Review of “Expanding the coverage of MISR aerosol retrievals over shallow, turbid, and eutrophic waters”

General comments:

This study designed an algorithm called “Shallow Water (SW)” to retrieve MISR AOD over shallow, turbid, and eutrophic waters to supplement the operational Dark Water (DW) algorithm that is only suitable for open ocean. This SW algorithm retrieve water-leaving radiance firstly, and ten retrieve AOD through minimize a cost function. The SW AOD is well validated against AERONET measurements, and it more reasonable than DW AOD over shallow, turbid, and eutrophic waters. This manuscript is well written.

Specific comments:

1. How BOA downward-directed irradiances and azimuthally-averaged upward transmittances from surface to the instrument are obtained in Equation (4) need to be explained. The two terms are affected by AOD, but AOD is unknown when it is to calculate remote-sensing reflectance in Equation (4).
2. In Fig. 1 and Fig. 2, AERONET-OC normalized water-leaving radiance corrected for bidirectional effects (LWN f/Q) at 667 nm. I suggest changing it to be surface reflectance (or remote-sensing reflectance called in manuscript); this may be better to demonstrate the impact of water-leaving reflectance on retrieving AOD quantitively.
3. In Fig. 5d, the transition from AOD retrieved by DW to SW is not very smooth at around 82.5W, 25N; DW AOD over deep water is larger than SW AOD over shallow. At 81.5, 25.7N, there are some hotspots. All these should be discussed.

This paper outlines an update to the MISR operational algorithm that facilitates retrieval over shallow and coastal waters. This is done by introducing an additional upwelling radiance to the radiative model, which is effectively retrieved alongside AOD (as its value is determined by that which minimises a cost function). The method is demonstrated for three cases studies, and validated against sun-photometer observations in appropriate locations, exhibiting superior performance to the baseline algorithm in most circumstances.

The manuscript is suitable for publication and I commend the authors on its quality. I include some minor comments below that may be considered at the author’s discretion. LX refers to line X of the preprint.

• L49: I found this sentence confusing. Perhaps, “However, there remains significant uncertainty associated with the representation of these waters in the global carbon budget.”
• L103: It sounds odd for an algorithm to “contain” something; perhaps “considers”.
• L106: “mixture” should be plural.
• Eq.2: There is a \tau on the LHS of this equation but not on the RHS. I believe \rho_m is the function of AOD. (Same the same is true of later equations.)
• L149: This sentence might be clearer as, “This derivative is then set equal to zero in order to find the value of $R_{rs}(l)$ at the minimum of the cost function”.
• L155: “an” rather than “a”
• L213: I would say “qualitatively selected” rather than “manually determined”. I think you placed the line where it made sense (in your professional opinion), but the latter could encompass evaluating an equation by hand.
• Eq.A5: A backslash was omitted before exp.
• App.B: In order to keep T and U in the same units, I would replace all U with U^2.
• L416,L428,L452,L484: Incorrectly capitalised titles.
• L416: Berenfeld et al. 2005 is article number GB1006.
• L448: Han et al. 2022 is article number 900694, not 664.
• L457: Kahn et al. 2005 is article number D10S04.
• L460: Kahn et al. 2010 is article number D23209.