Systematics in asteroseismic modelling: application of a correlated noise model for oscillation frequencies

The detailed modelling of stellar oscillations is a powerful approach to characterising stars. However, poor treatment systematics in theoretical models leads misinterpretations Here we propose more principled statistical for the be applied fitting individual mode frequencies with typical model grid. We introduce correlated noise based on Gaussian Process (GP) kernel describe given that frequency are expected highly correlated. show tuning GP can reproduce general features variations changing input physics and fundamental parameters. Fits better recover parameters than traditional methods which either ignore or treat them as uncorrelated noise.