10:00–10:30 am ERC 401
Constraining Models of Dwarf Galaxy Evolution with Observations and Bayesian Statistics
Advisor: Andrey Kravtsov
We present a statistical method for constraining the parameters of GRUMPY, a regulator model for dwarf galaxy evolution, using observed metallicities of stars in ultra-faint dwarf galaxies. Although the fiducial GRUMPY model successfully reproduces the stellar mass-metallicity relation, diverse star formation histories, and other observed properties of dwarf galaxies, certain model parameters remain poorly constrained due to uncertainties in observational and theoretical studies. To derive more robust constraints, we use maximum likelihood estimation to constrain GRUMPY model parameters with observational data of five well-measured dwarf galaxies: Bootes I, Ursa Major I, Hercules, Hydrus I, and Leo IV. We conduct preliminary tests of our method by constraining the parameters of the GRUMPY outflows model, which are especially uncertain due to the difficulty of detecting outflows with observations. To minimize computational cost, we constrain each parameter individually and use only a small set of simulated galaxies to build each GRUMPY model. Despite the limited scope of these tests, we demonstrate that our method is able to recover best-fit parameter values similar to the values from the fiducial model. We discuss the implications of these results and provide suggestions for how our method can be extended to allow for more complex constraints and/or more diverse observational data in future work.