Piotr Kowalczuk

Institute of Oceanology

Polish Academy of Sciences

Ul. Powstańców Warszawy 55

PL – 81 - 712 Sopot

July 28, 2026

Editor

Earth System Science Data

Attn: Review of the manuscript by Jean-François Berthon and Giuseppe Zibordi “Coastal Atmosphere & Sea Time Series (CoASTS) and Bio-Optical mapping of Marine optical Properties (BiOMaP): the hyperspectral absorption coefficients by optically significant constituents” submitted to Earth System Science Data and coded essd-2026-463.

Dear Editor,

After reading the manuscript by Berthon and Zibordi submitted to Earth System Science Data and coded essd-2026-463 I recommend to consider this paper for publication in this journal after minor revision.

General opinion

Authors have presented data sets of absorption coefficients spectra of particles suspended in sea water, phytoplankton pigments, and non-pigmented particles in the spectral range between 400-750 nm measured with a 3 nm spectral resolution at 2 nm increment, in water samples collected in the surface layer of numerous marine water around Europe. The particulate absorption spectra are accompanied with absorption coefficient spectra by Chromophoric Dissolved Organic Matter – CDOM, measured in same water samples volume in the spectral range 350-680 with 1 nm resolution. Presented data set consist with two parts: Coastal Atmosphere & Sea Time Series program performed from 1998 to 2016 in the northern Adriatic Sea at the Acqua Alta Oceanographic Tower (AAOT)  - CoASTS, and the Bio-Optical mapping of Marine optical Properties (BiOMaP) field program performed from 2000 to 2021 in the European Seas. The geographical extent of sampled marine basins included: Adriatic Sea, Ligurian Sea, Eastern and Western Mediterranean Sea, the Blac Sea, the Baltic Sea, Iberian Shelf, English Channel and North Sea, and Greenland Sea. Author have measured particulate and CDOM absorption spectra with use bench top spectrophotometer using well established methodology. Authors have also assessed uncertainty budget for measuring particulate and CDOM absorption spectra. They have assessed measurements uncertainty with two ways – first, by repeating spectrophotometric scans of the same sample and second measuring replicated sample taken from the same water volume. In case particulate absorption measurement of repeating sample the uncertainty was lower (the mean absolute percent differences (MAPD) ranged 2.8% - 7.4%) compared to replicated samples (the MAPD ranged 8.9% - 10.7%) and uncertainty increased toward longer wavelengths. The measurement uncertainty of CDOM absorption was higher for both for repeating and replicated spectrophotometric scans, and have the same increasing tendency toward longer wavelengths.

Author have presented data in the form of visualized spectra phytoplankton pigments absorption coefficient, a_ph(l), depigmented particles absorption coefficient a_dt(l), and CDOM absorption coefficient a_ys(l) for a given geographic locations. Authors have also calculated averaged and standard deviation of absorption spectra for a given constituents and given geographic subsets. Authors have also showed a tertiary plots for three constituents relative percent contribution to sum of a_ph(l), a_dt(l), and a_ys(l). These plots illustrated which constituent controlled a total absorption in a given European sea, or its subregion.

The last figure presented a relationship between a_ph(442) versus a_ph(676) to illustrate the impact of regional phytoplankton pigments composition. Data showed decreasing values of the average a_ph(442) / a_ph(676) ratio from oligotrophic waters of the Mediterranean Sea to more eutrophic Baltic Sea and the North Sea waters.

This is a very important contribution of Author, setting a benchmark for contemporary spectral values of absorption coefficient of three most important absorbing constituents. For these reason this manuscript deserves a prompt publication in Earth System Science Data journal. However I see one flaw of this presentation. Authors shall confront and compare their data with existing publications of same absorption coefficients measured with same methods. There exiting publication for many of considered regional European sea – especially in the Baltic Sea, the North Sea and the Greenland Sea. This way Authors could cross-validate quality of  their contribution.

I found one minor error:

Page 2 Abstract  Typo is Adriatic Sea at the Acqua Alta Oceanographic Tower

Should be Adriatic Sea at the Aqua Alta Oceanographic Tower

Page 8 line 120

The equation: is a_ph(λ) = a_p(λ)  ̶ a_ph(λ),

should  be: is a_ph(λ) = a_p(λ)  ̶ a_dt(l)

Best regards

Piotr Kowalczuk

The manuscript by Berthon and Zibordi describes the data release of the hyperspectral absorption coefficients for dissolved and particulate matter measured during the Coastal Atmosphere & Sea Time Series (CoASTS) and the Bio-Optical mapping of Marine optical Properties (BiOMaP) programs carried out in European waters from 1995 to 2022.

The hyperspectral absorption coefficients complete and enhance the data released by Zibordi and Berthon in 2024 (ZB24 hereafter) for the apparent and inherent optical properties, as well as the concentration of the optically significant water constituents for the same programs. The ZB24 dataset was limited to multispectral bio-optical data products distributed at spectral values restricted to key ocean color wavelengths.

The manuscript summarises several aspects of the ZB24 paper, including the details of the CoASTS and BiOMaP programs, the sampling station locations and the measurements methods and the measurement precision analysis. Then the hyperspectral absorption coefficients are summarised graphically and the spectral variability across marine regions is described, albeit very briefly.

This manuscript is suitable for publication in Earth System Science Data pending a revision addressing the following concerns and the suggested edits.

1) Comparison to other datasets

Related work was not adequately referenced. Throughout the text, a comparison of this new dataset to other available bio-optical data would enhance the manuscript. Further than ZB24, I suggest considering at least the COASTLOOC and the COASTCOLOUR datasets describing bio-optical data for European waters both described on ESSD (Nechad et al., 2015; Massicotte et al., 2023).

2) Data availability

At Lines 208-210, it is stated: “It is anticipated that the number of measurement stations included in the present dataset is slightly higher than the one constituting the multi-spectral dataset. This is explained by the quality control applied to hyperspectral data, fully independent of other measured quantities”

Then at lines 285-286 it is stated “Minor differences between Fig. 4 and the equivalent one published in Zibordi and Berthon (2024) are explained by the diverse number of spectra (see § 4).”

Nevertheless, in section 2, tables 1 and 2 and the relevant text provide the same station numbers as in ZB24. Can you please provide some more details on the number of available measurement stations for the new dataset? For which campaigns, dates are measurement available that were not distributed in ZB24?

3) Exponential decay parameters

When describing the equations of the estimate of exponential decay parameters for adt(λ) and ays(λ), at lines 258-259 it is stated “Those authors determined the parameters defining the best exponential fit versus wavelength of adt(λ) and ays(λ) within the412-665 nm spectral interval”. If this was necessary in ZB24, where the data was only provided at six nominal center wavelengths (412, 443, 490, 510, 555 and 665 nm), I’m not sure it still makes fully sense for this data release where the absorption coefficients are provided at 3 nm spectral resolution at 2 nm increment in the 350-680 or 400-750 nm interval. In several papers different spectral ranges are suggested for “optimal” regressions over hyperspectral absorption coefficients, for example to avoid the pigmented spectral regions in adt(λ). For instance, In Figure 2, several adt(λ) spectra depart from a exponential decay spectral shape and at least for BLTS it seems that there are a few adt(λ) spectra with phytoplankton absorption features in the 400-450 and the 650-700 nm spectral ranges. Can you please discuss this issue?

The estimated Sdt and Sys are only summarised with a reference to ZB24 at lines 268-271.  Perhaps, it could make sense presenting the estimate parameters  in tabular form a as in ZB24 or graphically as scatterplots of Sys vs ays(λ0) and Sdt vs adt(λ0)?  Also, it’s not clear to me if the estimates of the exponential decay parameters using the hyperspectral absorption coefficients yielded the same values and regression residuals as in ZB24. Please clarify.

4) aph(442) versus aph(676) relationships

In Figure the 5 scatter plots of aph(442) versus aph(676) are presented and briefly described at lines 287-293.

At line 288 it is stated  “As expected, data show”, can you please provide references for these expectations?

Can you please discuss these findings in reference to other published studies and/or available bio-optical data?

5) Figures 1-3

The Y-axis ranges in Figures 1-3 vary quite a lot for each of the ten panels, as the spectral variability across marine regions is one of the main findings of this work. I wonder if it would be possible to choose few (two, three or four) ranges for each figure to help the reader in comparing the spectra across the marine regions. if not, please consider adding to the figure caption a statement noting that the Y-axis ranges vary across the panels.

A few typos  needs addressing:

line 72: “leading to a total of 637 stations”. I believe they should be 617, as the sum of 481 and 136 reported in table 1and as reported in ZB24. Please check

Table 4: the table legend refers to MAPD, but in the table itself APD is used in the first column

References

Massicotte, et al., The Coastal Surveillance Through Observation of Ocean Color (COASTℓOOC) dataset, Earth System Science Data, 15, 3529–3545, https://doi.org/10.5194/essd-15-3529-2023, 2023.

Nechad, et al., CoastColour Round Robin data sets: a database to evaluate the performance of algorithms for the retrieval of water quality parameters in coastal waters, Earth Syst. Sci. Data, 7, 319–348, https://doi.org/10.5194/essd-7-319-2015, 2015.

Zibordi, G. and Berthon, J.-F.: Coastal Atmosphere and Sea Time Series (CoASTS) and Bio-Optical mapping of Marine Properties (BiOMaP): the CoASTS-BiOMaP dataset, Earth System Science Data, 16, 5477–5502, https://doi.org/10.5194/essd-16-5477-2024, 2024.