The Contribution of Multispectral Satellite Image to Shallow Water Bathymetry Mapping on the Coast of Misano Adriatico, Italy

Round 1

Reviewer 1 Report

For shallow water (as your study area in 5-6 metres), triangulating depths using stereo images is also common, the review should include this as well.

There is no result to support your third objective (in Conclusions).

The applications and results should be better organised for readers to understand.

I'd suggest you to try robust estimate techniques when you apply band ratio methods.

Your title starts "The Contribution of Satellite Images". It is expected that many different types of satellite images from different time and locations are going to be studied however, it is disappointed that only single satellite image was analysed through your paper. 

Since the vertical referencing data in the study area is available for last two decades, and some historical satellite images like Landsat are freely available, it's logic to conduct multiple epochs stud to see the changes for monitoring purpose using your approach.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

This is a well written paper that is based on substantial work by the authors. The paper evaluates the accuracy achievable with satellite bathymetry as a function of the different band ratios. The study concentrates on the difficult to map turbid near-shore waters where satellite bathymetry has the greatest potential to be useful. The results presented will therefore be of current interest to the community of scientists concerned with mapping near shore waters from existing multi-spectral satellite data. For the reasons above I recommend publication of this paper

There are some points which I have noted below which would in my opinion need to be adressed improve the overall usefulness of the paper to the community.

1) In equation 1 TVU needs to be defined precisely and its meaning and use in Hydrography described as most readers will not be familiar with it.

2) line 256 “the absorbed and energy” should be modified to read “the absorbed and backscattered energy”

3) Equation is completely wrong as written. The spectral reflectance involves multiplying the radiance by the solid angle of collection of the sensor which only concerns the distance of the satellite to the surface and in no way involves the distance to the sun as stated on line 264. The equation also does not include either the absorption by the atmosphere or the backscattering contribution from the atmosphere. This section needs to be revised completely.

4) Generally one proceeds from the TOA sensor signal and uses and atmospheric correction to produce the water leaving radiance rather than the opposite which is done here. I have trouble understanding why. This should be cleared up to avoid unnecessary confusion.

5) In equation 5 the factor n is quickly glossed over. One could say:”For the factor n we can use any positive number large enough that for all the pixels in the image both the logarithms used in the ratio are positive.”

6) The notation of the relative sdb (log ratio) is peculiar and awkward. One could use SDB as a subscript perhaps.

7) In figure 8 and 9 the points at a water depth of 3 meters seem like outliers. As mentioned later in the text they could be from a particular localized turbidity increase. This could be noted in the caption as an explanation to justify keeping them in the fit. The same comment applies to fig 13.

8) Line 530, do you mean water reflectance or water leaving radiance (the variable used in the logarithms)   

I have also found several minor error of turn of phrase or typos which are given below.

Line 356 “is not being reliable” should be replaced by: “is not reliable”.

Line 358 “transparence” should be replaced by: ”transparency”

Lines 409 and 530 , “low wavelength” and” high wavelength” should be replaced by ”short wavelength” and “long wavelength” to follow standard nomenclature.

Line 442 “can penetration” should be replaced by: ”can penetrate”

Line 446 “such environment” should be replaced by: “such an environment”

Line 480 “ for log-band ratio” should be replaced by: “ for the log-band ratio”

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

The article is well writen, the methodology is clear, and the results are interesting. This been said, I have some few comments that could help you to refine the paper : I think the discussion's a little poor compared to the other sections and given the significant number of results. And why not, give more details and recommendation on the use of Sentinel and Landsat (526-531). As you know, many researchers cannot afford the use of Quick Bird images, and will be interested to Sentinel & Landsat.

I have also some recommendations :

L 226. : Can you describe the projection method ? Which method, and the output at each resolution ?

L 253-254 : expressed in watts per unit per …. : Too heavy to read, you can remove it and the reader can understand it by reading the formuilation (W.m-2 sr-1 um-1).

Figure 11 is not saved with high quality (sufficient number of dpi). I suggest to resave it again with more « dpi ».

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Line 33:  R²=0.8927-0.9108 and RMSE= 0.35 m. PDF shows different font sizes and the space before "m" should be removed, similarly, some In of In(green) is italic and some are not (in equations), please check thoroughly.

How is the R2 calculated in Figure 10? What's difference between R2 in Figure 10 and Table 3?

In Table 3, results of "Blue and Red Bands" belongs to which method?

The caption of Figure 13 should state which method (log ratio band or OBRA).

Equation 12 is not RMSE, both Z and SDB should take out their means before insert to the equations.

Based on your OBRA's criteria: (Lines 328-330) More specifically, two spectral bands with a high coefficient of determination (R2) with the relative SDB estimated by equation (6) and in-situ bathymetry data among all possible band ratios. and the OBRA results (Figure 10), the optimal choice should be ln(Blue)/ln(Green) (you said so in Line 476), not blue and red bands (Line 497 and Line 563 ), how can this be explained?

Line 476: (green) should Ln(green)

Figure 4 gives relative SDB from band ratio method, where is OBRA's?

line 492: two datasets

Paragraph (Lines 467-472): Figure 10 doesn't have have depth information.

Paragraph (Lines 473-477): Figure 11 should be Figure 12. Figure 11 is Log-band ratio result, NOT OBRA's.

Figures 11, 12 should have the exact same size.

Instead of block dots, it's better to use white dots in Figure 5 since the water color is black.

Author Response

Please see the attachment

Author Response File: Author Response.pdf