Do Statistical models capture magnetopause dynamics during sudden magnetospheric compressions?

Under steady-state conditions the magnetopause location is described as a pressure balance between internal magnetic pressures and the external dynamic pressure of the solar wind. The question is, does this approximation hold during more dynamic solar wind features?

Under more extreme solar wind driving, such as high solar wind pressures or strong southward-directed interplanetary magnetic fields, this boundary is significantly more compressed than in steady-state, playing a significant role in the depletion of magnetospheric plasma from the Van Allen Radiation Belts, via magnetopause shadowing. Large step-changes in solar wind conditions enable the real magnetopause to have a significant time-dependence which empirical models cannot capture.

We use a database of ~20,000 magnetopause crossings, to determine how the measured magnetopause differs from a statistical model, and under which conditions. We find that observed magnetopause is on average 6% closer to the radiation belts,  with a maximum of 42%, during periods of sudden dynamic pressure enhancement, such as during storm sudden commencement. Our results demonstrate that empirical magnetopause models such as the Shue et al. [1998] model should be used cautiously to interpret energetic electron losses by magnetopause shadowing.