Drought Impacts on Vegetation in Southeastern Europe
, Célia M. Gouveia 1,2, Ana C. Russo 2, Roxana Bojariu 4, Sergio M. Vicente-Serrano 5 and Ricardo M. Trigo 2,6
Round 1
Reviewer 1 Report
This paper addresses the important topic about droughts and impacts of droughts on vegetation productions in southeastern Europe. Using satellite-derived vegetation indices and drought index, this work manifests where and when droughts exert direct and strong impacts on vegetation productivity. Combining land cover type product, the study found that croplands is the most affected land cover, while needleleaf forest is less directly affected by droughts. Thus, the results provide profound implications for regional agriculture and irrigation plans and developments.
While the subject of the paper has merit, there are a few concerns regarding the methods and results that need to be addressed before acceptance. Additionally, the writing is a bit disorganized and repetitive at times. Details of major and minor concerns are given below.
Major comments:
- The study conducted linear correlation analysis between NDVI and SPEI at 5 different timescales (1-, 3-, 6-, 9-, and 12-). It is not clear that whether is 1/3/6/9/12-month mean or is SPEI 1/3/6/9/12 months ago? According to the results, it reads like 1/3/6/9/12-month mean, if so, it would be appreciated to provide more detailed method about the calculation.
- In general, there is a 1-1.5 month lag correlation between NDVI and precipitation, which suggests that there would be changes in SPEI first, then this waster stress will shown in NDVI variations. How the results now manifest this lagged correlation ship between SPEI and NDVI in the study region?
- When studying the concurrence of vegetation stress and drought, the study quantified the concurrence as the percentage against the number of NDVIanom25. It seems more plausible to calculate the concurrence as the percentage of the number of NDVIanom25 to the number of SPEI days (NDVIanom25/SPEI) instead, for a drought event is identified based on the SPEI.
- The study also conducted vegetation responses to drought for different land cover types. Do the authors have any idea that whether human activities would affect vegetation production or drought events? For example, urbanization (land cover changes) would suppress vegetation greenness as well.
- According to the method section, there are 3 types of drought defined. Whether there are any results showing vegetation responses to different drought categories? Or the concurrence of vegetation stress and drought vary due to the severity of drought?
Minor comments:
Line 51: ‘monitoring vegetation’
Line 56: ‘at local, regional, or global scales’
Line 60-62: ‘The advantage of using multiscale drought indices is their ability to take differences in the response time of vegetation to water deficit into account.’
Line 68-69: The sentence ‘For example…a longer period tome’ is hard to follow.
Line 71-73: ‘In forests, a high percentage of monoculture plantations or semi-natural forests makes these regions less resilient to climate variability than natural forests.’
Line 79: To say ‘by agriculture is negative and becomes even more so’, do author mean the water deficit is more severe? It’s better to rephrase the sentence.
Line 81: ‘in recent climate in SEE’
Line 94: ‘using SPEI and satellite-derived NDVI at several time scales’
Line 96-97: ‘to study the impact of an extreme drought event on vegetation activity’
Line 100: ‘is located at 43°N - 49°N, 20.2°E - 30.1°E’
Line 102: What is the time period of the land cover type product? Does it only cover 2009?
Line 105: ‘representative land cover types’
Line 106: replace ‘land cover classes’ with ‘land cover types’
Lien 116: ‘the most common’
Line 118: ‘It is evident that forest types changes along with altitude’
Line 120: ‘It must be addressed’
Line 142: ‘to 2014’
Line 147: replace ‘disseminated’ with ‘extracted’
Line 169: ‘is particularly suitable to evaluate’
Line 174: ‘computed at monthly time step’
Line 182: The sentence is hard to follow.
Line 199: ‘thus should not’
Line 211: ‘vegetation stress’
Line 224: ‘The drought event in 2000/20001… the most severe drought during 1960-2013.’
Line 225: ‘not only due to’
Line 227: ‘particularly evident in Romania’
Line 228: ‘caused reduced crop’
Line 230: ‘resulted in’
Line 237-239: The sentence is hard to follow.
Line 265: ‘the higher concurrence’
Line 386: ‘relationship’
Line 514-521: The results are more evident about the positive correlation between SPEI and NDVI. Thus, it would be better to rephrase the points to not address the negative correlations.
Tables and figures:
Table 1: It is conventional to use ‘broadleaf’ and ‘needleleaf’. Please indicate the year for the analysis, i.e. land cover types during 2009 or during 19xx-20xx.
Figure 1: Which dataset was used to make the plots? GIMMS or SPOT? Please indicate in the figure caption.
Figure 2: To show the significance of the trends, it would better mask the parts not pass the significance test or make the parts pass the test.
Figure 3: The caption for this figure is not clear. And what’s the unit of the value in the figure?
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The paper study the impact of drought measures by different time-scales of SPEI index on vegetation activity by analyzing the NDVI anomalies activity over southeastern Europe (SE).
The manuscript is ok both by the area analyzed and the period used, It is also well written. But, I have a major concern that the authors should address.
Major issues.-
For defining drought, the authors used SPEI index with a p(x< -.084)=20% as a threshold to account for drought. But in the case of vegetation were careless in describe vegetation stress and no drought. They used an NDVI anomaly of -0.025 as a threshold for vegetation stress. I am not much confident with this index and threshold cause two issues:
1) it is arbitrary due that was defined for the 2005 drought in Portugal (Gouveia et al., 2009). Thus, not representative of the stress all around SE and either for the whole period (1998-2014). This anomaly, as stated in the manuscript, was calculated by subtracting the NDVI to its mean. Then the NDVI< -0.025 well for vegetation stress on Portugal for 2005 but not for SE and for 1998-2014.
2) I will ask the authors to change the NDVI< -0.025 for the z-index of NDVI (zNDVI). Thus, you can change the vegetation stress by drought and will ensure that the vegetation status was under its historical mean. Also, it will be a standardized value that allows for comparison with other parts of the world.
If the authors do not make this change, I will ask for the next revision to provide a comprehensive explanation of why not. Please address the arbitrary features of the NDVI < -0.025 anomaly.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The article's merit concerns an important problem of studying the influences of diverse climate factors like drought or precipitation index on the vegetation. The elaboration has scientific soundness and presents the problem in a professional way.
The material presented in the article is quite extensive. There are many clues and contents deeply analysed, discussed and appropriately concluded. However, while reading the article one can have an impression, that it may sometimes be overloaded. Such a huge amount of details especially presented in the discussion unit can sometimes be overwhelming. Generally, the discussion would need to be shortened. An overall impression is that the authors provide a reader with lots of in-depth studies, analyses and details which sometimes lead to predictable, rather known conclusions. Of course, the problem presented is important and interesting, especially considering such a long time series analysed. Nevertheless, the results obtained confirm the state-of-the-art. In my opinion, it should be underlined, that the authors' assumptions (ex. the potential influence of drought on the vegetation etc.) had been confirmed and supported with modern, advanced, remote-sensing based analyses.
Technically, the material is well prepared and contains only a few mistakes like blind spaces or missing commas. Hence, proofreading is recommended before the text will be published.
All figures need information about the sources (own elaboration or based on someone's analyses etc.).
Line 48 Does it concern the authors' studies?
Line 100 would need an extra explanation about the study location (general map showing the investigated area at least in Europe).
After introducing all necessary changes, the text can be considered for publication.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The manuscript and content have been largely approved. The authors have worked hard to address the questions and made recommended changes. The comments on the latest version of the paper only need to quickly correct some minor linguistical mistakes:
Line 73-74: This sentence still reads not correct. I suggest to rewrite it as 'In the forest regions, there are patches consist of high percentages of monoculture plantations or semi-natural forests. These areas are less resilient to climate variability than natural forests.'
Line 183: Instead 'accordingly with this definition', it would be better 'Considering the return periods of different droughts, the ...'
Author Response
Reviewer 1
1 - The manuscript and content have been largely approved. The authors have worked hard to address the questions and made recommended changes.
Response 1 – We thank the Reviewer for the positive assessment of our work.
The comments on the latest version of the paper only need to quickly correct some minor linguistical mistakes:
2 - Line 73-74: This sentence still reads not correct. I suggest to rewrite it as 'In the forest regions, there are patches consist of high percentages of monoculture plantations or semi-natural forests. These areas are less resilient to climate variability than natural forests.'
Response 2 – We have made the proposed change.
3 - Line 183: Instead 'accordingly with this definition', it would be better 'Considering the return periods of different droughts, the ...'
Response 3 - We have made the proposed change.
Reviewer 2 Report
No more comments. All have been addressed.
Author Response
We thank the Reviewer for the positive assessment of our work.
Reviewer 3 Report
The authors have introduced all the recommended improvements and responded to the reviewer's questions sufficiently. Hence, the text, in its new outlook, can be published. Congratulations!
Author Response
We thank the Reviewer for the positive assessment of our work.
