Retrieval of Suspended Sediment Concentrations in the Pearl River Estuary Using Multi-Source Satellite Imagery

Round 1

Reviewer 1 Report

Overview and General comments:

The manuscript titled, “Retrieval of Suspended Sediment Concentrations in the Pearl River Estuary using multi-source satellite imagery” by Cao et al. demonstrates the seasonal and interannual variations of suspended sediment concentrations (SSCs) in the Pearl River Estuary using a combination of field observations, Landsat imagery, and Sentinel-2 imagery. This study formulates an empirical SSC algorithm for obtaining reasonable maps of SSC from the Landsat imagery. Then these maps are used to investigate the spatial distribution and long-term variability of SSCs. Finally, the effects of runoff, wind, channel dredging and artificial facilities on the temporal and spatial variability of SSCs are discussed. Overall, the manuscript is scientifically sound. However, the Introduction is needed to be better written and organized and the authors should emphasize the novelty of this study. The literature review seems inadequate. I have observed some unreliable results, especially verification of the atmospheric correction algorithm and SSC inversion model. There are also a large number of typos, grammar mistakes, and unnecessary words and sentences at several locations. I believe several conclusions and analyses are unreliable. I will reconsider the manuscript after major revisions. Please see my comments below.

 

Major comments:

Point 1: Introduction. The first paragraph should come straight to the point. The introduction should lead directly to the importance and significance of suspended sediment studies for estuaries rather than rivers. Line31-Line35 is unnecessary and meaningless. The authors should focus on summarizing the work related to suspended sediment studies in estuaries.

Point 2: Introduction. The second paragraph of the Introduction (Line42-Line57) is not well written and organized. This study focuses on the study of suspended sediment in the Pearl River Estuary. However, it is not appropriate to list more relevant studies on suspended sediment in rivers. In fact, there are a lot of studies using satellite remote sensing data to study suspended sediment problems in estuaries, such as the Pearl River Estuary (the study area in this paper), the Yangtze River Estuary (Du et al., Remote Sensing, 2021, https://doi.org/10.3390/rs13050929), the Yellow River Estuary, the Yalu River Estuary, the Río de La Plata Estuary, etc. The authors will take inspiration from the following related studies of these estuaries:

[1] “Changes in suspended sediments in the Yangtze River Estuary from 1984 to 2020: Responses to basin and estuarine engineering constructions”.

[2] “The spatial and temporal variation of total suspended solid concentration in Pearl River Estuary during 1987–2015 based on remote sensing”.

[3] “Using Landsat 8 data to estimate suspended particulate matter in the Yellow River estuary”.

[4] “Application of the Geostationary Ocean Color Imager to mapping the diurnal and seasonal variability of surface suspended matter in a macro-tidal estuary”.

Also, there is no correlation between the contents of line60-line70 and those of line58-line60. In short, the first two paragraphs of the Introduction should be fundamentally rewritten. The literature review in other parts should also focus on estuaries rather than rivers.

Point 3: In the last paragraph of the Introduction, the authors propose three research objectives. I wonder what is the difference between this study and the previous study (Wang et al, 2015, Changes in suspended sediments in the Yangtze River Estuary from 1984 to 2020: Responses to basin and estuarine engineering constructions)? What is new in this paper? Please mention and clarify the novelty in the Introduction.

Point 4: Line96-Line107: I suggest the authors first explain the importance of studying suspended sediment in the Pearl River Estuary and then introduce the research objectives of this paper.

Point 5: 2. Materials and Methods

2.1. Study area:

To let readers better understand the Pearl River Estuary, the authors should give general information about the characteristics of tides, winds, waves (height and period), sediment resource (e.g. riverine, marine, continental), sediment type, typical SSC concentration in the study area, and so on.

Point 6: 2. Materials and Methods 2.2. In-situ data collection

What are the guidelines or protocols for in-situ data collection and normalized water surface reflectance calculation in Section 2.2? Why did you decide to do it in this way and not the other way? Whose methodology did you follow?

Point 7: 2. Materials and Methods 2.3. Satellite data and image pre-processing 2.3.1. Satellite data availability

What are the criteria for downloading satellite images? What was the percentage of cloud coverage of the satellite images used for the analysis? What are the meanings of "1", "2" and "N" in Figure 5? Please clarify.

Is the SSC inversion model used by Sentinel-2 the same as that used by Landsat? Please clarify. You said the two satellites have different spatial resolutions, so how do you complement the missing Landsat images?

Point 8: 2. Materials and Methods 2.3. Satellite data and image pre-processing 2.3.2. Satellite image correction

I can not see any evidence of the effectiveness of atmospheric correction algorithms? Have these algorithms been validated in the Pearl River Estuary? Why did you use different atmospheric correction algorithms? If the atmospheric correction is not accurate, your results are not reliable. What exactly did you evaluate when you say 'model performance assessment'? I didn't see it. To sum up, more evidence is needed for the validity of the atmospheric correction algorithm used in this study.

Point 9: 3. Results

Line300: Table 2 only lists the SSC data for 13 sites. Where are the other data? What point are you trying to make?

Line316-L319: I can't see the results of the ???? and RE. Therefore, in the current state of the manuscript, your results are not credible.

I personally think that the analysis of the impact of river discharge on the temporal and spatial distribution of SSC in the Pearl River Estuary in the Result section should be moved to the Discussion section.

Point 10: 3. Results 3.3.2. Mann-Kendall test results

Only the results |Z| ≥ 1.96 (95% confidence level) are credible, the rest of the results are not to be trusted. Please show and discuss the results of |Z| ≥ 1.96 (95% confidence level).

Reference: ‘Decadal Variation and Regulation Mechanisms of the Suspended Sediment Concentration in the Bohai Sea, China’.

Point 11: 4. Discussion

The Discussion section is somewhat simplified and superficial, and a large number of well-known results are not mentioned. The literature review is biased.

 

Minor Comments:

Line14: SSC should be SSCs. ‘river estuaries’ should be ‘estuaries’.

Line19-Line20: delete this sentence.

Line18: You only established one model, so ‘models’ should be ‘model’.

Line26: significant decrease of SSC.

Line59: Please specify ‘large scale’ and ‘long-term’.

Line92: Pearl River Estuary (PRE).

Line96: ‘Pearl River Estuary’ should be ‘PRE’, same in Line99.

Line131: delete this sentence, it’s well-known.

Line132: ‘filed’ should be ‘field’.

Line135: ‘closing to time’ means 12 hours or a day or whatever? Please clarify.

Line139: R_rs should be in italics.

Line164: Figure 2 is not mentioned in the paper.

Line172: delete ‘the’.

Line174: Missing a space in ‘?_rsis’, same in Line290.

Line184: Figure3: SSCs corresponding to the three curves cannot be distinguished.

Line187: Grammar mistakes.

Line200: delete ‘However’.

Line227: You mean few Landsat images are obtained in other Path/Low?

Line250: ‘Figure 4’ should be ‘Figure6’.

Line287: The position of Figure 7 should appear after the paragraph in the paper in which it is first mentioned.

Line307: Which band does R_rs correspond to in Equation 4? Is it (B3+B4)/(B3/B4)? If not, what are you trying to say in Line205-Line212? It is not clear.

Line312: What is x in Figure 8? I don’t know.

Line316-Line319: These sentences should be moved to Section 2.

Equation 5: ‘i=0’ should be ‘i=1’.

Line344: Does that include data from Sentinel-2?

Line345: ‘Figure 111’ should be ‘Figure 10’. Obviously, the labeling of the figures is completely confusing in this paper.

Having said all that, I won’t point out any more mistakes. A large amount of grammar, spelling, and writing mistakes need to be checked and corrected fundamentally. The author can ask a native English speaker to help solve these problems.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This is a pretty solid manuscript that describes long-term (decadal) trends in SPM for the Pearl River Estuary. While similar papers have been published, this provides useful new information in that it (a) updates the time-series, (b) discusses why things have changed, such as new infrastructure, and (c) merges several satellite products into a single time series. Having said that, the manuscript needs some work, specifically ensuring that the methods are reproducible by others. For example, they develop SPM algorithms for three sensors, but only show the results for one sensor. It’s also a bit uneven. The introduction is quite long and detailed (and could perhaps be shortened) while the methods alternate between too much detail (how to collect above-water radiometry with an ASD) and then lacking some key specifics, such as exactly how the skylight correction was performed. The figures should all also be interpretable from the figure legend without having to read the main text to figure out what is being shown. In many cases not everything in the figure legend is described, or it’s ambiguous as to exactly what is being shown (e.g. Figure 16). I think all of these could be corrected with revision. I don’t see any fundamental issues with the methodology (other than better describing it) so I left it at minor revision. 

 

The English is generally fine but there are occasional odd word choices or incorrect grammar, which could be corrected with a light edit by someone proficient in written English. 

 

Table 1: please provide the appropriate units for runoff and discharge

Figure 1: Elevation (spelling)

 

 

Section 2.2.2. This is both too detailed and yet leaves out important information. Most of the description can be removed and replaced with any of a number of references that describe using an ASD for above-water spectra. What’s missing, however, is how the sky contamination was conducted. Many people use Mobley 1995, but there are more recent updates that refine the sky correction. Exactly how were the data processed? 

 

Line 221: S2 is higher spatial resolution for some bands, but not all of them. Which bands were used, and what was the native resolution? 

 

Figure 6: if you are simply subtracting the residual noise, why does the corrected spectrum increase substantially in the blue relative to the green? Based on equation 3 the correction should uniformly reduce the spectra, not change the band ratios so dramatically .

 

Figure 8: is that for L8 OLI? It would be nice to see either a figure or table of the fits for the other sensors, since the bands and RSRs are not identical. If someone wanted to reproduce these results, it would be necessary to provide all of the regressions.

 

Figure 10: It would be more useful to show the mean and then the deviation (each year minus the mean) which would better show the interannual differences

 

Table 3: what does “Bool” mean?

 

Figure 12: the “turning points” appear to be subjective, but there are statistical tests such as cumulative sum (CUSUM) that will identify significant points in a time-series using a bootstrap analysis. 

 

Figure 14: the cumulative anomaly method is not described anywhere in the methods

 

Figure 16: more information is needed. Presumably these are averages of SSC before and after the bridges were added, but there’s no description of what is actually shown. 

 

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

The manuscript titled " Retrieval of Suspended Sediment Concentrations in the Pearl River Estuary using multi-source satellite imagery" aimed to investigate the spatial patterns of SSC distribution across the Pearl River Estuary, exploring the Spatial-temporal variation in SSC over the past 25 years and analyze the factors of changes in SSC and their relationship with human activities. The study area has great relevance, and this kind of research demands great operational effort to be carried out, revealing precious results for conservation actions in the region of the Greater Bay Area. The results are interesting for publication in the Special Issue “Remote Sensing Observation on Coastal Change”, but the manuscript needs to be more well-written overall (English grammar, spelling errors, unclear sentences). There are two main points throughout the text: some results presented were not mentioned how were processed in the section “Materials and methods" and, need matchups between Landsat and Sentinel-2 data acquisitions and field measurements to validate the SSC values estimated using satellite data. How many uncertainties on estimated SSC values are due to atmospheric corrections? I strongly encourage authors to perform cross-validation of the satellite dataset and field measurement. The manuscript has the potential to be published, but I strongly suggest a major review to make clearer the methodological framework of the research.

 

Specific comments:

Change keywords. Do not use the same words as in the title.

Section 2.1 (Study area). Needs to include information about Geology and Precipitation.

The tables and figures present must be cited in the body of the text.

Separate values from measurement units, for example, 0.5 mg (line 154 and others).

Line 170: The sentence “the reflectance is an inherent property of the waterbody, independent of the light source” is wrong. The reflectance is an apparent optical property, dependent on light geometry, for example.

Line 181: cite the reference used for the lookup table used for determining r_sky

Line 183: (2.2.3 Spectral characteristics of turbid water) change to the results section

Figures 3: Indicate which curve has low, medium, and high SSC; where it was collected, and the date.

Figures 4: Indicate the value of Chlorophyll concentration; where it was collected, and the date.

Line 210: Explain how the band correlation analysis was made.

What does “N” mean in Figure 5?

Line 249: Figure 6. Correct the number of figures.

Line 252: Cite the studies.

Line 286: (3.1. Initial results): I suggest changing for “Spectral characteristics of turbid water”

Table 2: What is the reason for this table? Explain in the manuscript.

Line 307: What is the wavelength for the Rrs is used in the model shown in equation 4 and Figure 8?

Line 316: Equations 5 and 6 need to be moved to section “material and methods”.

Line 324: “The trend is consistent with previously reported results [15]”. Move to section “discussion”.

Line 329: Explain in the section M&M how the authors obtained this figure (average of the entire historical series, as shown in the Figure below?). What do the letters inside the figure mean?

Line 345: Correct “Figure 111”

Line 367: Explain in the section M&M how to use the Mann-Kendall test.

Line 376 and 377: Move to section “discussion”.

Section 4.1 (line 436): The authors can analyze any other extreme climate events?

Line 481-484: I suggest removing it.

Section 4.3 (line 487-499): It is not discussed about it. What is the dredging frequency? When was a bridge built?

Section 5 (line 518):  the decrease in the SSC shown in the manuscript need to be further discussed.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors have answered most of my comments and the paper can be accepted after a minor revision if they can improve the following two aspects.

(1) the related literature can be further improved with the similar topic, for example the recent publications: Du, Y.; Lin, H.; He, S.;Wang, D.; Wang, Y.P.; Zhang, J.Tide‐Induced Variability and Mechanisms of Surface Suspended Sediment in the Zhoushan Archipelago along the Southeastern Coast of China Based on GOCI Data. Remote Sens. 2021, 13, 929. https://doi.org/10.3390/rs13050929

Zhou, Y.; Xuan, J.; Huang, D. Tidal variation of total suspended solids over the Yangtze Bank based on the Geostationary Ocean Color Imager. Sci. China Earth Sci. 2020, 63, 1381–1389.

 

(2) The language still have some mistakes and should be improved.

Author Response

Please see the attachment.

Author Response File: Author Response.docx