2015MNRAS.449..296B

The inherent statistical and operational properties of the Lick-index spectrophotometric system are investigated to facilitate a more appropriate use for astrophysical studies. Non-Gaussian effects in the index standardization procedure suggest that a minimum S/N ratio has to be met by spectral data, such as S/N ≳ 5/px for a spectral resolution R ∼ 2000. In addition, index (re-)definition in terms of narrow-band colour should be preferred over the classical pseudo-equivalent width scheme. The overlapping wavelength range among different indices is also an issue, as it may lead the later ones to correlate beyond any strictly physical relationship. The nested configuration of the Fe5335 and Fe5270 indices, and the so-called Mg complex (including Mg1, Mg2 and Mgb), is analysed, in this context, by assessing the implied bias when joining entangled features into global diagnostic meta-indices, like the perused [MgFe] metallicity tracer. The perturbing effect of [Oiii]5007 and [Ni]5199 forbidden gas emission on Fe5015 and Mgb absorption features is considered, and an updated correction scheme is proposed when using [Oiii]5007 as a proxy to appraise Hβ residual emission. When applied to the present-day elliptical galaxy population, the revised Hβ scale leads, on average, to 20-30 per cent younger age estimates. Finally, the misleading role of the christening element in Lick-based chemical analyses is illustrated for the striking case of Fe4531. In fact, while iron is nominally the main contributor to the observed feature in high-resolution spectra, we have shown that the maximum responsiveness of the Fe4531 index is actually to titanium abundance.