Remote Retrieval of Suspended Particulate Matter in Inland Waters: Image-Based or Physical Atmospheric Correction Models?

Round 1

Reviewer 1 Report

The research target may rouse interest many readers interest however, I recommend some revisions on the paper for being more reliable and the results able to apply to more general cases.

 

Specific comments:

Line 180-182 “the COST uses an additional correction for transmittance along the path from the ground toward the sensor (known by TAUz). The correction of this module, which is part of the multiplicative effect of the atmosphere, is made by the computation of the cosine of the 179 solar zenith angle. According to Chavez [34], the solar zenith angle cosine is a good approximation of TAUz.”:

The assumption of “the solar zenith angle cosine is a good approximation of TAUz” seems strange if TAUz is defined by the above.

TAUz may be“Atmospheric transmittance along the path from the sun to the ground surface”, and TAUv may be “Atmospheric transmittance along the path from the ground surface to the sensor” in Chavez [34].

 

Line: 211-212 equations:

The lambda seems to indicate not the wavelength but the reflectance values.

So, it should be Reflectance(lambda)?

 

Figure 3 and Line 212:

I could not understand where the "Mirror" is used.

Can you explain more about the usage of the mirror ?

 

Figure4:

Y-axis of ToA seems strange because it is reflectance and usually in 0.0-1.0.

Why the values of COST and ones of DOS are the almost same?

The transmittance correction is properly applied to the COST? Can you check the processing?

 

Figure 5 and Line 303-305 "It is, however, important to emphasize that according to the 1:1 line, except for ToA, all models tend to underestimate high SPM concentrations with different variability levels.":

The samples seem to be consisted by the two groups (<20mg/L and >35mg/L), and the regression results may depend on the shape (order) of the regression curve.

Could you explain about the SPM range difference, and Can you show the regression equation?

 

Figure 7:

Can you enhance the intensity of the RGB image to see the water areas clearly?

The layout (order of COST, DOS, ToA, ACOLITE, ATCOR, and FLAASH) of the figures should be the same for the Figures 4-8, I think.

 

Line 420-428 “Plus, a reciprocal relationship exists between SPM and clear waters (the less waters are laden of SPM, the clearer they are [3]). On the other hand, SPM in waters causes turbidity. Turbid waters are known to be well correlated to the red part of the spectrum [40]. Thus, the overcorrection of the blue band by the physical models likely destroys this natural physical dependence existing between the concentration of SPM and the absorbed energy by waters in the blue part of the spectrum.”:

This part seems to say important things, but I could not understand what you are saying in the sentences (and cannot easily access to [3]). Could you explain more about this part?

I suppose the blue bands can be influenced strongly by the colored dissolved organic matter (CDOM) which sometimes correlates with SPM but sometimes independently distributes.

Could you mention about the influence of the CDOM in the sample waters analyzed in this study?

 

Author Response

To reviewer

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript presents a comparison of the performance of image-based atmospheric correction models with the common physical models most used for retrieving SPM concentration. The topic is interesting, and it should be published. I find the manuscript rather well written and illustrated. I have only two suggestions to improve the readability of sections 2 and 3. The section 3 also presents methods, so it can be a part of section 2 (Materials and Methods). An additional separate section with illustrations and basic information about the study area should be provided.

Author Response

To reviewer 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Line 430-439: "an inverse correlation between the reflectance captured by the blue band";

->

The correlation between the blue band and SPM can be positive when SPM does not have strong absorption in blue. the negative correlation is appeared when the SPM has the absorption in blue or the high SPM is coincident with high CDOM or algal blooms.

So, the negative correlation between blue absorption and SPM in this paper is a kind of "local" relationship about SPM, CDOM, and algal blooms, I think.

 

In the last answer of the cover letter, “We agree that CDOMs are strongly correlated with the blue part of the spectrum, especially at 440 nm, but this is in terms of absorption coefficients, not reflectance. Therefore, CDOMs in water do not affect the signal captured by Landsat's blue band as much.”:

->

The water leaving reflectance (rw) is correlated with the total of the back scattering coefficients (bb) / total of the absorption coefficients (a) (rw ~ bb/a). So, CDOM absorption can affect the Landsat's blue band.

e.g., https://www.ioccg.org/reports/report5.pdf (page 10)

Could you re-consider the explanation about the explanation? 

(Your final conclusion, “The over-correction of the blue band reflectance by physical models probably destroys this natural physical dependence that exists between the SPM concentration and the energy absorbed by waters in the blue part of the spectrum …”, seems reasonable.)

Author Response

To reviewer

Author Response File: Author Response.pdf