标题： 用球状星团检验恒星群体拟合的要素 I：恒星光谱库 显示英文标题

标题： Testing Stellar Population Fitting Ingredients with Globular Clusters I: Stellar Libraries

摘要： 恒星系统的综合光谱包含了丰富的信息，对其分析可以揭示诸如金属丰度、年龄和恒星形成历史等基本参数。 广泛使用的分析这些光谱的方法基于将星系光谱与恒星群体（SP）模型进行比较。 尽管这是一个强有力的工具，但SP模型包含了许多成分，每种成分都有其假设和不确定性。 在多种可能的不确定来源中，很难直接确定哪种成分主导了模型中的误差。 在这项工作中，我们提出了一项研究，即独立于其他成分来研究SP模型的一个成分——光谱恒星图书馆。 为此，我们将使用具有可用色-亮度图（CMDs）的球状星团的综合光谱。 从这些CMDs中，可以对这些物体的综合光谱进行建模，而无需采用或假设SP模型的另外两个主要成分——演化轨迹或IMF。 在这里，我们测试了四个常用的恒星光谱库。 我们发现，在平均通量不确定性为5%的情况下，这些光谱库能够再现30个星团光谱中的18个。 对于所测试的更长波长范围，理论光谱库在比较中优于经验光谱库。 如果没有蓝光部分的光谱，经验光谱库的表现优于理论光谱库，特别是在个别特征方面。 然而，对于理论光谱库的结果是有希望的，它们同样有效地再现了整个光谱。 显示英文摘要

摘要： The integrated spectra of stellar systems contain a wealth of information, and its analysis can reveal fundamental parameters such as metallicity, age and star formation history. Widely used methods to analyze these spectra are based on comparing the galaxy spectra to stellar population (SP) models. Despite being a powerful tool, SP models contain many ingredients, each with their assumptions and uncertainties. Among the several possible sources of uncertainties, it is not straightforward to identify which ingredient dominates the errors in the models. In this work we propose a study of one of the SP model ingredients -- the spectral stellar libraries -- independently of the other ingredients. To that aim, we will use the integrated spectra of globular clusters which have color-magnitude diagrams (CMDs) available. From these CMDs it is possible to model the integrated spectra of these objects without having to adopt -- or make assumptions -- on the other two main ingredients of SP models, evolutionary tracks or an IMF. Here we tested four widely used stellar libraries. We found that the libraries are able to reproduce the integrated spectra of 18 of the 30 cluster spectra inside a mean flux uncertainty of 5%. For the larger wavelength range tested, a theoretical library outperforms the empirical ones in the comparison. Without the blue part of the spectra, empirical libraries fare better than the theoretical, in particular when individual features are concerned. However, the results are promising for theoretical libraries, which are equally efficient to reproduce the whole spectrum.