SIF-Based GPP Is a Useful Index for Assessing Impacts of Drought on Vegetation: An Example of a Mega-Drought in Yunnan Province, China
Round 1
Reviewer 1 Report
Dear authors,
Your paper “SIF-based GPP is a useful index for assessing impacts of drought on vegetation: An example of a mega-drought in Yunnan Province, China” is still not ready for publication as there are several un-solve problems and issues. After describing these issues, I‘m suggesting other small corrections
- The GPP-SIF data – in lines 138-146 you describe how you utilize several flux-towers to fit this GPP-SIF for local vegetation types, then you are taking one of these flux-towers and utilize it to show this data is more accurate compared to other products (lines 266-270, Fig 5). Verification should be conducted with non-calibration sites, so your message in the Results, Section 2 - “GPP-SIF has better accuracy than other GPP products” is biased and un-accurate. As so, please remove this section, or evaluate your data with a flux tower that was not part of the calibration process.
- Section 3.1 – (1) your title is that SIF data are more accurate than NDVI and EVI, yet in the first paragraph (lines 194-199) you explained that your study period is peak growth months as in these months “SIF is more sensitive to drought in the peak growing months”. If you chose a specific period or location, it should NOT be because it fits your dataset. Again, your results will be biased. (2) Yet, from the following sections it can be seen that you utilized all the data from 2009-2011 (Figure 10 for example), and in line 203 you mentioned 15 months. What is the correct study period? And if it changes from one section to the other, why? Further, if you chose specific months as you mentioned in line 203, why specifically these months and not others? (3) the correlation results – you classified all correlation values above 0.2 as the best value (Table 2) and stated that correlation values above 0.2 as “significantly positively correlated area (R2≥0.2)” (Line 217). This is un-accurate and most readers will not agree, as good values should be >0.7-0.75. The mean values you presented in Table 2 0.43-0.08 fit what I can see from Fig 2 (there is no linkage between the SIF, NDVI, and EVI to the drought index) and Fig 3 (large areas are classified as minus correlation, meaning that these indices do not represent correctly the drought).
- Section 3.3 – the GPP respond to the 2010 drought. According to the data you present in Figures 10 and 11, the GPP (all the 3 products) do not change a lot because of the drought. Your new feature, comparing the GPP from the drought months to the average GPP is good (the GPPd), but from your figures, it is difficult to see it. I would suggest changing Figure 12 so instead of a lot of small maps, make a graph with the x-axis being the time in months and the y-axis being the average province GPPd, and the average sc_PDSI. Further, you should change Figure 13 to a scatter plot: instead of two maps that is difficult to see the linkage between them, a scatter plot of all pixels will show if there is a linkage or not. From the data you present, your message that the GPP-SIF responds well to drought is not accurate.
Other issues:
- The writing includes long sentences, which enable the reader to follow what is the main message in that sentence. For example: “Furthermore, the spatial distribution of the three GPPs (Figure 7) is basically the same, with the Yunling-Yuanjiang valley as the boundary in general, and the GPP in the western region is significantly higher than that in the east, showing the blocking effect of the vertical ridge and valley area, showing a decreasing trend from southwest to north-east.” (Lines 282-286). In addition to a long sentence, the connection word, “showing” is repeated. It makes the impression that you did not edit what you wrote.
- You introduce the acronyms NDVI, EVI, and NDWI in lines 63, without the full name that generate these acronyms and no citation for the source of each one of them. Although you do it in line 108, and again in 112-113, you should do it the first time you wrote any acronym.
- R, the correlation coefficient – in Section 2.3.2 you describe this parameter as the main statistic, yet trough out the paper you are using the symbol R2, representing the coefficient of determination. The R-squared is essential in the scatter plot as Figures 4 and 5, but it needs to be correct in Table 2 and elsewhere. Also, the title of Fig 4 should be correct from “Correlation analysis of SIF, NDVI, and EVI with sc_PDSI at Yunnan site” to “Scatter plots of SIF, NDVI, and EVI with sc_PDSI at Yunnan site”, because the statistic you present in this figure is the R-squared.
- Line 177, Eq. 7 – GPP =SIF_total * Si. What is the SIF_total? As you wrote in the previous equation that SIF is determined for each wavelength, is the meaning of the total here is the sum up of the SIF from all wavelengths or all bands? If so, please write it. Else please explain what “total” really means. In addition, you should write that these Si values generated from the procedure you describe in lines 138-146
- In Lines 470-473 you present results in the Discussion part. Please rewrite so the Discussion will use only for interpretation.
Author Response
Dear reviewer,
Thank you very much for your careful work, which has helped us a lot in identifying problems in the manuscript We have carefully revised the manuscript according to your comments. The detailed response to your questions is in purple. Thanks again.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This paper compared the sensitivity of SIF, EVI, NDVI to drought, The results show that SIF has a stronger correlation with drought than other vegetation indices, the accuracy of GPP based on SIF is higher, which can capture the whole process of drought impact on vegetation more quickly and accurately.
The paper is well structured, The research method of this paper is scientific and the conclusion is credible. I think it's worth publishing to readers.
Author Response
Dear reviewer,
Thank you very much for your approval of this article, which improves the possibility of the editor and the journal accepting my article, which is very good news for me, and I thank you again on behalf of our entire team.
Have a nice life!
Author Response File: Author Response.docx
Reviewer 3 Report
This MS discusses "SIF based GPP is a useful index for assessing impacts of 2 drop on Vegetation". Generally, the article better reflects the role of SIF based GPP. However, before publication, there are still some contents that need to be modified and improved:
1) Introduction. At present, the introduction is not well written. Each paragraph lacks a topic sentence, and there is a lack of necessary connection between paragraphs. It is suggested that this part should be carefully revised. And There are many kinds of drought indexes, such as SPI, spei, etc. why only PDSI is selected in this paper? It is suggested to make some summary and difference comparison.
2) For vegetation type data with 1km spatial resolution, using bilinear method to resample to 0.05 degree will cause many problems.
3) At present, there are many SIF based GPP products. Why not use them directly? Can the products made in this paper be compared with these products?
4)The types included five categories: arable land, forest land, grassland, scrub, and other types. Are the effects of drought on these vegetation types the same in the study area?
5) L213 Table 2. The scatter chart or box chart should be more direct and intuitive than the table form. It is suggested to add another comparative chart.
6) L246, eight flux towers, why only four sites are selected for analysis? No specific distribution can be seen in Figure 1.
7) L258 section 2.1?
8) L139 site data 2003-2010, figures 2 and 3 time 2009-2011, Figure 7 time data from 2007-2010, figure 4 data no time. The summary highlights:
The drought has a serious impact on GPP, and the monthly average values of the 25 effect of drop on GPP (gppd) in Yunnan Province in 2009, 2010, 2011.
It is recommended that all data cover 2009-2011.
9) L363. According to 2.3.2?
10) L466, Temperature is the main stress factor for vegetation in this region. Why do you say so? Is it the conclusion of this paper or the result of others?
11) Are there any limitations in this MS? It is suggested to supplement in the discussion section.
Author Response
Dear reviewer,
Thank you very much for your careful work, which has helped us a lot in identifying problems in the manuscript We have carefully revised the manuscript according to your comments. The detailed response to your questions is in purple. Thanks again.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Dear authors,
Please correct the following issues:
Line 198 – what is the Long-Ridge flux tower? In lines 191-192 you mentioned 8 sites, and wrote that Changling flux site was used for verification. Is this site is also called Long-Ridge? Please correct
Lines 366-368 - You introduced Figure 7 prior to Fig 6. the reader should NOT jump back and forth. I suggest you will remove these lines, as you refer and explain this figure from line 374
Line 384 – The beginning of the sentence is missing a word: “The RMSE of the three during the non-drought …..”. I’m assuming the word should be “products”. Please correct
Lines 467-468 – you refer the reader to Fig 10-11 but actually it is now Fig 11-12.
Lines 504-510, beginning of Section 3.3.3 – from Figures 11 and 12 in Section 3.3.2, it is clearly seen that all the GPP products do follow the drought curve, especially in summer of 2010 (as I wrote in the first review). To lead the reader to this 3.3.3 Section, I recommend to re-write this sentence. For example, say that even if the GPPsif did not respond well to the drought (Section 3.3.2), you further evaluate the difference between normal month to drought month to see if this change will present the drought accurately.
Lines 610-611 – You wrote: “…indicating that GPPSIF responded well to drought.”. there are two issue to correct here: please remove the word "well" (R2=0.46 is not good) and change the GPPsif to GPPd, so the end of sentence should be: “…. indicating that GPPd respond in some degree to the drought”.
Author Response
Dear reviewer, Thank you very much for your meticulous work, which corrected many problems for me and greatly improved the accuracy of the content of my manuscript and the possibility of its acceptance by the journal. I will learn from your rigorous and serious attitude in my future research work, thank you again! Best wishes, Chuanhua Li, on behalf of all authors.Author Response File: Author Response.docx
Reviewer 3 Report
For vegetation type data with 1km spatial resolution, using bilinear method to resample to 0.05 degree will cause many problems. Now you said that for vegetation type data you used resampling by the nearest neighbor method. In fact, the nearest neighbor method still has problems. Different methods will have different results. Which method did you use? If there are 25 pixels, 24 of which are class A, but the nearest one is class B, which category do you think they should belong to after resampling?Author Response
Dear reviewer,
Thank you very much for your meticulous work, which corrected many problems for me and greatly improved the accuracy of the content of my manuscript and the possibility of its acceptance by the journal. I will learn from your rigorous and serious attitude in my future research work, thank you again!
Best wishes,
Chuanhua Li, on behalf of all authors.
Author Response File: Author Response.docx
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
