Abstract details

We present a new approach for simulatneous extraction of stellar population parameters and internal kinematics from the spectra integratedalong a line of sight. The idea of the method is a fitting of a templatespectrum into an observed one in a pixel space. The core of the method is a non-linear $\chi^2$ minimization in the multidimensional parameter space, including characteristics of a line-of-sight velocity distribution (LOSVD) and parametrized star formation history (SFH). At every step of the minimization procedure a template spectrum is derived by interpolating among a grid of pre-computed models, corresponding to parameters of SFH, and after broadened using LOSVD. Our present implementation of the method is limited to solar $[\alpha/Fe]$ abundance ratios, because we make use of recent evolutionary synthesis models computed with the PEGASE.HR package using the high-resolution ELODIE.3 stellar library. Representing SFH by a single stellar population, characterised by only age and metallicity allows direct comparison with results obtained using measurements of the Lick indices. Formal statistical errors confirmed with Monte-Carlo simulations point out that precision of stellar population parameter estimations with our technique is two to three times better compared to Lick indices in the spectral band of 4100\AA\ and 5500\AA, and spectral resolution about R=2000. We appled our technique to the 3D spectroscopic observations for a handful of dE galaxies in the Virgo cluster and nearby groups, obtained with the Russian 6-m telescope, and multi-object spectroscopy of a large sample of low-lumonisity early-type galaxies in the Abell 496, made with Giraffe at VLT. We obtained undoubted evidences for a presence of embedded kinematically and evolutionary decopuled substructures in dE and dS0 galaxies, which strengthen the connection between early and late-type galaxies and shed a light on their formation and evolution.