On the accuracy of selected empirical magnetopause models

Empirical magnetopause models generally aim to predict its location as a function of upstream solar wind parameters. Many such models have been developed to date, typically based on the fitting of individual magnetopause crossings identified in spacecraft data by prescribed empirical functions deemed to reasonably characterize magnetopause shape and position dependences on selected control parameters. We use a unique list of more than 60,000 magnetopause crossings identified in THEMIS A-E, Magion, Geotail, and Interball spacecraft data to evaluate the performance of some of the most popular such models (Formisano et al., 1979; Petrinec and Russell, 1996; Shue et al., 1997; Lin et al., 2010). Differences between observed and model magnetopause locations are investigated as a function of solar wind dynamic pressure, interplanetary magnetic field magnitude, clock angle, and cone angle. A particular attention is paid to the magnetopause shape. This is studied both in terms of the level of the tail flaring, assuming a rotational symmetry around the aberrated x-axis, and in terms of asymmetries present in the real configuration. We show that although the Lin et al. (2010) model performs arguably the best, some systematic deviations are still present.