In this Appendix, we provide two examples of the results for different methods of sky subtraction for a subset of the measured indices. As discussed in the main text, night sky variability and the location of some GCs in crowded fields did not allow us to always find a suitable region close enough to the target that could be used as an external sky frame. When it was possible however, the difference in the results with respect to the sky estimate from the edges of the science frame was very small, especially in the case of metal lines. We note here that for the same two clusters, Puzia et al. (2002) found a larger difference between the two ways of accounting for the sky background. This maybe the consequence of the target region chosen to measure the sky spectrum. However, based on our tests, we agree with the results of these authors that the internal derivation of the background is more reliable.
Uncorrected Lick indices. Example of same-frame sky estimate versus external frame.
We show the uncorrected Lick indices (after sky subtraction) measured along the two directions for each individual cluster (Table B.1) as well as their average (Table B.2). We refer to the main body of the paper for details. We recall that, by “uncorrected” we mean data that have not (yet) been corrected for any offset between our and the Lick system caused by, e.g., residuals in the sky subtraction, and systematics in the wavelength calibration. The example is limited to a subset of the measured indices, the remainder of the sample being available upon request.
Lick indices measured along EW and NS directions and their associated statistical uncertainties.
Uncorrected Lick indices averaged over the two directions.
We present the uncorrected indices averaged over the two directions measured as if our observational set-up were in the native Lick resolution. The entries in Table C.1 should be compared to those in Table B.2. Here we also measure the indices assuming that our resolution precisely matches that of the Lick system (e.g. Worthey & Ottaviani 1997). Our resolution is slightly lower, hence a small correction is in principle needed (see Sect. 3), which however we do not apply here. The difference (Table C.2) in the end-products is remarkably small, typically being < 1% for Hβ and Mg2, and 3–4% for the other indices shown in the table.
Uncorrected indices averaged over the two directions – Indices measured as if they were at the native Lick resolution.
Applied velocity dispersion corrections.
© ESO, 2011