Extraction of Spatial and Temporal Patterns of Concentrations of Chlorophyll-a and Total Suspended Matter in Poyang Lake Using GF-1 Satellite Data
Round 1
Reviewer 1 Report
In the introduction, the previous works for water quality mapping are insufficient. Many previous algorithms include analytical and empirical approaches. Please the critical references. Moreover, this study considers the regression-based model. Mapping water quality requires a spatio-temporal nonstationary model.
The method’s not clear for us. In line 127, semi-analytical/semi-empirical approaches were used to estimate the concentrations of Chl-a and TSM. What do you mean the semi-analytical/semi-empirical approach using in this study?
In Chla, the model is used from the APPEL model. Table 4 shows the quadratic polynomial and power-law equations. However, the authors did not mention the model for TSM. Could the authors provide the equations in the methods?
Compared to the similar studies, what is the new finding in this study?
Figure 4a shows reflectance spectra from three sampling sites with approximately the same Chl-a concentration and different TSM concentrations. What is the purpose of getting result of the figure 4? Moreover, another correlation coefficient plot is shown in Fig 6. What is the implication of the figure 6?
Fig 6 shows the correlation coefficients between the spectral index (reflectance ratio and reflectance difference) and Chl-a concentration. Table 3 shows spectral response characteristics of Chl-a. Then, GF-1 bands 1, 3, and 4 were used to establish APPEL model in this study. Are they independent? Or what are the relation between the APPEL model and results from Table 3 and Fig 6?
Most studies achieved satisfactory Chl-a estimation and focused solely on the spectral regions from near-infrared (NIR) to red spectral bands. Table 3 shows the combinations of spectral wavelengths. The highest correlation ones are not near-infrared (NIR) and red spectral bands. This finding is similar to the study from the following.
References:
Van Nguyen, M.; Lin, C.-H.; Chu, H.-J.; Muhamad Jaelani, L.;Aldila Syariz, M. Spectral Feature Selection Optimization for Water
Quality Estimation. Int. J. Environ. Res. Public Health 2020, 17,
272. Salem, S.I.; Higa, H.; Kim, H.; Kazuhiro, K.; Kobayashi, H.; Oki, K.; Oki, T. Multi-Algorithm Indices and Look-Up Table for Chlorophyll-a Retrieval in Highly Turbid Water Bodies Using Multispectral Data. Remote Sens. 2017, 9, 556.
Author Response
We appreciate the insightful and constructive comments and suggestions. We hope that the concerns and suggestions have been addressed to satisfaction in the revised manuscript. The point-by-point responses to the specific comments are presented in the attached Word file.
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This is an interesting paper where authors used data from the Gaofen-1 (GF-1) satellite, in situ measurements of the reflectance of the lake water, and the analysis of the Chl-a and TSM concentrations of lake water samples to investigate the spatial and temporal variation and distribution patterns of the concentrations of Chl-a and TSM.
I think the paper show novelty, especial because validating data from a new remote sensing satellite with in-situ measurements. However, I consider some further explanation is needed in the introduction and the results. Also, I consider major issues the lack of citation in the results/discussion section when the authors describe the chl a and TSM Retrieval models and their results.
Also, I consider the paper will be improved if the bands elected from the GF 1 satellite and the equations used in the retrieval models of this paper are compared to previous work done with other satellites such as Landsat and MODIS for estimating chl a and TSM. An intent of doing this can be observed between lines 260 and 270 for chl a; however, this can be improved.
Some specific comments:
Lines:
48. High temporal resolution is needed too.
55. Add some where it said: In SOME lakes chl a can be use as a proxy of primary production.
Chl a can not be use to estimate PPR in oligothrophic and ultraoligothrophic lakes.
The introduction is lacking a good description of the study lake. Information as regards trophic state, mean Secchi disk depth, chl a, and TSM seasonal patterns from previous works is important in this section.
100 to 110 A better description of how chl a and TSM field data was obtained is needed. It especially is needed to say how much water was filtered. When were water samples taken in each season? All the samples on the same date? How close are the sampling dates to the date of the remote sensing images?
166. What does it mean: after removing abnormal reflectance data? More explanation is needed on this. If not seems, the author has chosen data that fit their hypothesis.
260-273. The equation defined in the Appel model needs a further explanation as well as the bands elected in that equation. Why subtracting one band to others, why multiplying one band by other...
Figures 9 and 13 can be a single figure. The standard error bar is needed for Chl a and TSM concentrations.
326-366. Citations are needed. The authors are comparing their results to the general pattern of the lake, which has been studied and described in previous work. Eg. Line 336 The relative high chl a concentration... is due to.... need citation, as well as most of the sentences in this section.
375-402 It does not thoroughly explain how the equation elected for TSM is related to TSM optical properties and why for two seasons, the authors use power-law equations, and for another season, they use a quadratic polynomial equation. I understand that is because the equations were found to have the best fit. But how is that related to the observed spatial variation of TSM or the optical properties of TSM in the field?
409-466. Citations are needed. See previous comment for the chl a section.
442. Where it says figure 13... Is it correct? Or should it said figure 12?
Author Response
We appreciate the insightful and constructive comments and suggestions. We hope that the concerns and suggestions have been addressed to satisfaction in the revised manuscript. The point-by-point responses to the specific comments are presented in the attached Word file.
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
1.The most important finding is here. The study suggested that the wavelengths corresponding to bands 1, 3, and 4 of the GF-1 images were the most sensitive to changes in the concentration of Chl-a. The results showed that the correlation between the reflectance and TSM concentration was the highest for wavelengths corresponding to band 3 of the GF-1 satellite images. But they are still not clear for us. We suggest that the result should be included the figure or table number. If possible, please explain them clearly.
2. This study identifies the correlation coefficients between the spectral index and water quality. However, spectral index is from measured water reflectance, not GF-1 image. The authors used the GF-1 images to estimated the water quality parameters. How did we make sure that the GF-1 reflectance is the measured water reflectance?
3. In line 163, the preprocessing of GF-1 images includes geometric correction, radiometric calibration, atmospheric correction.In this paper, image data were processed 165 using the ENVI 5.3 software. Could you give more information for the algorithms and validations?
Author Response
We appreciate the insightful and constructive comments and suggestions. We hope that the concerns and suggestions have been addressed to satisfaction in the revised manuscript. The point-by-point responses to the specific comments are presented in the attached Word file.
Please see the attachment.
Author Response File: Author Response.docx
