(Department of Astronomy & Center for Galaxy Evolution Research, Yonsei University)