The manuscript titled "Duration of vegetation green-up response to snowmelt on the Tibetan Plateau" by Ni and colleagues investigates the complex interactions between spring phenology and snowmelt dynamics on the Tibetan Plateau (TP), a region noted for its ecological sensitivity to climate change. The authors employ satellite-derived phenological data along with various statistical analyses to explore the spatiotemporal patterns and drivers of the time difference between snowmelt and vegetation green-up. This work offers valuable insights into TP ecosystem dynamics, aligning with the research interests of EGUsphere readers. However, there are several issues that could affect the robustness of the conclusions. The following major suggestions aim to enhance the paper’s scientific impact and clarity.

Major concerns

1. The first concern pertains to the snow coverage. While the TP experiences frequent snowfall, snow cover duration can be brief due to sublimation and wind dispersal. It is essential to verify that the study areas experience sustained snow cover throughout winter, not just isolated pixels as depicted in Figure 2. Additionally, consider streamlining the main text by moving certain figures (e.g., Figures 1 and 2) to the supplementary materials.
2. The second concern is about the statistical analysis. Firstly, the variables—spring mean temperature, total spring rainfall, and daily snowmelt—are likely to exhibit multicollinearity, given the interdependence of temperature/rainfall and snowmelt. A multicollinearity check is recommended to identify and potentially exclude highly correlated variables. Employing a structural equation model could provide a more nuanced understanding of these interdependencies.   Secondly, the chosen significance level (p < 0.1) was too large. Despite this, a substantial number of pixels on the TP show non-significant trends, suggesting an absence of robust relationships between green-up and snowmelt. To address this, consider categorizing pixels by significance, explaining the underlying causes for these patterns in each category.
3. The third concern is about the structure of results. The results section currently includes an extensive number of figures, which may hinder clarity. Consider reorganizing this section, for example, separating spatial and temporal characteristics into distinct parts and using concise figures. This restructuring could improve readability and emphasize key findings more effectively.

Minor concerns

1. L15: Clarify whether you mean the "duration or date" of vegetation green-up.
2. L18-19: It is unnecessary to listing all methods here.
3. L50-55: too many abbreviations make this paper hard to follow
4. L118: Provide more details for this treatment. For instance, if merging 10 pixels with various plant functional types (PFTs), specify which PFT the combined pixel represents. Confirm if all PFTs were included, and consider excluding bare and arable land, which lack seasonal dynamics relevant to this analysis.
5. Figure 3: please make sure two figures have matched pixels and adjust the colorbar in 3b for better visibility.
6. Figure 4: With 18 subplots, distinguishing annual differences is challenging. Move this figure to supplementary materials and replace it with a simplified version, such as a comparison between two periods (e.g., 2001-2009 vs. 2010-2018).
7. Figure 5: Similar suggestion as Figure 4—consider a more concise format.
8. The direct relationship between temperature/precipitation and green-up may be more pronounced than that of snowmelt. If so, this would suggest a lesser role for snowmelt, especially given the year-round snowfall on the Tibetan Plateau.
9. Figure 7: Why the figure 7b use a different pattern unlike 7a? it is much better to testing Tspring and Sstog, and Pspring and Sstog effects on deltaT effect on ∆T separately for clarity.
10. Conclusion: Condense to focus on primary findings for a stronger impact.

GENERAL COMMENTS

The manuscript Duration of vegetation green-up response to snowmelt on the Tibetan Plateau by Ni et al. addresses the responses of date of snowmelt and date of green-up - and the time lapsed between the two - to climate change. The study raises up questions relevant within the scope of BG, and present some novel results that highlight the importance of studying snowmelt patterns in alpine regions. The scientific methods are clearly outlines, but the statistical analysis is described in too little detail, which does not allow to reproduce the results. The title reflects the content of the article, and the abstract is good, though with too many details, it can be written in a more concise way. Some more references could be added, but in general they provide good and appropriate references.

The are two major concerns I have about the article.

1. The statistical anlyses are not actually described. In the Material and methods there are no details describing what program was used, what functions, how they dealt with problems in the different models... And in the results they do not provide any details either, they only indicate whether something is significant or not. They should at least mention the number of replicates, p-values, confidence intervals... They could include all these details in one table, for ease of reference. And they use different significant thresholds for different tests. They should provide the reasoning behind that, otherwise it looks like they were chosen a posteriori, which is not scientifical. As the other referee points out, in one of the tests they use a significant threshold of 0.1, which seems quite high.

2. They do not write anything about what the objectives of the study are, and what their hypothesis are either. I think they need to state them clearly in the introduction, and the rewrite the structure of the Results and Discussion to answer the questions they pose. At the moment the structure is not completely cohesive. Having some clear hypothesis would help to streamline the flow of ideas. For example, the first subsection in the discussion is about a point that has not been raised before, and is not mentioned either in the Introduction or in the Results. I reckon this would also help the authors decide which results and figures to show in the Results section, something the other referee also mentions.

I think these points should be addressed before working any further in the manuscript. I consider that these points, though they require quite some work, do not change the content of the manuscript in a major way, so I think just a minor revision (though extensive) is needed.

SPECIFIC COMMENTS

1. L30-31. You say that the TP plays a role in maintaining global biodiversity and ecological security. Maybe add short sentence explaining what that role is?

2. L50. Regarding snow phenology in the TP, I would appreciate a short description of when the first snow tends to arrive, for how long it lasts and when it starts melting. I think it would make it easier for the reader to picture it. Maybe include a table showing the different values for the different regions? Seeing those values would help appreciate the changes experienced by the different regions

3. L61-65. This is part of the materials and methods. Here you should indicate what your objectives and hypotheses are

4. L89-90. Where was the dataset obtained from? Where did the data come from?

5. L143, Figure 2 caption. I write it here, but it applies to other captions as well. Could you write a more detailed caption?

It is explained in the text, but looking only at the figure I do not know what "Filtered" means, or what the 47 days, 3 days and 6 days indicate, for example. It would be easier for the reader if these values are explained, shortly, in the caption was well

6. L152. When calculating I_I, how do you determine what the neighboring regions are? Is it just the adjacent pixels, or is it more complex?

7. L186-187. Do you mean that, at temperatures below 0 degrees, the correlation between SCED and Tgu was low, while at temperatures above 0 the correlation was high? I think you could rewrite this sentence to make the message clearer

8. L188. "a strong negative correlation prevailed". Between T_SOM and T_GU? So, in areas of high temperature, a late T_SOM was related to an early T_GU, and an early T_SOM to a late T_GU? Or am I misunderstanding it?

9. L189. What environmental conditions? The factors mentioned afterwards are either temperature-based or precipitation-based.

10. L190. Regarding "influencing factors". Why did you end up choosing these factors? Was it based on the literature, did you perform any tests to see what factors had the largest coefficients...?

11. L195. Any reasons for not considering interactions between the factors?

12. L224. When you say that a test or model was significant (here and later on) can you provide the statistical values connected to the test (p-value, confidence interval...), in order to give a better idea of the strength of the effect? And can you say anything about the goodness of fit of the model in the supplementary material?

13. L249. You indicate how much ΔT advances near the freezing point. But do you mean that it advances 3 days if we consider an increase in temperature of 1 K? Or what do you mean?

14. L250. "ΔT fluctates with temperature". Do you mean that there is not a consistent trend, i.e., that higher temperatures do not always have shorter ΔT? Or something else? Could you make it clearer?

15. L264. Just to make sure I understand your point. Even though higher T_spring was connected to lower ΔT values, in more than 50% of the plots that had a significant correlation a greater average temperature in spring led to a delayed green-up after the start of the snowmelt, right?

16. L286. [This comment refers to the whole section 4.1]

This is the first time you mention that this study will look at different ways of identifying T_SOM. You should include a paragraph, or at least a couple of sentences talking about it in the introduction, describe in the material and methods how the values are obtained, and present the results in the Results section

This section is good, but comes up abruptly. I was not expecting this after reading the rest of the article.

17. L311. It would have been interesting if you could have had at least one sample area per land cover type (10 in total), as I imagine they differ in the snow cover pattern and properties. But maybe it was not possible?

18. L318. I reckon you could first have a section where you discuss T_SOM and T_GU by themselves (section 3.1 in the Results): their spatial distribution, connecting it to the different land cover types (and maybe other factors); and their temporal distribution and how you find very few areas with a significant trend.

I think it would be best to cover these points before discussing the ΔT (which is section 3.2 in the Results).

19. L324. Maybe you could add a short sentence here, after "respectively", explaining that the larger deltaT in the TP is due to its characteristics - Which would link this paragraph to the following ones

20. L360-363. I think you should write shortly what this actually means for the different vegetation types. What would happen to the vegetation (both higher and lower) if temperature increases and/or if snowmelt happens earlier?

21. L373. You could name what areas in the TP experience a negative ΔT, and explain shortly how they might be affected by climate change.

22. L378. The conclusion includes to many details. You do not need all the specific numbers here, they appear in the rest of the manuscript. Here you should only include those details and number that more strongly support your conclusions.

TECHNICAL CORRECTIONS

1. L73, Figure 1. This applies also to some of the other Figures. They are a bit difficult to read, especially the text. Could you make the font slightly larger, or the figure?

2. L73. It is difficul to see the difference in colour between "Shrub" and "Coniferous forest" in the legend, and impossible (at least for me) in the Figure. Would it be possible to use colours with more contrast?

3. L120. IIt says "Subsection (as Heading 3)". I assume this is a typo, and you wanted to write only "Calculation of ΔT"

4. L191. T_spring (here and in the whole manuscript). It might be best to use a different letter for temperature. In most cases you use T to denote time (T_SOM: time of start of snowmelt; T_GU: time of green-up). Using then T for T_spring could be confusing for the reader.

Or use a different letter for denoting time. Maybe D, as you are always talking about dates?

5. L211. "more concentrated"? The range is almost the same (30 days)

6. L212. Is the progress not from East (earlier T_GU) to West (later T_GU)?

7. L223. "to gain more about spatial distribution". To gain what? More insight? More information?

8. L233, Figure 5. Have you tried using the darkest colours to represent the high-high and low-low clusters (since they are more common), and paler colours for the other ones? Pale colours (at least in my case) are more difficult to distinguish from each other in a small image

9. L248. Do you mean "constant" instead of "consistent"?

10. L249. "rises" instead of "rise".

11. L253. Instead of "amplify" I think "strengthened" would be a better word. It implies that the reduction is greater, but also more consistent, which seems to be the case in your data

12. L280. "taller" instead of "higher".

13. L330-331. This is actually not that easy to see in the figure, maybe this figure should be larger.

14. L351-352. I suggest to rewrite this sentence. Something like this:

"Conversely, in colder regions increased temperatures can reduce cold stress on vegetation, resulting in a larger effect on T_GU (-0.27), similar to that on T_SOM (-0.28)."

15. L353. "enhance" might be a bit of a misleading word. The reader might think that a better Time of start of snowmelt is an earlier Time of start of snowmelt. You could say "do not significantly lead to later T_SOM", for example.

16. L364. "the growth rate of soil temperatures"? What do you mean?