Spatio-Temporal Evolution of Inland Lakes and Their Relationship with Hydro-Meteorological Factors in Horqin Sandy Land, China

Round 1

Reviewer 1 Report (Previous Reviewer 2)

The revised manuscript has been significantly improved in response to reviewers’ comments. The revised manuscript is now suitable for publication.

Author Response

We would like to express our sincere appreciation for the positive feedback and affirmation you provided on our manuscript. We are also grateful for the thorough and constructive comments and suggestions you offered during the review process, which have significantly enhanced the quality of our manuscript and made it more suitable for publication in Remote Sensing.

Reviewer 2 Report (New Reviewer)

Dear authors pls refer to the PDF attached for my comments. 

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

We feel great thanks for your constructive review work on our manuscript. According to your nice suggestions, we have made extensive corrections to our previous manuscript, the detailed corrections are listed below.

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report (New Reviewer)

Dear Authors,

your work is interesting and might be of potential interest also for scholars working on lakes outside the study region. However, how to translate your approach/results to other contexts should be better emphasized in the Discussion section.

As you can see below, I have a number of comments that should be addressed before publication. I would like to stress the need for better shaping both the Introduction and the Discussion, to show what is new in your work (besides applying well-known methods to another case study) and what we can learn from your work.

- add the study country in the title and the abstract

- in the Abstract, please add more details on the Landsat data (what mission?)

- line 97: SWOT and not SWTO

- I suggest rephrasing the Introduction by moving from the international picture to the local one, pointing out what are the major challenges in monitoring lakes first, and then what is missing in the study region and why your work is timely and needed. This can help in stressing the novelty of your investigation

- please provide high-quality figures. In the present quality it is impossible to read them

- I suggest combining sections 2 and 3 in a single "Materials and Methods" section
- please add some references for eqs 1 and 2 (i.e., have you developed them or are they taken from the literature?)

- line 231-237: please provide more details on how you selected the control lakes and the control points. Adding a map with such information might help in better evaluating the quality of your check

- in Table 1, please check the spelling (the first column should read "Elevation [m asl]")

- please check the numbering of the tables. You have Table 1 twice (line 277 and line 301)

- in Figure 6, please name the panels, otherwise, it is hard to understand what is reported in the caption

- at line 470 you wrote "With the strengthening of human activities..." How can a reader see such an increase in human pressure? I suggest adding more comments on that, potentially showing the increase via a figure. Maybe the correlation between changes in lake surface and human presence can be better addressed and moved to the Discussion section

- providing the GEE scripts in an open repository can help in guaranteeing the study's reproducibility, which is needed in scientific writing

 

Author Response

Dear Reviewer,

We sincerely appreciate your positive comments and valuable suggestions to improve the quality of our manuscript. We will respond point-to-point carefully. Hope it evolves into a publishable article. The detailed corrections are listed below.

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report (New Reviewer)

Dear Author,

thank you for having addressed the majority of my comments.

I am still convinced that the Discussion section can be better developed, to further show what is the novelty of your work if compare with the actual state of the art, for eventually increasing the impact of your study.

Author Response

Dear reviewer,

We would like to express our sincere appreciation for the positive feedback and affirmation you provided on our manuscript. We have revised the Discussion section of the manuscript to increase the impact of our study (lines 489-492, lines 513-517, lines 528-530, lines 547-549, lines 569-571 and lines 580-583). The GEE platform not only enables rapid access and acquisition of publicly available remote sensing images and datasets but also facilitates the integration of multi-source data and the extraction of lakes within the platform. This approach offers enhanced speed, convenience, and applicability for reference and utilization in other regions. Furthermore, our research methodology presents novel insights for comprehensive lake monitoring in areas lacking hydrological data. Additionally, our long-term and continuous monitoring endeavors provide a more precise understanding of the evolving patterns exhibited by small lakes in the region and can aid in formulating effective strategies for the sustainable management of water resources. The outcomes of our study furnish a regionally representative scientific foundation for investigating global surface water changes. While previous investigations have explored the interplay between meteorological elements and lake dynamics, uncertainties persist. Through quantitative and qualitative analyses, we have elucidated the potential relationship between hydro-meteorological factors in the region and the sandy inland lakes, thereby addressing some of these uncertainties. We are grateful again for the thorough and constructive comments and suggestions you offered during the review process, which have significantly enhanced the quality of our manuscript.

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.

Round 1

Reviewer 1 Report

Thanks for the opportunity to review this manuscript. The topic (remote monitoring of lake surface area and change in volume, plus connection to meteorological processes) is interesting and important. The authors have clearly done a lot of work on this, and I would like to see this evolve into something that's ready for publication. In my comments below, I will focus on weaknesses or problems in the manuscript that would need to be addressed:

A. Overall, the manuscript tries to cover too many things without fully covering any of them - there is the monitoring component (surface area, volume change), the meteorology analysis, a sort of qualitative comparison of the lake monitoring data to the meteorology, a wavelet-based quantitative comparison of the lake and meteorology data, and then another comparison that appears in the Discussion section. I would recommend trimming the breadth in order to focus more in depth on the important parts.

B. The lake monitoring parts of the paper (surface area and changes in storage) are interesting, but there is a very large literature on this topic that is mostly not cited, using many different optical sensors for lake surface area and a variety of methods to estimate water level, and then volume by combining area and water level. I think it would help the authors to review some of the main papers over the past couple of decades that have done similar analysis of area/water-level/volume/storage in lakes in different parts of the world. The same is true for the literature review of lake/meteorology interactions - the papers cited (refs 26-30) don't really represent the body of significant work in this field.

C. The descriptions of the Landsat data analysis process need to be expanded, made more complete, and clarified:

  C1: Lines 141-143 refer to downloading 2777 images from GLOVIS, but later in that paragraph (and elsewhere) there are references to the analysis being done in GEE (Earth Engine). If the analysis was done using GEE-hosted imagery, why were images downloaded through GLOVIS?

  C2: Line 189 the extraction rules seem confusing to me ... (NDWI > 0) and (EVI < 0.1) are clear, but I'm not sure what is meant by the contents of the parentheses at the end of the line (MNDWI > EVI or MNDWI > NDVI).

  C3: I take it that seasonal vs permanent lake status was based on WF within individual calendar years. Were the images cloud-screened, and if so, how? There is very little description of the image processing methods, yet these are extremely important. If the authors used the "Land Cloud Cover" or "Scene Cloud Cover" metadata in GEE to filter out cloudy images, that can be unreliable, leading to lakes "disappearing" and "reappearing" due to clouds ... for example, at the Horqin site of this paper, in the first year (1984), one image (1984-07-26) is listed at 7% cloud cover but is really almost 100%.

  C4: Likewise, how many images were used each year, and how were they combined? Two WRS-2 paths (119 and 120) cover the region, but many images only cover part of the area. I would recommend including a table with the # of images used per path/row, per year (1984, 1985, ... 2021) and a description of how they were aggregated to what appears to be a single lake area value per year.

  C5: The material on p. 6 between figures 2 and 3 is very confusing to me. It sounds like the authors generated contour lines from the SRTM DEM at 1-m intervals above the Feb 2000 water level of each lake (is that what the authors mean by S[SRTM]? The lake area at the SRTM water level?) This allows increases in water storage to be calculated, but it doesn't work for lakes whose water levels drop below the level of SRTM.  I am *guessing* that the authors somehow either (a) extrapolated those changes in storage for each lake downward below the SRTM level (on a lake-by-lake basis, until storage is 0) or else (b) combined data on storage change per 1-m band from lakes with different surface areas ... is this what figure 3(b) shows? Are those data points from the same lake, or different lakes?  

This latter point is extremely important - the authors are applying some kind of hypsographic curve to calculate changes in water storage, but don't really explain all the full details of how it was derived. If it was created by combining area/change-in-storage data from different lakes, that would seem very unreliable, because different lakes have different basin morphometries. It is critically important to know whether the hypsographic curves being used to relate area and water level (storage) are accurate, because all the analysis of water storage change depends on this.

This is particularly important because there is no validation or accuracy assessment.  If there are no in-situ data available on water level change for specific lakes, perhaps the authors could look for lakes where ICESat-2 altimetry data are available?  

D. The validation issue applies to lake surface area data as well. The authors do not appear to have attempted to assess the accuracy or level of error in the lake area time series. Previous studies have done this various ways, e.g., by measuring area of selected lakes in much higher-resolution imagery and comparing them to the automatically estimated area from a Landsat image close to the same date. Lyons et al. (2013) have a nice overview of typical errors in remote sensing lake mapping studies (https://doi.org/10.1080/01431161.2013.827343).

E. The authors state that lake area fluctuates more in the pre-2002 period than the post-2004 period. Aside from one outlier year (looks like 1995?) when lake area and number of lakes spikes, is the difference in variance between those two time periods meaningful?

F. More clarity is needed on the time series of water storage changes (e.g., figure 6). What is the "change" for each year relative to? Are they 1-year changes from the previous year (in which case storage increased every year), or changes from the SRTM level (winter 2000), or changes from the first year in the series (1984)? None of the years are at 0 (neither 1984 nor 2000) so I'm confused about what the baseline is for measuring change.  And, again, the authors need to explain how storage change was calculated when water levels dropped below the SRTM elevation.

G. Somehow an additional analysis - with its own methods and results - has worked its way into the discussion section (lines 495-513). If that's included in the paper, it should probably be presented in the methods and results sections.

Other minor comments:

H. Line 27 and elsewhere - I would recommend not using "randomly" to refer to distribution of lakes, since it is probably not truly random (dependent on surficial and bedrock geology, etc.)  Maybe use "irregularly"

I. Line 47: Are you certain this is true of "most inland lakes" globally? Cited source is one very specific location, may not be true everywhere.

J. Lines 78-84: Actually, I believe that previous studies have shown it is possible in many cases to classify lakes with high accuracy using single metrics, particularly in relatively flat regions and arid regions.

K. Maybe mention the newly launched SWOT satellite in the introduction, because it will make doing this kind of work much easier and more effective in the near future.

L. Line 116: This section is numbered 3.1, but is within section 2. Actually, the description of the study area probably belongs at the end of section 1, since it is not "Materials and Methods".

M. Lines 120-122: How is annual mean ET 4X larger than annual mean precipitation? Be clear this is potential evapotranspiration. Also, I think many of these lakes are shallow, but if any of them are deep, actual evaporation rates from the lake may be much lower than as measured in the shallow dish.

N. Line 172: "Ice-off period" rather than "unglaciated period".

O. Lines 238-239: I would drop the description of wavelet methods in the first sentence ("mathematical microscope" ... "focus on any details of the object") and go straight into the following setnence ("...widely used to analyze the periodicity...")

P. Figure 4 and Figure 6 - avoid using 2 Y axes; split each into two figures, with a single Y axis on each one.

Q. Line 286: There is a reference here to "analyze the spatial differentiation characteristics" in 4 years, but I don't think there is much analysis, just mapping them in Fig 5. Maybe "... were selected to visualize [or map] the spatial distribution of lakes (Fig 5)."

R. Line 287: "As shown in Figure 5, lakes resemble stars or dots..." - I believe the authors are not showing the actual shapes of the lakes on the map, just circles of different sizes based on lake area. I am not sure what the comment about "resemble stars or dots" is referring to.

S. Figure 9 and accompanying description - this is the wavelet analysis relating meteorological and lake measurements. The authors either should drop this part of the paper, or provide a substantial additional explanation of the figure and the analysis behind it. Are the periods (Y axis) years? I am generally skeptical of wavelet analysis on datasets with small numbers of points (38 years?). The authors have already provided a lot of qualitative interpretation of the relationship between P/ET and water storage, and there is more to come in the discussion. I think an unconvincing and not deeply explained wavelet analysis doesn't add a lot of value to that.

T. Line 426: I do not think "most" is correct here. There are many studies around the world that have looked at large numbers of images of lakes to estimate lake surface area time series. The two cited at the end of this sentence may have only used a few, but that is not representative of most prior studies.

U. Lines 480-483: I think this is an important point that the authors need to be clearer about in earlier sections of the manuscript. How are they relating P/ET and water storage?  It is probably change from the previous year that relates to the meteorology.

Reviewer 2 Report

I found the topic of this paper very important, but there are some things that could be improved.

 

Here are many details:

1.        Abstract: Please use the correct expression for square kilometers (km² not km2).

2.        Introduction (Lines 94-105): In this paragraph, the authors listed many related work. Here, I think, the authors lack a brief introduction to the development of this work.

3.        Figure 1: Remove the pointer and scale from the above two panels in this figure. It is meaningless here.

4.        Formulas 1 and 2: Add references.

5.        Lines 193-194: Why 0.25 and 0.75? Reference [34] does not seem to mention the threshold here. Please add more details.

6.        Figure 2 and Lines 216-232: This part is very interesting. But how much does the vertical accuracy of SRTM affect this result?

7.        Figures 3, 4, 5, 7, and 8: The resolution of those Figures does not appear to reach 300 dpi.

8.        Figures 5 and 7: Same with comment 3.

9.        Lines 374-375 and Figure 8: Add the slope and confidence interval of the trend.

 

I recommend minor revision.