The data description seems thorough and well-organised and I suspect this product will serve many users. Please, however, can the authors attend to a few points:

1) I feel somewhat surprised to see that these authors have ignored two coccolithophore data sets recently published in ESSD - Loveday & Smyth, https://doi.org/10.5194/essd-10-2043-2018 which seems quite relevant to the remote sensing aspects albeit using AVHRR rather than ocean colour, and Daniels et al. https://doi.org/10.5194/essd-10-1859-2018 which addresses the issue of calcification rates. In one of their responses the authors mentioned that the impacts of these blooms on pCO2 represented the unique contribution of this data. But these data also both draw on and contribute to the other two data sets?
Such a comparison may also prove useful for validation (e.g. see https://doi.org/10.5194/essd-10-2275-2018).

2) Please also attend to these comments from a third reviewer:

"I found the paper interesting and useful. The authors have made every attempt to validate their products and build on previous work, and to provide a theoretical basis for the algorithms where possible. They have brought together a large body of different types of data to generate their products, and to validate them. They have taken the trouble to re-process some of the OC-CCI data, when it appeared that the masking applied may not have been appropriate for their particular application.

While this paper was under review, another paper on the same topic, but using AVHRR paper, has been published in ESSD (Loveday and Smyth, ESSD 2018). It would be good to refer to this publication. That other attempts have been made recently to address a similar problem in no way deters from the value of the paper under review: the users should be given the opportunity to accept the product that best suits their requirement, and to evaluate the products themselves.

The writing style leaves some problems with the grammar, and some instances where the statements lean towards the hyperbole (especially in the introduction). I do not know if ESSD editors and copy editors can help the authors deal with them?

Assuming that such minor problems with the language can be fixed, I recommend the paper for publication."

Pending appropriate responses, I may also ask one of the reviewers who volunteered to read a revised version to take a quick final look.

Thank you for considering ESSD.