A STOCHASTIC APPROACH TO CHEMICAL EVOLUTION

Observations of elemental abundances in the Galaxy have repeatedly sho wn an intrinsic scatter as a function of time and metallicity. The sta ndard approach to chemical evolution does not attempt to address this scatter in abundances since only the mean evolution is followed. In th is work, the scatter is addressed via a stochastic approach to solving chemical evolution models. Three simple chemical evolution scenarios are studied using this stochastic approach: a closed box model, an inf all model, and an outflow model. These models are solved for the solar neighborhood in a Monte Carlo fashion. The evolutionary history of on e particular region is determined randomly based on the star formation rate and the initial mass function. Following the evolution in an ens emble of such regions leads to the predicted spread in abundances expe cted, based solely on different evolutionary histories of otherwise id entical regions. In this work, 13 isotopes are followed, including the light elements, the CNO elements, a few alpha-elements, and iron. It is found that the predicted spread in abundances for a 10(5) M. region is in good agreement with observations for the alpha-elements. For CN , the agreement is not as good, perhaps indicating the need for more p hysics input for low-mass stellar evolution. Similarly for the light e lements, the predicted scatter is quite small, which is in contradicti on to the observations of He-3 in H II regions. The models are tuned f or the solar neighborhood so that good agreement with H II regions is not expected. This has important implications for low-mass stellar evo lution and on using chemical evolution to determine the primordial lig ht-element abundances in order to test big bang nucleosynthesis.