Assessing the performance of magnetopause models based on THEMIS data

The magnetopause is the boundary between the Earth’s magnetic field and the interplanetary magnetic field (IMF), located where the supersonic solar wind and magnetospheric pressure are in balance. Although empirical models and global magnetohydrodynamic simulations have been used to define the magnetopause, each of these has limitations. In this work, we use 15 years of magnetopause crossing data from the THEMIS (Time History of Events and Macroscale Interactions during Substorms) spacecraft and their corresponding solar wind parameters to investigate under which solar wind conditions these models predict more accurately. We analyze the pattern of large errors in the extensively used magnetopause model and show the specific solar wind parameters, such as components of the IMF, density, velocity, temperature, and others that produce these errors. It is shown that (1) the model error increases notably with increasing solar wind velocity, decreasing proton density, and increasing temperature; (2) when the cone angle becomes smaller or |B_x| is larger, the Shue98 model errors increase, which might be caused by the magnetic reconnection on the dayside magnetopause; (3) when |B_y| is large, the error of the model is large, which may be caused by the east–west asymmetry of the magnetopause due to magnetic reconnection; (4) when B_z is southward, the error of the model is larger; and (5) the error is larger for positive dipole tilt than for negative dipole tilt and increases with an increasing dipole tilt angle. However, the global simulation model by Liu ZQ et al. (2015) shows a substantial improvement in prediction accuracy when IMF B_x, B_y, or the dipole tilt cannot be ignored. This result can help us choose a more accurate model for forecasting the magnetopause under different solar wind conditions.