Dear Dr. Dmitriev,

thank you thank you for submitting your manuscript "On the radiation belt location in the 23–24 solar cycles", and for your efforts to provide us with detailed answers to the reviewer comments and with an early, draft version of the revised manuscript. As you may judge from the interactive discussion, both reviewers consider your work very interesting and the results original. However, they have raised many outstanding issues that need to be resolved before a final decision can be made for the acceptance of your article for publication in An. Geoph. For this reason, I believe we can proceed to the follow-up stages of peer review by inviting you to submit your article updated along the lines of the responses and the draft revision you provided during the Interactive Discussion session. After you submit your revision, I will send it to at least one of the two previous referees for a second round of review, as an editorial assessment of your answers by me may not be sufficient.

While the answers you provide appear satisfactory, I would strongly recommend to add further discussion on several points, in order to improve further the chances for a positive outcome of your submission.

1) On the subject of linear fitting of the ORB location and the conclusion about its large, 2-3 deg latitudinal "offset". While I understand that variations seem to a first order random and a linear fitting is a good first approximation, the errors of the linear fitting parameters are comparable to the derived, averaged ORB displacements, even though for certain longitudes absolute values of errors are smaller. Inspection of the corresponding Fig. 4 and 5 plots give the impression that the linear fitting parameters could be misleading. E.g., it should be investigated how fittings are affected by outlier points: in Fig.4 there are clear, extreme low ORB values in 2017 (top panel) and 2012 (bottom panel). As the are potential outliers which may control the fitting and impose a negative latitudinal gradient, their origin could be discussed. Also, a few deg change in ORB over more than a solar cycle should be obtained even if you e.g. remove 1-1.5 years of data either from the start or the end of the data series - that could be a further test of your findings. It was also not clear to me if error propagation was used in the linear fitting (i.e. do you take into account the uncertainty in each data point?).

2) On the subject of the definition of ORB: one of the reviewers asked how is this obtained. You clarified that you define it as a starting point that a continuous increase is seen in the fluxes. However, I may have missed if you also require that a certain flux limit or S/N has to be be crossed. If so, please elaborate further. E.g. a continuous increase of fluxes can be slow or sharp - more impulsive. Do you define the boundary for such occasions in a different way?

I hope you find these additional comments helpful and I am looking forward to receiving your formal manuscript revision.

Kind regards,

Elias Roussos