It is apparent that much effort has gone into this manuscript in responding to the reviewer's concerns and incorporating the suggested changes. I am almost satisfied that the manuscript has reached maturity and can be published in GMD.

I would like to ask the authors for one final effort before we can go ahead with publication: following the comments of reviewer #2 the authors should try to use more quantitative assertions when disussing model performance.

To give an example: In the discussions (p.5346, l.22 and onwards) the authors state that "the phenology model performed __quite well__ at temperate sites...", and similar qualitative terms are used throughout the text.

While I appreciate that this "policy" makes the text more readable and approachable it also undermines somewhat the author's efforts in quantitaively assessing the JULES model's performance.

I suggest to ammend the text by introducing a consistent termonology for the very same terms. E.g., the authors could define "quite well" as meaning that the bias is less than 1 gC/m2/s and the RMSE is less than 2 gC/m2/s. Similar definitions can be found for other terms like "poor", etc., and other metrics could be applied, too. My numbers are fairly arbitrary at this point and the authors will have to choose their own. The main advantage in my view, however, is that the qualitative terms also gain a quantitative or at least semi-quantitative meaning which can than be debated and discussed in the follow-up literature.

The definition of a consitent terminology can be achied either by adding a table or by adding the metrics in parenthesis at least at their first introduction in the text. For instance, a single paragraph fairly early on could be enough to define the terms.

I would like to ask the authors for this "final push" and then we shal go ahead with publication.

At this point I would also like to thank the anonymous reviewers for their sterling job and all their efforts.