Dear M. El Montassir,

Based on input from the reviewers, your responses, and my own review of this manuscript, I am pleased to accept this paper for publication in GMD pending a few revisions noted below. Congratulations!

With Kind Regards,
-Travis-

Minor revisions:

* Reviewer 1 suggested avoiding use of the term "residual modeling" and I agree with their reasoning. Your revised manuscript avoids the use of this phrase in one place, but it remains in others (e.g., Figure 1). I suggest doing a global search-and-replace, using the phrase "error modeling" instead.
* line 319 - "and EXIM is advecting the last observation while keeping the same level of details" <-- I agree that's true, but HyPhAICC is also advecting the last observation, so this explanation seems incomplete. Is it because EXIM uses a more accurate (and less diffusive) advection scheme? Can you reference EXIM documentation that would support this argument? If not, the discussion here should be revised.

Thank you for the revisions: this looks great!