Cosmic chemical evolution: here, there, and everywhere

Metals play a very important role in star formation and stellar evolution. Amongst other things
they control the cooling of the interstellar gas, thus allowing the formation of stars; they affect the radiation transport, through the opacities involved in the different microscopic processes; and they have the most important role in the dust formation and in the mass loss from stars. Metals are formed inside stars. Therefore, it is expected that metal enrichment in the universe should start soon after the formation of the first massive stars that, after a very short life, would return to the interstellar medium (ISM) in explosive events newly synthesised chemical elements, heavier than the primordial hydrogen and helium. It is thought that this ejecta, at some point mixed with the surrounding gas, would help to its cooling and propitiate the appearance of a new generation of stars, hence giving rise to the cycle of cosmic chemical evolution. All this wealth of detail is somehow imprinted in the distribution of abundances in galaxies. The absolute quantities of metals a galaxy (or a region of a galaxy) possesses, the relative abundances of the different elements, and their spatial distribution in a given galaxy are important constraints for the verification of our models and scenarios of the formation and evolution of galaxies.

There are two ways of handling and exploiting this information. One is to derive abundances in galaxies “in situ”, i. e. in galaxies at their different redshifts. In that way, we can observe the gas abundances at the time of the emission of radiation and thus obtain the value of the mean abundances of the gas at a given epoch. The other one is to derive the gas abundances at the present epoc , that is, observing galaxies at redshift zero and then use chemical evolution models to infer their abundances at previous epochs.

Eventually both approaches must converge in a single galaxy formation and evolution scenario. Once that convergence is achieved, the models can be extrapolated back to even earlier times and predictions can be made which can be contrasted with planned observations.