The stellar population histories of early-type galaxies. II. Controlling parameters of the stellar populations

This paper analyzes single stellar population (SSP)-equivalent parameters f or 50 local elliptical galaxies as a function of their structural parameter s. The galaxy sample is drawn from the high-quality spectroscopic surveys o f Gonzalez (1993) and Kuntschner (1998). The basic data are central values of SSP-equivalent ages, t, metallicities, [Z/H], and "enhancement" ratios, [E/Fe], derived in Paper I, together with global structural parameters incl uding velocity dispersions, radii, surface brightnesses, masses, and lumino sities. The galaxies fill a two-dimensional plane in the four-dimensional s pace of [Z/H], log t, log sigma, and [E/Fe]. SSP age, t, and velocity dispe rsion, sigma, can be taken as the two independent parameters that specify g alaxy's location in this "hyperplane." The hyperplane can be decomposed int o two subrelations: (1) a "Z-plane," in which [Z/H] is a linear function of logo and logt and (2) a relation between [E/Fe] and sigma in which [E/Fe] is larger in high-sigma galaxies. Velocity dispersion is the only structura l parameter that is found to modulate the stellar populations; adding other structural variables such as I-e or r(e) does not predict [Z/H] or [E/Fe] more accurately. Cluster and field ellipticals follow the same hyperplane, but their (sigma, t) distributions within it differ. Most Fornax and Virgo cluster galaxies are old, with a only a small sprinkling of galaxies to you nger ages. The field ellipticals span a larger range in SSP age, with a ten dency for lower sigma galaxies to be younger. The present sample thus sugge sts that the distribution of local ellipticals in the (sigma, t) plane may depend on environment. Since the (sigma, t) distribution affects all two-di mensional projections involving SSP parameters, many of the familiar scalin g laws attributed to ellipticals may also depend on environment. Some evide nce for this is seen in the current sample. For example, only Fornax ellipt icals show the classic mass-metallicity relation, whereas other subsamples do not. The tight Mg-sigma relations of these ellipticals can be understood as two-dimensional projections of the metallicity hyperplane showing it ed ge-on. At fixed sigma, young age tends to be offset by high [Z/H], preservi ng Mg nearly constant. The tightness of the Mg-sigma relations does not nec essarily imply a narrow range of ages at fixed sigma. Although SSP parameters are heavily weighted by young stars, modeling them still places tight constraints on the total star formation history of ellip tical galaxies. The relation between [E/Fe] and sigma is consistent with a higher effective yield of Type II SNe elements at higher sigma. This might occur if the IMF is enhanced in massive stars at high sigma, or if more SNe II-enriched gas is retained by deeper galactic potential wells. Either way , modulating Type II yields versus sigma seems to fit the data better than modulating Type Ia yields. The Z-plane is harder to explain and may be a po werful clue to star formation in elliptical galaxies if it proves to be gen eral. Present data favor a "frosting" model in which low apparent SSP ages are produced by adding a small frosting of younger stars to an older "base" population (assuming no change in sigma). If the frosting abundances are c lose to or slightly greater than the base population, simple two-component models run along lines of constant sigma in the Z-plane, as required. This favors star formation from well-mixed pre-enriched gas rather than unmixed low-metallicity gas from an accreted object.