This study built an MLT-dependent magnetopause location model expressed in terms of L* and driven by the Kp index. This model can be used in the ring current and radiation belt simulations. However, this model doesn’t consider drift shell splitting and has no dependence for the second adiabatic invariant, which has been demonstrated to be important in simulations of radiation belt dropouts. Several concerns need to be addressed before I can recommend this manuscript for publication.

1. Due to the drift shell splitting, L* values for particles measured at the same locations with different pitch angles are distinct. However, the proposed magnetopause model does not include pitch angle or second invariant (K) dependence. Previous studies (Tu et al., 2019) have demonstrated that K-dependent last closed drift shell is critical to simulating radiation belt dropout. The authors should discuss how the absence of K-dependent in the model might affect its accuracy.
2. Line 94-95. Xiang et al. (2017) is the first study to suggest that LCDS is more reliable than magnetopause standoff position to evaluate the impact of magnetopause shadowing.
3. Line 115. The effects of magnetospheric electric fields on particles are emphasized to support the necessity of building a 4D radiation belt modeling. However, the proposed magnetopause model seems not account for the effects of these magnetospheric electric fields.
4. Figure 10, Figure 11. These two figures demonstrate that the 4D model has better performance than the 3D model. However, the differences between 4D model and satellite observations at some locations have large values (>2). The RMSE is always >0.5 at all MLTs. This raises questions about the model's accuracy in representing the magnetopause L* location. The authors should provide a more detailed discussion of these discrepancies
5. The new magnetopause locations model is developed for the 4D Salammbo model. However, there is not any simulation results with the new model in the study. Some comparisons between the simulation results using the new and previous magnetopause location model in the 4D Salammbo model are needed to validate the advances of the new model.

Reference:

Xiang, Z., Tu, W., Li, X., Ni, B., Morley, S. K., & Baker, D. N. (2017). Understanding the mechanisms of radiation belt dropouts observed by Van Allen Probes. Journal of Geophysical Research: Space Physics, 122, 9858–9879. https://doi.org/10.1002/2017JA024487

Tu, W., Xiang, Z., & Morley, S. K. (2019). Modeling the magnetopause shadowing loss during the June 2015 dropout event. Geophysical Research Letters, 46, 9388–9396. https://doi.org/10.1029/2019GL084419

This paper developed a magnetopause location model which considered the drifting effects based on statistical analysis of observational data. The produced magnetopause model is dependent on L*, MLT, and Kp index. The model is an important component to global Fokker-Planck models, providing constrains on the simulation boundaries and magnetopause shadowing. The semi-analytical format of the model also makes it an efficient model. This paper studied an important subject and is recommended to be published after the following minor issues are resolved.

1. In Figure 5, the author used a linear fit for the Kp dependence at MLT of 10. The curve shown, however, seems closer to a parabolic fit. Can the author explain why a linear fit is used instead of any higher order fit?
2. Line 93: “modelmodeling” is a typo.
3. Line 318 “The documented dawn-dusk asymmetry of the magnetopause during disturbed times.”: This sentence is incomplete.
4. Figure 10: The color-bar makes it hard to distinguish the performances of 3D and 4D models. It is suggested that the author changes a color-bar or chooses a different format for the figure.
5. The author provided many discussions on the importance of a MLT dependent magnetopause model in the FP simulations. However, it is not shown or discussed, how the new 4D model will affect the simulations, aside from reducing the boundary error to below the grid sizes. The author should present some discussions, or preferably simulation results with the new 4D model.