Spatiotemporal Comparison of Drought in Shaanxi–Gansu–Ningxia from 2003 to 2020 Using Various Drought Indices in Google Earth Engine
Round 1
Reviewer 1 Report
I propose that the topic be rephrased a little bit (to something explicit around droughts and indices for assessing droughts). This is because the manuscript focuses on investigating the Spatio-temporal characteristics of drought intensity and frequency in Northwest China from 2003 to 2020 using the Standardized Precipitation Index (SPI), Vegetation Condition Index (VCI), Temperature Condition Index (TCI), Vegetation Health Index (VHI), Normalized Vegetation Supply Water Index (NVSWI), Soil Moisture, Condition Index (SMCI), and Soil Moisture Agricultural Drought Index (SMADI). As is, the topic is currently misleading. Some indices showed a wetter period, others depicted a drier period.
Overall, I felt that the manuscript was unbalanced; there are no strong research questions as well as why this study has to be done with all these indices. The focus of the manuscript is blurred. In the beginning, one gets the impression that it is focusing on “trends and impacts of a wetter climate in the Shaanxi- 2 Gansu-Ningxia region from 2003 to 2020”. After reading the introduction, one gets the impression that it is focusing on assessing the ability of each as well as comparing performances of the 9 indices for assessing the Spatio-temporal characteristics of drought intensity and frequency in Northwest China from 2003 to 2020. Thus, the manuscript still needs to mature. The methods and materials section has lots of mathematical details about the relevant indices. These details are already published in the literature. Thus, only concise summaries should be provided in this section. In my view, published or textbook equations details do not add much value to this manuscript. It is also not clear what exactly are the research questions driving this analysis including the choice of 9 indices. I propose that for each analysis step, there should be a clear research question or objective under investigation and the steps followed during such an analysis should be explicit. Furthermore, no quantitative results were reported in the results section [p –values missing, etc., no statements of effect sizes, or whether the observed trends or patterns are significant or not]. Thus, inferences made appear entirely speculative or qualitative
Abstract:
Some method summary details are missing. For example, how were the comparisons done? There is a need to provide a number around effect size is a number measuring the strength of the relationship between indices or at least just some p-values for readers to tell if the performances were significantly different or not.
It is also important to provide information on whether differences (wetness or dryness) between the periods assessed were statistically different. Broadly, the take-home message from this work has to be re-worked again. The Abstract is not clear on which indices are better or worse for their purposes.
Introduction:
This section is poorly presented. I suggest rewriting this section more clearly outlining:
(1) What the problem is
(2) Why it’s important and
(3) What are the gaps in knowledge being addressed?
(4) What are the proposed research questions/objectives?
Line 38 -39, there is repetition. Keep one of the two sentences.
The justification of using all these 9 indices is required. Seems to me this is data dredging.
A listing of the indices in this section and their relevant authors does not advance much in the understanding of this manuscript. I propose outlining the merits/advantages and/or disadvantages of each index for specific purposes. This should give context as to why it is important to investigate these indices for his area.
Materials and Methods:
Section 2.2.1 lists several data sources used in this study but provides no details of why all these data are required.
In section 2.2.4, more details should be provided on why Pearson’s correlation.
“To compare the applicability of different drought indices for drought monitoring in the SGN area, we selected the Palmer Drought Severity Index (PDSI) based on the TerraClimate datasets as the reference data and mapped the spatial distribution of the Pearson correlation coefficient of each drought index and PDSI”. Provide clarity on why PDSL based on the TerraClimate datasets was chosen as the reference point.
“To further explore the impact of drought on agricultural productivity, we separately analyzed the correlation between the annual average drought index of the cropland in the SGN region and the per hectare yield data for grain crops”. It is not clear anywhere in the manuscript what the question under investigation here is. What is driving this part of the analysis?
Section 2.3.1. Authors should make this section concise. Most of the mathematical details are already published elsewhere.
Results:
This section needs to be more focused. The results should be presented as guided by your research question/objectives.
Provide details around defect size or at least just report p-values. At least provide the correlation coefficients and related p-values here.
Report on some details concerning the outcome of the comparisons of the indices
“We also compared the relationship between the annual average for each drought index and annual grain productivity (Figure 6). The manuscript reports significant changes to certain aspects’’. It seems to me that a lot of analyses were done without clear research questions or clear motivation provided in the manuscript introduction section)
Discussion:
The discussion section is not comprehensive; it should clearly establish the link between your introduction, the aims/objectives, and your results. In other words, it should answer objectively the question of whether a specific objective had been addressed or not, and provide the meaning of the findings in the context of the aim (s). I propose that the authors need to:
- Assess their findings in light of their objectives/questions.
- They should tell readers whether findings contradict or support what has been found/published by others doing similar work.
- Then provide a take-home message from that assessment.
It is clear that the absence of clear research questions leads to considerable conjecture in the discussion section. This fact needs to be addressed.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The aim of the work is understand the causes of changes in drought in Northwest China by analyzing the spatial-temporal characteristics of drought intensity and frequency in Northwest China from 2003 to 2020. The manuscript is very well written in all its parts. The bibliographic review is treated fairly well, making the scientific background of the manuscript solid.Materials and methods, results, discussions are also clear and well explained.
Only the conclusions need to be improved because they show only a summary of the work instead of dealing with the problem at a more general level, that is, to pull the strings of the discourse and offer a more global perspective of the results achieved.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
GENERAL COMMENTS:
This paper analyses the recent historical trend of climate in the Shaanxi Gansu-Ningxia region and its impacts on meteorological and agricultural droughts. The subject is of potential interest for the Remote Sensing readership. Nevertheless, in my opinion, there are some issues that should be addressed before its potential publication:
INTRODUCTION:
The introduction has to be improved. The authors should make an effort to clarify the novelty of the paper, from a methodological point of view in a broader context. They should include a more clear definition of the knowledge gaps, indicating the main novelties of the proposed objectives and their relevance.
DISCUSSION: In the discussion should be also improved.
I suggest to include in the discussion a limitation section where the hypothesis assumed and the limitation are summarized, making references to other previous works.
For example, is the period analyzed long enough to withdraw draw conclusions about drought statistic from a statistical point of view? Previous studies point out the necessity to analyze long periods to withdraw conclusions about drought statistics, in the analyses of historical periods (Pedro-Monzonís et al., 2015) and future horizons (Collados-Lara et al., 2020).
How do the author explain this trend to a wetter climate? Should it be related to climate change impacts? The author should include any comment about it. Many recent research work analyze the impacts of climate change on droughts in different regions (Tramblay et al., 2020).
REFERENCES:
Collados-Lara; AJ, Pulido-Velazquez; D. Pardo-Igúzquiza, E., 2020. A statistical tool to generate potential future climate scenarios for hydrology applications. Scientific Programming 2020(4) DOI: 10.1155/2020/8847571
Pedro-Monzonís, M., Solera, A., Ferrer, J., Estrela, T., and Paredes- Arquiola, J., 2015: A review of water scarcity and drought indexes in water resources planning and management, J. Hydrol., 527, 482–493, https://doi.org/10.1016/j.jhydrol.2015.05.003, 2015.
Tramblay, Yves & Koutroulis, Aristeidis & Samaniego, Luis & Vicente-Serrano, Sergio & Volaire, F. & Boone, Aaron & Le Page, Michel & Llasat, Maria & Albergel, Clement & Burak, Selmin & Cailleret, Maxime & Cindric, Ksenija & Davi, Hendrik & Dupuy, Jean-luc & Greve, Peter & Grillakis, Manolis & Hanich, Lahoucine & Jarlan, Lionel & Martin-Stpaul, Nicolas & Polcher, Jan. Jordi Martinez-Vilaltaq,r, Florent Mouillote, et al. (2020). Challenges for drought assessment in the Mediterranean region under future climate scenarios. Earth-Science Reviews. 210. 10.1016/j.earscirev.2020.103348.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
"To compare the effectiveness of each drought index, we calculated the Pearson correlation coefficient between each drought index and Palmer Drought Severity Index"(PDSI).
1) Indicate why the PSI was used as the point of reference
The wetting trend of VCI index characterization is the most obvious (R2 = 0.9606); During the period 2003.
SPI-12 showed the best correlation with PDSI, with R values greater than 0.6 in almost all regions. SMADI had the weakest correlation with PDSI, with R values ranging between -0.4~-0.2. The results of this study clearly clarified the spatiotemporal variation characteristics of drought in the SGN region from 2003 to 2020 a
2) Indicate P-values in the above statements.
3) Whilst attempts are acknowledged, the authors still need to provide more convincing arguments around the use of the 9 indices (Why all the 9?)
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 3 Report
The authors have properly addressed my comments.
Author Response
Thanks very much for your kind work and consideration on publication of our paper. On behalf of my co-authors, we would like to express our hreat appreciation to your work.
