Spatial and Temporal Characteristics of NDVI in the Weihe River Basin and Its Correlation with Terrestrial Water Storage
Round 1
Reviewer 1 Report
Review of manuscript number remotesensing-1982335
General note:
The authors aimed to find the relationship between vegetation and water storage in an inland basin using the NDVI to analyze the spatiotemporal characteristics of vegetation changes, and then adopted two models to derive water storage changes in the study area. the methodology is straightforward and it was implemented correctly. However, the manuscript lack of novelty, some comments are found as follows:
Specific notes:
Abstract:
· Please add a sentence showing the data used and the duration (time frame).
· Give a line or two to briefly introduce the models.
· Line 58: make it a new paragraph.
·
Introduction:
· Line 36 “on the other hand” >>> meanwhile,.
· Line 39 - 40 “on the other hand, vegetation …… [10]” >>> It is also essential in terrestrial ecosystems is or regulating water balance at regional and global scales.
· Lines 80-106 is some kind of “fill in” paragraph, please revise and incorporate properly in the introduction. You can remove the words about WRB and then include them in the study area section.
Methods
· Please provide statistical measures of the average annual rainfall, temperature evapotranspiration…..
· What are the main vegetation types and cultivation?
· Line 130: Its geographical location is shown in Figure 1. Delete replication
· Line 140: add: Weihe River basin (the study area)
· Line 150: what is GSM?
· More description of the GRACE – FO is needed. What is its spatial resolution
· Line 163: MODIS is a low spatial resolution data.
· Line 184: of this paper >>> this study.
· In the trend analysis, using the ZERO as a reference thresholding value, might be misleading. The values of NDVI change but the vegetation type may not. So, it is preferred to used a Range (from-to) for thresholding the classes of change. Mean while that section has no references to support the thresholding.
· What was your base to classify the F-test classes. Needs justification and references.
Results:
· Figure 3: the months should be added to the graph to match its description in lines 222-227.
· Line 232: what is the multi-year average? Is the annual average.
· Lines 254-257: delete and start by Table 1…. You mentioned that you will use the trend analysis in the methodology part.
· Figure 5: add the time frame to the caption.
· Table 1 and Fig 6: did you figured out any “no change, Slope = 0) in the study area. please refer to the previous comment above.
· Figure 13: provide explanation of the obvious variation in the years 2016-2020.
· Most of the results lack of a discussion.
·
Conclusion:
· This is not a conclusion
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The authors presented a case study in the WRB focusing on the NDVI and TWSA and their spatial-temporal relationships. The vegetation change in the Loess Plateau has been widely noticed in previous studies related to climate change and forestation. This study provides an insight from the water storage change. It would be interesting since water availability is the one of the major driving factors that regulating vegetation change. The analysis is well implemented especially with detailed information on NDVI. However, challenges generally remain in similar studies due to: 1) The mismatch of spatial scales of NDVI and TWSA, and 2) The driving factors that beyond TWSA. My detailed comments include:
1. Title: ‘Vegetation’ is not necessary since NDVI already indicates vegetation (V).
2. Figure 4: It might better to add standard deviations (error bar) for the annual mean NDVI.
3. It seems strange that the TWSA time series in Figure 9 and Figure 13 represents large differences. Besides, Figure 9 seems not necessary if TWSA will be provided in Figure 13.
4. Figure 13: Abrupt changes in TWSA since 2015, while NDVI remains stable change. Why? More interpretation is needed if additional information can be provided.
5. Since the study area is largely covered by cultivated plant, how would human intervention such as crop farming and irrigation affect the NDVI beyond TWSA?
Others:
1. Figure 1: in the legend, should be tributaries and not ‘V tributaries’.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The work is well written, it is clear and meets the structuring requirements of the journal's scope.
We emphasize the use of data from the GRACE project with the NDVI (MODIS) index.
There are some points that can be improved. The suggestions are to facilitate communication with readers. I make the following considerations.
1 – In the summary, indicate the location of the Weihe River basin. In view of the international audience, not everyone is familiar with the given description. In this case, mention that it's in China.
2 – On the issue of evapotranspiration. It would be up to the authors to discuss the limits of associating evapotranspiration with NDVI. The authors address the issue punctually in the introduction. I suggest referencing the reasons for not using evapotranspiration in the study.
3 – In the same sense, what are the reasons that led the authors not to use other spectral indicators that take into account the water factor in the soil, as is the case of SAVI?
4 – I consider that there was no adequate explanation for the phenomenon indicated in Figure 13. I refer to the period from 2016 onwards. The inversion of the dynamics of the TWS anomaly and the NDVI in this signaled period is surprising. Would it be possible to address more details that help to identify the phenomenon?
5 – Conclusion 2 indicated by the authors is not properly explained. The authors conclude that there are more factors to the water storage issue. But what would these factors be?
Author Response
Please see the attachment.
Author Response File: Author Response.docx
