A Multi-Band Atmospheric Correction Algorithm for Deriving Water Leaving Reflectances over Turbid Waters from VIIRS Data
Round 1
Reviewer 1 Report
This manuscript evaluates a possibility of using an atmospheric correction algorithm in Ocean Color previously developed by authors to make retrievals of the water leaving reflectance in four water areas with high reflectance due to shallow waters and high turbidity as well as algal blooms, which are not routinely retrieved by NASA and NOAA processing algorithms. The success is demonstrated in a qualitative manner since the authors did not have fully digital data available from NASA and NOAA consistently for TOA and surface level imagery.
The manuscript can be published as a demonstration of the approach with corrections below.
In Eq 2 it is not clear why the transmittance coefficient for the downwelling light is included in the water leaving reflectance term.
Results are presented in different formats for different areas. Where it is possible, it would make sense to have color bars to get better understanding of results.
Author Response
Please see attatchment!
Author Response File: 
Author Response.pdf
Reviewer 2 Report
While novel, the main issue I have with this paper is that it looks like your algorithm can produce results when other atmospheric correction algorithms would fail (mask out pixels), but how do we know your results are accurate? The paper does not include any ground truthing, coincident in situ data analysis.
Additionally, it would be interesting to see how well your algorithm performs compared to the traditional Gordon/Wang algorithm for VIIRS, for datasets that consist of clear scenes. Studies could be conducted at MOBY and AERONET-OC locations where water-leaving radiances are collected. A multi-year study at a couple of locations would be ideal.
The study feels incomplete. I think it needs a layer of ground truthing to verify that your results are useful. Once this is added to the paper, I would be happy to produce a more thorough review document.
Author Response
please see attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
Paper Summary and Recommendation
The manuscript presents a novel atmospheric correction algorithm for deriving water leaving reflectances over turbid waters from VIIRS images. Current algorithms do not focus on those waters and could potentially result in biased low chlorophyll concentrations over some geographic regions. The algorithm uses three SWIR wavelength bands (1.24, 1.61, and 2.25um) of VIIRS to derive aerosol information with extrapolation to obtain aerosol optical properties in visible for deriving water leaving reflectances. The authors demonstrated the algorithm for four VIIRS scenes over turbid waters. They concluded that the spatial coverage of water leaving reflectances with the algorithm can be improved significantly compared to the current NOAA operational algorithm. The method is sound. The manuscript is well written. The research work suits the journal of Remote Sensing’s aims and scope (e.g., multi-spectral and hyperspectral remote sensing). I recommend the manuscript for publication in Remote Sensing with reservation depending on the authors’ response to my comments below.
Recommendations for Revision: General Comments
Although the paper is well written, and the physics of the algorithm is sound, potential readers would ask how well the algorithm performs for open ocean where the conventional algorithms produce results. I do notice that the authors did mention that the retrieved water leaving reflectances over deeper ocean waters compare well with those derived by NASA MODIS operational algorithm (lines 197-200). It would be nice to include a figure to demonstrate this.
Authors state that “the SWIR bands centered near 1.24, 1.61, and 2.25 um are used in a spectrum-matching processes to obtain aerosol information, which is subsequently extrapolated to visible region for the derivation of water leaving reflectances of visible bands” (lines 17-19). The word extrapolation or extrapolated appear several times in the paper. It could mislead readers. I understand the this is not a simple extrapolation process. For my understanding, the reflectances in three SWIR bands are used to derive aerosol information such as aerosol models, which are used to for calculating water leaving reflectance in visible bands. It would be helpful to clarify this.
Some minor comments:
Line 36: Define NIR (near-infrared)
Line 58: What are wavelengths for the RGB image? Several RGB images are presented without mentioning the wavelength combination. It is helpful to specify the wavelengths for all RGB images.
Line 76: OC is introduced without definition.
Line 82: Labels for latitude and longitude in Fig. 1 (C), (D), (E), and (F) are barely visible. Suggest using larger font and perhaps in red color.
Line 134: “apparent reflectance at a satellite level” suggest using “apparent reflectance at top-of-atmosphere (TOA)”
Line 137: “E0 the downward solar irradiance at the top of the atmosphere when the solar zenith angle is equal to zero”. Suggest using “E0 the incident spectral solar irradiance at TOA”.
Line 148: “Eqs. (1) and (3)” would be better just using “Eq. (3)” or “Eqs. (1) or (3)” . Also, replace Eq. (2) to Eq. (3) in line 151.
Line 200: A period is missing after water surfaces.
Line 287: Latitude and longitude labels for Fig. 5(A) and (B) are too small to see. It would be helpful to make it the same size as Fig. 5 (C) and (D).
Author Response
Please see attached file for our responses!
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Thanks for addressing my concerns. You provided good explanations.
Author Response
We made minor modifications to the manuscript. Thanks for help reviewing our manuscript!
