|The authors have made very significant changes to the paper, and have addressed the reviewer comments in a relatively profound way. The article contains a very detailed and rich analysis, and I think that the balance between the year 2015 and other years is now adequate. The authors also improved the discussions of links to previous studies on the relations between the carbon cycle, vegetation greenness and climate anomalies. The scientific methods are very well documented and appear reasonable to me, and the results support the author’s conclusions.|
There is one aspect that I think the authors did not address so well, which is to work out a clear motivation for their study and a message that emerges from their results. As a meticulous case study for the year 2015, I think that the paper can be useful for other scientists in the same field. Having said that, the authors may reconsider if the visibility of their paper might be improved by some additional changes, especially in the abstract, introduction, and conclusions section.
To me, the fact that greening and carbon loss can occur together due to different seasons, locations and processes, invokes the question what the future trends of these changes are going to be. For example, would an increased ENSO frequency or amplitude have the potential to overcompensate the carbon uptake due to warming in the extratropics? Can the author’s results be used for a simple back-of-the envelope calculation (for example, an extrapolation of their results into the future) to estimate a range of possibilities? Which sources of uncertainty in particular should be reduced in order to get a better estimate? Can the increased respiration in autumn in mid-latitudes overcompensate the increased uptake in spring, or does this only reflect the larger annual cycle with increasing overall uptake?
I understand that these may be questions too general and complex to answer. But my point is that using such kinds of questions at least as an orientation to guide the reader through the manuscript can improve its readability by pointing out the benefits that readers would have from spending time on the article. The aspects above are implicitly (and sometimes explicitly) already mentioned in the manuscript, but they are hidden in a large haystack of details. The very long and detailed documentation of results still leaves me a bit confused about how to link these results to the big questions about the carbon cycle. I believe that the authors have the expertise to work out these links in a more explicit way, without the need for much extra analysis.
- The overall language of the article is good, but I would recommend that a native speaker takes a final look. There are some issues with missing articles especially.
line 482: outside the growing season,
line 486: a strong transition to a carbon source,
line 487: observations showing a strong dependence of...
line 558: The validity period of ... is 2004-2015, although data for the whole time span is available.
line 785: still not good English – should read “apparent paradox”
line 859: more common would be “linked to”, or “associated with”
line 969: tropical vegetation shows
line 1034: decoupled outside the growing season ... The transition into a carbon source
- lines 559-562: I actually like the explanation in the response to my previous review more than the version in the article because it really eplains what the validity period is and why the estimates deteriorate outside. Readers who are less familiar with these technical details might not understand this paragraph as it is.
- line 683: The reason to remove year 1993 seems to be that it is an outlier and the authors conclude that the method fails in this case. But what assures the authors that a similar problem does not occur in other years as well, with a smaller effect on the results?
- line 822: What is meant with a “first-order difference”? Couldn’t one just remove “first-order”?
- line 1041: What is meant with opposing ENSO events?
- line 1044: I am still quite sceptical about the use of the term “abrupt” in this context. I would prefer the term “large climate anomalies”. An abrupt shift usually implies consequences that are permanent.
- line 1073-1075: This statement goes into the direction that I am driving at above and should be explained a bit more. It almost reads like an afterthought, but could be a major argument for the purpose of this study in the introduction. Why is this study an interesting test bed, and how? And after having done the analysis, how should one proceed to evaluate the models and gain understanding? Hence, the authors may want to fill this interesting statement with some more flavour and serve it as an appetizer instead of a dessert.