Synchronous Atmospheric Correction of High Spatial Resolution Images from Gao Fen Duo Mo Satellite

Round 1

Reviewer 1 Report (Previous Reviewer 2)

see attached file

Comments for author File: Comments.pdf

Author Response

Dear Professor:

Thank you for your comments concerning our manuscript entitled "Synchronous Atmospheric Correction of High-spatial Resolution Images from Gao Fen Duo Mo Satellite" (ID: remotesensing-1854116). We appreciate that you gave us a chance of minor revision. Those comments are all valuable and very helpful for revising and improving our paper. We are sorry for mistakes in English grammar, and We have checked the manuscript, and corrected the errors accordingly. Now I response the comments with a point by point. We sincerely hope that you find our responses. Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report (New Reviewer)

Overview and general recommendation:

The paper “Synchronous Atmospheric Correction of High-spatial Resolution Images from Gao Fen Duo Mo Satellite” by Lingling Xu et al., describes the performance of the Synchronization Monitoring Atmospheric Corrector (SMAC) onboard the high resolution multimode imaging satellite Gao Fen Duo Mo (GFDM) applied for compensating the influence of the atmosphere and reduce the adjacency effects in satellite images. The proposed Syn-AC method allows synchronous application of atmospheric correction to the GFDM images. The corrected by a Synchronization Atmospheric Correction (Syn-AC) method surface reflectance has been compared with the in situ measured values, and with a classical correction method, the Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes (FLAASH), which can work with most hyperspectral and multispectral sensors. Corrected by the Syn-AC method multispectral reflectance was in better agreement with the in situ spectral reflectance than the images corrected by the FLAASH method. Thus, synchronous atmospheric correction can significantly improve the quality (accuracy) of remote sensing (RS) images.

For evaluation/comparison of the performance of the Syn-AC method, instead of absolute errors I suggest to present differences between methods using relative errors or averaged relative errors for different testing sites.

Symbols used in the article are inconsistent.

1)      In equation (1) to (3), indices denoting the wavelength band λ are superscripts, in following equations subscripts. I believe that using subscripts is preferable.

2)      For the radiance usually symbol L is used, as in (1). In equations (15) to (17) Rad has been used instead, and RSRR in (20).

3)      Long symbols like RSRR used for the relative spectral radiance response in equation (20) should be avoided. Instead of long symbols for radiance, L with suitable subscript would be better choice.

4)      For example, instead of absolute error AE, I suggest E_A, and instead of relative error RE, E_R.

The use of units and presentation of values should also be reviewed. The rules of “SI Brochure -The International System of Units (SI)”, pp 148-151, Rules and style conventions for expressing values of quantities should be followed. In particular, unit names and symbols are normally printed in upright type; between numbers and units, a single space shall be left. When dividing the values of quantities using a solidus, brackets are used to avoid ambiguity: (a/b)/c, not a/b/c.

I recommend that after revision by authors the article may be published in Remote Sensing.

Minor comments:

1. Row 27: For performance comparison, relative error is more suitable. Instead of AE: absolute error.
2. Rows 27to 29: Instead of long rows of numbers, I suggest comparison of methods to be based on the mean relative errors.
3. Row 56: Instead of “the resolution” “ground sample distance” could be considered.
4. Row 98: Instead of “analysis of the correction precision” “evaluating the performance of correction methods”.
5. Row 109: 0.42 m
6. Row 127: five polarized bands are centered at (490, 670, 870, 1610 and 2250) nm, and marked with superscript P in Table 2.
7. Figure 3: Should the flight directions of (a) and (b) be the same?
8. Row 127: The radiance unit is equal to µW cm^-2 nm^-1 sr^-1 or µW/(cm² nm sr).
9. Row 127: The degree of linear polarization P_λ …
10. Row 205: least one of the equations from (4) to (7) are satisfied.
11. Equation (8): Arrows as superscripts before arguments: symbols  and , also valid for rows 214 and 215, equation (12), (13), row 250, etc.
12. Equation (9): If   is equal to , then 
13. Row 268: and the radiance unit is equal to µW cm^-2 nm^-1 sr^-1 or µW/(cm² nm sr), ...
14. Row 327: using Eq. (20)
15. Equation (20): Multiplication sign * looks confusing.  
16. Table 5: The number of significant figures in the presented standard deviations is redundant (two figures are normally sufficient).
17. Row 366: measurements was calculated for each band following Eq. (21).
18. Equation (20): Instead of AE, I suggest E_A, and instead of RE, E_R.
19. Calculation formula for relative error?
20. Table 7: Redundant number of significant figures for relative error values.
21. Table 8: According to International Vocabulary of Metrology (VIM), Measurement precision is usually expressed numerically by measures of imprecision, such as standard deviation, variance, or coefficient of variation under the specified conditions of measurement. Instead of absolute error, I suggest to use relative error. The Table’s capture could be: Performance comparison between Syn-AC and FLAASH, basing on Example 1.
22. Table 9: Performance comparison between Syn-AC and FLAASH, basing on Example 2.
23. Table 8 and 9: Redundant number of significant figures for absolute error.
24. Figure 10, 12, 13, 15, 16, 18, 19, 21, 22, 25, 26 and 27: Instead of absolute errors, I suggest to present data using relative errors.
25. Table 10, 11, 12 and 13: Redundant number of significant figures for absolute error. Instead of absolute errors, I suggest to use relative errors.
26. Rows 576-577: Values in text have redundant number of significant figures. I suggest using a Table or a way to present data more compactly.
27. Rows 586-593: To compare or evaluate an absolute error, the compared values are also needed as reference. Therefore, for such analysis using relative error is preferable.

 

Author Response

Dear Professor:

Thank you for your comments concerning our manuscript entitled "Synchronous Atmospheric Correction of High-spatial Resolution Images from Gao Fen Duo Mo Satellite" (ID: remotesensing-1854116). We appreciate that you gave us a chance of minor revision. Those comments are all valuable and very helpful for revising and improving our paper. In the revised manuscript, the relative error between the reflectance retrievals and the field-measured values were calculated and listed in the tables for performance comparison. Considering that the same value of absolute error have different significance for bright and dark targets, the absolute error values are retained.

Now I response the comments with a point by point. We sincerely hope that you find our responses. Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report (New Reviewer)

Test images from the high resolution and multimode imaging satellite, Gao Fen Duo Mo (GFDM), are atmospherically corrected using the data from the Synchronization Monitoring Atmospheric Corrector (SMAC), a multispectral and polarization detection device that can retrieve atmospheric parameters in the same field of view. A synchronization atmospheric correction (Syn-AC) method is proposed to remove the influence of the atmosphere and the adjacency effects to retrieve the surface reflectance. The surface reflectance data observed for three test sites are compared with the field-measured values. In addition, the results are compared with those from the FLAASH algorithm. The visual effects after the Syn-AC are reasonable, and the multispectral reflectance agrees with the field-measured spectral reflectance. Although this manuscript describes some new insights on the atmospheric correction of high-resolution satellite imagery, the authors should consider the following issues to improve the quality of the paper.

 

(major)

L27 “The AEs” – please spell out AE as “absolute error”. Also, please pay attention to the significant digits of errors (See the comments about Table 8, for example). Also, it is questionable if showing the AE values best describes the relative uncertainties of the acquired results. For example, the same value of AE may have different significance for bright (high reflectance) and dark (low reflectance) targets.

L75, L81, L117 (Table 1), etc. “30m” – “30 m”, “0.42m and 1.6m” – “0.42 m and 1.6 m”. Please note that a space is indispensable between a number and its unit since these two (a number and its unit) are considered different words in English. Many more similar cases are found in this manuscript, so please check the manuscript carefully.

L102 “A diagram of the GFDM satellite is illustrated in Figure 1(a) [13]” – Since the figure with two panels seems to be exactly the same as the one shown in ref. [13], it is recommended to omit showing this figure in this paper. Besides, the same figure is available from a website:

http://aircas.ac.cn/dtxw/kydt/202109/t20210924_6214541.html

L126-127 “a spatial resolution of approximately 6.7 km” – Please explain briefly why the resolution of 6.7 km has been chosen, irrespective of the fine resolution of the GFDM main sensor.

L142 Please explain the meaning of the green triangle in Fig. 3(a). It is better to rotate Fig. 3(b) by 90 deg so that the flight direction matches with Fig. 3(a). The caption of Fig. 3 should contain more information about the symbols used in panels (a) and (b).

L163 “… is calculated using Eq. (1), as follows:” – “… is calculated as”. Here please note that in English, each equation should be part of a sentence. Thus, “using Eq. (1), as follows:” should be replaced with “as”. The same applies to all the equations hereafter.

L191 “The main conclusions and corresponding thresholds are as follows.” This part should be replaced with the following sentence (L204-205) “A pixel is recognized as a cloud-covered pixel when at least one of the following conditions on \rho^TOA_490, \roh^TOA_1380, Normalized Difference Dust Index (NDDI), or Normalized Difference Snow Index (NDSI) are satisfied.” Otherwise, the reader may wonder about the relation among different conditions.

L222 Please give an explicit explanation about how the simulation of L_p is implemented. Probably the model (UNL-VRTM) is employed for this purpose, but the current description is not clear enough.

L226 Why only the “volume fine-mode fraction” is parameterized here? Please add a brief explanation.

L227 An important aspect of atmospheric correction is the appropriate choice of aerosol model. Later, in Table 4, it is seen that all the aerosol models are assumed to be continental, but a more general remark should be mentioned when the simulation method is described.

L256 Please revise Fig. 5 so that all the rays are received by the satellite sensor within a very limited FOV (parallel rays). This is because the figure should indicate the radiance components received by a single pixel. In the case of the adjacency effect, an additional atmospheric scattering should be involved after leaving the adjacent pixel.

L316 Please indicate scales (10 m, 100 m, etc.) on each of the panels in Fig. 7. Please give a brief explanation of the small squares seen in Fig. 7(a).

L325-L339 Since later analyses (such as Table 8 and 9) show results for each band, it is not much meaningful to consider the average reflectance defined in Eq. (20). Instead, it would be better to indicate the surface reflectance as a function of wavelength. The same approach would be useful for illustrating the results of Tables such as Tables 8 and 9.

L350 Please explain the exact meaning of “imaging time” and “observation time” in the header of Table 6. Do they refer to the main sensor image of GFDM and the SMAC data, or else? Also, the relation between Tables 6 and 7 is not clear enough. If both AOD and CWV were measured in the ground observation (as suggested in Table 7), please add an explanation about the instruments in relation to Table 6.

L393 Table 7 header : “AOD(550nm)” – “AOD (550 nm)” (two spaces inserted), “CWV” – “CWV (g cm^-2)”

Please pay attention to the significant digits in the relative error. For example, (0.2-0.164)/0.164=0.036/0.164=0.22; thus RE = 22%, not 21.95%.  

L427-428 The general rule of a journal paper is “do not show the same results using both a table and a figure”. In the present case, Table 8 should be deleted. Figure 10 should be retained, and the overall precision can be briefly summarized in the text. The same applies to all the test sites.

L542 “the visual effects of corrected image was improved with the blurring effect removed” –“the visual effects of the corrected image were improved with the blurring effect removed”. It would be desirable to give a more quantitative evaluation of the removal of the adjacency effect and improvement in the image quality.

L576 “The relative errors of AOD were 21.95%, ...” Please pay attention to the significant digits. Moreover, it is questionable to show just the absolute error here since the value of AOD may change, say, between 0.1 and 1.5. The same remarks apply to all similar statements hereafter.

L602 “polarization in-formation” – “polarization information”: The advantage of including polarization information (polarized radiance) has not been sufficiently discussed in the present text.

 (minor)

L23 “field – measured” – “field-measured”

L26 “the visual effects … are enhanced” – please brush up this part so that the reader can understand what is meant precisely.

L38-39 “The gaseous absorption and diffuse reflection by aerosols and molecules” – “The scattering and absorption of molecules and aerosols”: please note that both molecules and aerosols exhibit scattering and absorption.

L42-43 “the atmosphere and adjacency effect” – “the atmospheric and adjacency effects”

L47-48 “acquisition of synchronous atmospheric parameters” – “synchronous acquisition of atmospheric parameters”

L54-56 “the spectral information of the sensor is used for aerosol retrieval, such as the Moderate-resolution Imaging Spectroradiometer (MODIS) [8]” – “the spectral information of the sensor such as the Moderate-resolution Imaging Spectroradiometer (MODIS) is used for aerosol retrieval [8].”

L59 “and generally” – “though generally”

L94 “three radiometric calibration” – “three radiometric calibration sites”: Also, it would be better to cite a reference about these sites, if available.

L99 A reference should be cited for “FLAASH”, and it would be better to spell out this abbreviation since this is the first part this appears in the text (a general rule is that spelling out an abbreviation is necessary for both Abstract and text.)

L117 “Width” – “Swath width”

L125-126 “490 nm(visible)” – “490 nm (visible)”

L128 “different atmospheric missions were accomplished” – probably more clear wording is desirable.

L130 “polarized angle” – “polarizer orientation”

L137 (golden-filled boxes” – please clarify which boxes are meant in Fig. 3(b).

L139 “The central longitude and latitude of the two pixels are represented by the green dots.” – It is not clear from Fig. 3 what are indicated as “green dots.” The same for L141.

L152-153 “following the illustrated steps in …” – “following the steps illustrated in …”

L157 “anomalous attributes, for example the working mode” – “anomalous attributes, for example, the working mode”: Please add some more explanation on the “working mode.”

L165, L268 “The radiance unit is equal to \muW / cm^2 / nm/ sr” – In SI, all the units should be roman (non-italic). Please use \muW cm^-2 nm^-1 sr^-1 instead of repeated “/” symbols.

L167-168 “I ,Q ,and U are expressed using Eq. (2), as follows [17]:” – “I ,Q ,and U are expressed [17] as”. The same applies to all the equations hereafter.

L168 Please check if the red minus signs are correct in Eq. (2). They should appear black, not red.

L169 “polarized azimuth” – “polarization orientation angle”, or “polarization angle.” Please explain that “channel i” denotes the channels with different polarization angles, as mentioned in Table 2.

L171 The comma after eq.(3) should be a period.

L182 Table 3: “Sen-land flag” – “Sea-land flag”, “pixel(A and B)” – “pixel (A and B)”. Please explain the meaning of “A and B”. “not polarized bands” – “non-polarized bands.”

L192-205 Please use present-tense verbs (is, are) instead of past-tense ones (was, were). The same applies to the following part that explains the theoretical treatments.

L208 In eq.(8), “T” should be italic.

L210 “angle ,” – “angle,” (remove a space)

L213 “P. Litvinov et al.” – “Litvinov et al.”

L214 “transmittance items” – “transmittance factors”

L216 “aerosol optical depth” – “AOD” (already defined at L45.)

L228 “AOD(550nm)” – “AOD (550 nm)” (insert two spaces.)

L238-241 Please use non-italic fonts for chemical species (H_2O, O_3, CO_2, and OG).

L247, L250 It would be more reasonable to define the variable T as the right-hand side of eq.(13).

L250 “1/ (cos(\theta_s )) 1/ (cos(\theta_v ))” – “(1/cos \theta_s)+(1/cos\theta_v)”

L251 Please indicate the units of parameters A, B, and C (g cm^-2).

L260, L272 “path-radiance” – “path radiance”

L262-263 “the influence of atmosphere and adjacency effect” – “the influence of atmospheric and adjacency effects”

L273 “T(\theta_v) depicts” – “T(\theta_v) represents” or “T(\theta_v) stands for”

L274 “Eg(0)” – “E_g(0)” (subscript)

L282 “depicts” – “is”

L284 “which is a measure of r from ...” – “which depends on the distance, r, from ...”

L288-289 The subscript “t” in “rho_t” is italic at L288 but is non-italic at L289. Please be consistent.

L290 “q” should be italic since this is a variable.

L303, L309-310, L318, L330 “field-based measurement experiments” – “field-based test measurements”

L305 “As shown in Table 4, the ... images are listed.” – “Table 4 lists the ... images.” The same applies at L330-331.

L314-315 “In the Baotou calibration site, three targets with different reflectivity [38] shown in Figure 7(c), were ...” – “In the Baotou calibration site shown in Figure 7(c), three targets with different reflectivity [38] were ...”

L318 “a spectra radiometer” – “a spectroradiometer”

L328 Please eliminate the indent before “where”.

L329-330 Please insert a space between each number and its unit “nm”. The unit “nm” should be non-italic.

L340 “AOD(550nm)” – “AOD (550 nm)”

L344 “time matching” – “temporal matching”

L346 “Example 1” should appear after introducing Table 6, not before.

L353-534 “The absorption and scattering ... the blurring of satellite images” – It is understandable that the atmospheric scattering is responsible for the image blurring, but is it the case for absorption? If so, why?

L366-367 “Eq.(22)” should read “Eq.(21).” Since symbols || are employed in Eq. (22), “ABS” is not necessary on the right-hand side. The comma after the equation should be a period. The sentence “In the equation” should be placed after Eq.(21), not before.

L369-370 Please show a reference for FLASSH here.

L373-374 The abbreviations “IDL”, “VNIR”, and “SWIR” should be explained.

L377 Please explain what parameters are considered here as “atmospheric parameters.”

L390 “The results indicates that ...” – “The results indicate that ...”

L398-399, L4-6-407 “The AC parameters are retrieved from SMAC with AOD of ...” – “The AC parameters retrieved from SMAC data are AOD of ...”

L400 “g/cm2” – “g cm^-2”: please use a superscript, not a slash. The same applies to all similar cases hereafter. “Figure 8(a-c) present” – “Figures 8(a)-(c) present”

L401 “B3, B2, and B1” – “band 3, 2, and 1 for red, green, and blue, respectively”

L404-405 The order of images should be in line with Table 7 (Dunhuang, Songshan, and Baotou.) (Why is it “AOE-Baotou” in Table 7?)

L446-449 Figures 12 and 13 can be merged together to show both the white and black targets in the Songshan site (August 26, 2020, AOD=0.225). Table 9 can be omitted.

L465-468 The same applies to Figures 15 and 16 (Songshan on February 4, 2021, AOD=0.421). Table 10 can be omitted.

L482-485 The same applies to Figures 18 and 19 (Songshan on February 8, 2021, AOD = 1.19). Table 11 can be omitted. It seems that the choice of aerosol parameters is insufficient in the FLAASH analysis, particularly for this high AOD case.

L503-509 The same applies to Figs. 21-23 for the white, gray, and black targets in Baotou (august 26, 2020, AOD = 0.338).

L521-527 The same applies to Figs. 25-27 for the white, gray, and black targets in Baotou (august 31, 2020, AOD = 0.092).

L538 “The paper demonstrate that ...” – “The present paper has demonstrated that ...”

L549 “atmospheric model and aerosol model” – please explain how these models are different in this context.

L554-555 “the band settings of main sensor may not meet the requirements of FLAASH to retrieve atmospheric parameters” – It is not clear what is meant by “requirements of FLAASH.”

L561 “experiment results” – “experimental results”, “more image data is required for analysis of ...” – “more image data are required for the analysis of ...”

L574 “In the study” – “In the present study”

L594 “Preliminary analyses ... indicates” – “Preliminary analyses ... indicate”

L598 “for the RS application” – “for remote sensing applications”

 

Author Response

Dear Professor:

Thank you for your comments concerning our manuscript entitled "Synchronous Atmospheric Correction of High-spatial Resolution Images from Gao Fen Duo Mo Satellite" (ID: remotesensing-1854116).

We appreciate that you gave us a chance of minor revision. Those comments are all valuable and very helpful for revising and improving our paper.Thank you for your careful reading, and the errors were modified as the suggestions.

Many changes have been made to the way to present the results. The figures are merged together to show both the bright and dark targets. Considering that the same value of absolute error have different significance for bright and dark targets, the relative error between the reflectance retrievals and the field-measured values were calculated in the revised manuscript. Then, the absolute errors and relative errors were listed in the tables. The Figures of absolute error were removed. As the suggested, we have merged the figures.

Now I response the comments with a point by point. We sincerely hope that you find our responses. Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report (New Reviewer)

The manuscript has been improved substantially. For further improvements, however, the authors should reconsider the following issues. No more review is necessary.

L30-31 "19.34%" and "76.58%" - please reconsider the significant digits given to these error values (see examples below).

L83, L110-111 "resolution are" - "resolutions are"

L98-99 "the average reflectance of the selected targets in the surface reflectance image of Syn-AC were compared with" - "the average reflectance of the selected targets in the surface reflectance image of Syn-AC was compared with"

L100 "the evaluating of" - "the evaluation of"

L124 Please add a reference for the SMAC sensor.

L132 The acronym "DOLP" (degree of linear polarization?) should be mentioned as a footnote to the table.

L139 "Therefore, the golden-filled boxes cover ..." - "Therefore, in Figure 2(b), the golden-filled boxes cover ..."

L145 "(a) Synchronous detection mode": in the figure caption, please add a brief explanation of the green and red lines in Fig. 2(a).

L168 "is calculated as:" - "is calculated as" - the same omission of a colon applies to all the similar cases hereafter. Please note that in an ordinary sentence, a colon is not needed just after "as".

L171 The importance of including polarization information must be explicitly stated around here (Response 20 in the reply sheet). Generally, it is strongly recommended to include the major points from the reviewer comments in the revised manuscript since such points are important also for readers. Please re-check the response sheet from this viewpoint.

L173 In Eq. (2), "cos2(\alfa_{\lambda,1})" should be "cos(2\alpha_{\lambda,1})". The same applies to all nine cases.

L188 Please add a footnote (or some appropriate remark) about what is meant by “pixels A and B”.

L218 "...depicts the surface polarized reflectance" - "... is the surface-polarized reflectance"

L224 In Eq. (11), the parenthesis in the argument of the "cos" function can be omitted, just as in the case of the denominator in Eq. (9). The same applies to L256.

L231 Please add an explanation of the parameter "d" in Eq. (12).

L247 "O3,CO2" - "O3 and CO2".

L253 In Eq. (14), the equal symbol in "=T" should be replaced with \equiv (three lines, not two lines) that symbolizes the definition.

L256 Please remove the space before each comma.

L265 "effect).The" - "effect). The"

L281 A reference should be added for the "6SV model".

L309 Please add the explanation on the "image time" - is this time and date of image acquisition? Also, please add a brief explanation of why the aerosol model has been fixed to "Continental" in these three cases.

L310 "Time(UTC)" - "Time (UTC)"

L318 The map scales are too small and too complicated. It would be better to show a simple bar indicating the scale of 100 m for each panel in Fig. 6.

L319 In the caption of Fig. 6(a), it would be better to mention the numerous small black structures that appear on the right side of the panel (though they are not used in the present study).

L322 "The spectra radiometer" - "The spectroradiometer"

L339 "changes gently" - "changes insignificantly"

L341 In Table 5, it is strongly recommended to list the average value of field-based reflectance calculated using Eq. (21). Otherwise, it is not much meaningful to list just the standard deviation.

L352 Please insert a space just before each "(UTC)", as in the case of L310.

L358-359 "the visual effects of the satellite images before and after atmospheric correction were compared in the study": the calculation of modulation transfer function (MTF) or similar indices may contribute to evaluating the effect of AC more quantitatively. This is just a remark for future study.

L368 The period just after Eq. (23) should be a comma since the sentence still continues.

L396 In Table 7, the data "0.2" (occurring twice) should be "0.200" in consideration of the significant digits.

L409 "Image (a) ~ (c) are original images, and image (d) ~ (f) are ..." - "Images (a) - (c) are original images, and images (d) - (f) are ..." Please use “ - “ instead of “ ~ “ throughout the text.

L444 "slightly polluted" - "the slightly polluted atmosphere". The same at L460 and L478.

L457 It is recommended to retain up to only two or three significant digits for showing the error values. This is because of the common understanding that "the error values cannot be so precise" (e.g., not 22.69 but 22.7 or 23). Showing five digits for E_R is considered inappropriate (e.g., not 124.94 but 125 or even 130). Please revise the digits in the text and other tables.

L462 "are displayed Figure 13" - "are displayed in Figure 13"

L566-567 "under different atmospheric states and underlying surface and the influence of atmospheric model and aerosol model on correction accuracy" - too many "and" appears in this part. Please consider rephrasing this part.

L588 "and it representing" - "representing"

L590 "19.34% and 76.58%" - "19.3% and 76.7%" (or "19% and 77%" would be more reasonable because of the inherent uncertainty in these error estimations).

Author Response

Dear Professor:

Thank you for your comments concerning our manuscript entitled "Synchronous Atmospheric Correction of High-spatial Resolution Images from Gao Fen Duo Mo Satellite" (ID: remotesensing-1854116).

We appreciate that you gave us chances of minor revision. As your suggestions, we have revised the manuscript (significant digits of error values, grammar, reference, etc.), and added more explanations to the important points. Thanks for your suggestion for future study.

Thank you again for your careful reading, and those comments are very helpful for improving our manuscript. Now I response the comments with a point by point. We sincerely hope that you find our responses. Please see the attachment.

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

The paper describes the atmospheric correction approach for high-resolution GFDM sensor. While the manuscript shows a complex problem in remote sensing, the atmospheric correction approach is not novel, empirical thresholds and decisions were not well justified, and it does not provide a new operational product as well. The manuscript can be improved in many ways, especially the structure (e.g., separating of experimental & method) and English (fluency/grammar) lacks rigor. The manuscript requires intense editing to improve readability and lacks details and analysis in methodology. Several satellite systems contain useful bands for atmospheric parameter retrieval, and the authors lead the reader to believe GFDM has unique characteristics for accurate surface reflectance retrieval, but the atmospheric parameter retrievals will present uncertainties as any other available products. The paper does not have a discussion section, which is highly appropriate to connect the previous results with the current advance. The authors provide empirical thresholds for cloud, cirrus clouds, desert areas. However, no further analysis was performed to define these thresholds and they are likely to be local and scene-based dependent; this is a limitation for operational implementation. The mixing of equations (8), (11), (15) to express Rtoa and Ltoa is quite confusing, and the presentation of these equations should be improved. The current description of aerosol retrieval is not sufficient to understand the applied approach, and this step is critical for any atmospheric correction. The image-based aerosol retrieval is a challenging topic and has a long-term effort from researchers. This aerosol section lacks a lot of details about the surface assumptions, AOD inversion, and simulation process, and readers cannot reproduce the methodology as it is. Experiments for aerosol retrieval should be performed. The authors included methodology in the experimental section such as equation 18 and table 7, description of QUAC and FLAASH, which is not appropriate or at least difficult to follow. For these reasons, I recommend the rejection of this manuscript.

Comment 1: Citations with numbers should be placed within [X].

Comment 2: The paragraphs are short and can be connected. For instance, paragraphs (1 and 2) or paragraphs (3 and 4) have similar content and flow for connection. Please revise the introduction and make this connection as needed.

Comment 3: Line 63: What do you mean by “instrument for atmospheric detection”? This is very confusing.

Comment 4: Line 64 - I don’t think “realize” is the correct verb for this sense. Please verify it.

Comment 5: Line 88-89You should add 2-3 sentences about the experiment design (period, number of images, location).

Comment 6: Unless the authors provide a better figure (quality and description), figure 1 has no value for the paper and can be removed. For instance, Fig1c has no description what are the components and what is represented.

Comment 7: Line 109: Could you confirm that GFDM provides global revisit < 2 days?

Comment 8: Line 113-114: the sentence “high definition of the image texture” is not common way to express the sensor characteristics (texture is target property). Please improve it.

Comment 9: Please check the units of line 168. I believe it’s more common m2 instead cm2, but maybe that’s why you gave the uW.

Comment 10: please provide a reference for stroke equation in line 171

Comment 11: You mention “judgment method”. What is this method? Do you have reference? I believe you are saying “empirical” = “judgment”.

 

Comment 12: The authors provide empirical thresholds for cloud, cirrus clouds, desert areas. However, no further analysis was performed to define these thresholds and they are likely to be local and scene-based dependent. Please provide further information how did you define these parameters.

Comment 13: line 223: Please provide reference for “Vector Radiative Transfer Model”

Comment 14: Line 282: the r measure depends on many factors and is a huge source of uncertainty in this approach using equation 17. This description is superficial and not even present the used values for the equation. Please provide further comments about the uncertainties of this method and used parameters from cited study.

 

Comment 15: Figure 5 and 6 are useful for comparison and should be placed together for quick assessment.

Reviewer 2 Report

See attached file

Comments for author File: Comments.pdf

Reviewer 3 Report

In this paper, synchronous atmospheric correction of High-spatial Resolution Images from Gao Fen Duo Mo Satellite was carried out based on retrieved atmospheric parameters form the Monitoring Atmospheric Corrector (SMAC) observations. The manuscript needs careful editing in grammar, spelling, and sentence structure in order to make the readers understand the manuscript more clearly. Following shows some specific questions after I reviewed this paper.

(1)  A period at the end of the Abstract is missed.

(2)  Does the number “1” in Line 37 means the cited reference? Please check the requirement on references citation of the Remote Sensing journal.

(3)  Lines 140: There seems no “yellow-filled boxes” in Fig.2.

(4)  Equation (1) shows how to covert the DN numbers to the radiance with the absolute spectral response and gain of each band. I am wondering where the absolute spectral response and gain of each band comes from? Does it determined at laboratory before SMAC launched or calculated with onboard calibrator or other ways?  The same requirement should also be explained for equation (2).

(5)  Equation (10) shows the cost function between observed and simulated radiance. How to calculated the observed radiance ( ) at the polarized bands (490, 670, 870, 1610, 2250 nm). Equations for how to calculate  is recommended to make the readers clearer. Besides, the simulated radiance ( ) in equation (10) is calculated from the simulated in the LUT? Please make more explanations.

(6)  For the water vapor retrieval, I am confused whether it retrieved from the High Resolution Camera or SMAC sensor? It is mentioned that the WVC was retrieved from bands centered at 865 nm and 910 nm, however there is no window channel centered at 865 nm if it retrieved from SMAC observations.

(7)  Lines 258, please make sure the reference citations 272829 are correct.

(8)  In section Spectrum Analysis, the surface (SURF) reflectance image after Syn-AC is compared to the reference surface reflectance from ENVI spectral library is unreasonable except for tell the readers that they have similar spectral shape. However, for quantitative analysis, comparison with ground measured surface reflectance is more meaningful. Please add the validation analysis results with ground measured surface reflectance.

(9)    In section Quantitative Analysis, the retrieved AOD from SMAC on 2021/8/26、2021/2/4、2021/2/8 are 0.225, 0.421, and 1.19, which means the weather on 2021/8/26 is clearer than the other two days. However, the corrected surface reflectance on 2021/8/26 has larger difference than that on 2021/2/8 when compared with ground measured values. The authors must give the reasonable and detail explanations on the phenomenon. Similar questions are also should be explained for the quantitative analysis at AOE-Baotou site. 

(10)  I am confused why only the white targets were used to validate the corrected surface reflectance at Songshan and Baotou site since there are black and gray targets as well. I suggest add the validation of the corrected surface reflectance at the black and gray targets. Besides, validation before and after adjacent effect correction should also conducted since the contrast of the targets at the two sites. Moreover, it is recommended to add references related to the sites, so that readers can refer to relevant information.  

(11)  The title of Figure 21 is incorrect, please check it.

Comments for author File: Comments.pdf