Fitting surface-independent phase shift differences

With the great inference power asteroseismology brings to model fitting of stars, one of the greatest challenges is the asteroseismic surface effect. While it has previously been circumvented using combinations of individual frequencies, most often the frequency separation ratios, they are derived using several approximations, which decreases the precision of the inferred parameters. Here, we present the epsilon differences method, which relies on how specific combinations of individual frequencies cancel out the contribution from the surface, derived using less approximations than frequency separation ratios, and show how it performs in comparison to the established methods.