Investigating inner Galactic Bulge stars through the near infrared

Sammanfattning: Context: The formation of the Galactic Bulge is a topic of active research. There are many scenarios based on observations and Galactic evolution models. The key properties which need to be well constrained observationally are the metallicity distributions of stars and the spatial metallicity gradients. The metallicity distribution of stars in the inner Galactic Bulge (|b|<4o) is a subject of ongoing debate, with only a few spectroscopic studies based on small samples of stars. Many studies in the literature reported a narrow metallicity distribution, with one study which concluded a lack of a vertical metallicity gradient in the innermost regions (Rich et al. 2012). However, new studies from large scale surveys (Schultheis et al. 2015) are beginning to challenge this with larger dispersions in the metallicity distributions reported. These large scale surveys utilise automated pipelines to analyse spectra, which are known to have issues. As such, one would like to investigate and validate the reported results. Aims: Due to the large discrepancy in reported metallicity distribution in the inner Bulge and the challenging nature of spectral analysis (due to high extinction and cool effective temperatures), the stellar samples from two opposing studies (Rich et al. 2012; Schultheis et al. 2015) will be re-analysed to present a homogeneous dataset of stars in the inner Galactic Bulge. A benchmarked line by line analysis method would be used to determine the metallicities of the stars. One would also like to compare the results to those reported in an automated pipeline method, given there are known calibration issues with metallicities. Method: A set of iron lines were benchmarked against the Sun, Arcturus, and a sample of 96 high SNR stars using spectra in the near infrared (H-band). The iron lines which were found to produce accurate abundances were utilised to determine the metallicity of the inner Galactic Bulge stars. This was conducted by the fitting of synthetic spectra to the observed spectra within the program Spectroscopy Made Easy (SME), which employs a chi-squared minimisation algorithm to determine parameters. Results: The determined metallicities for the Schultheis et al. (2015) sample were within ~ 0.2 dex of the reported values. Both of the metallicity distribution were dominated by a peak at ~ -0.2 dex. The reported APOGEE values also show a bump in the distribution at ~ -0.8 dex, but is absent in the distribution based on the metallicities determined in this study. Preliminary tests on the Rich et al. (2012) sample gave large discrepancies between the synthetic and observed spectra. Investigation on the cause of this is still ongoing and hence no meaningful metallicities could be determined. Conclusions: Overall, we corroborate with APOGEE on the result of a large spread in the metallicity distribution. However, some deviations are found at the metal poor ([Fe/H] < -1.0 dex) and metal rich ([Fe/H] > 0.1 dex) end of the distribution, depending on the methodology. Considering the low sample size and uncertainties, the presence of a metal poor population ([Fe/H] ~ -1.0dex), as claimed in Schultheis et al. (2015), is weakened based on the results of this study.