Thank you very much for attending completely and reasonably to all the points of the reviewers.

I have just a minor revision request. It might be worth mentioning in your text that setting nlayer = l for the case that you discuss, i.e. R1=1 and R2=2, leads to the radial dimension being around half of the lateral dimensions (at least for H20, others might not be quite as bad). This is typically not ideal in numerical methods. Clearly your method does not require nlayer to equal l, so this can be adjusted easily. I think it might be worth reminding the reader that nlayer=l is not always ideal basis for numerical methods (definitely not ideal for H20); and that they might consider using nlayer = 0.5*l (which is ideal for H20), which your code can easily accommodate.

I would to thank you for an excellent paper. I am sure that the code that you provide openly will be used widely as the basis of developing further codes, like the one you mention in your response.

Many thanks.

Dear Cedric

Excellent work. Excellent work also in responding fully to the reviewers comments. Many thanks.

Huw

Dear Cedric

Many thanks for submitting this work to Solid Earth!

Congratulations on an excellent paper.

Best wishes

Huw