Dear authors

Many thanks for your diligent consideration of the reviewers' concerns. I am content with your points and modifications.

Thank you very much for your diligent consideration of the reviewers' concerns.
It gives me great pleasure to state that the Topical Editor and I are happy for this paper to be published subject to a few very minor technical corrections.

Here are the few very minor (hence labeled technical) additions coming out of your reply to Koulakov's review that the Topical Editor and I would like you to incorporate for the final manuscript:

- Please mention in the text that source localisation is done, and where in the workflow this takes place. Readers should not need to consult the manual to understand what kind of basic operations the code is performing.

- In the discussion, please also mention the limitations and subjective choices that are necessary to utilise this method, including for instance the limited scope of synthetic tests, choice of parameterisation (as in any tomographic approach).

- "stable" 2D inversion: One shouldn't give readers the impression that a stable or robust, convergent inversion has more to do with reality than parameterisations or 3D for which inversions may become unstable: It may simply come from oversimplification of reality. Again, please make sure to mention something along these lines in the discussion, related to limitations of interpreting the 2D approximation rather than reality, and relate this to the uncertainty in interpreting tomographic results.

- it doesn't hurt to mention that you have done extensive 3D checkerboard tests and modeling even though the residual reduction is small.

Congratulations again, and many thanks for considering Solid Earth for your submission.

Best wishes

Huw