Dear authors, dear community, 

the first version of the manuscript "A 16-year global climate data record of total column water vapour generated from OMI observations in the visible blue spectral range" by Christian Borger et al. has been submitted on 22.09.2021 (i.e. almost two years ago). During several revisions, the authors have put a lot of additional effort in improving the paper and also answering quite difficult questions by two expert reviewers.

The main point of criticism, however, was a bias w.r.t. reference datasets (here: ERA5) particularly across the tropical landmasses. And this bias is still present in the revised and most actual version of the manuscript and dataset and, hence, requires a closer look. Therefore, in comparison to the initial version, the authors have now included a detailed (and objective!) description of this bias including a discussion about possible reasons. Moreover, the authors also published product quality flags, with which users can filter out "bad data". 

This solution is indeed preferable to, e.g., simply removing all data with high deviations because:

1. The usefulness of data depends on the application. As also explained and shown by the authors, the observed systematic differences are less important for the analysis of long-term trends, which is a main application area of the product. Therefore it is better to leave additional filtering to the users.

2. As shown in the paper, the observed differences/uncertainties are in line with what is typically expected from satellite data products and also with the requirements for greenhouse gas estimates. 

Therefore, I am convinced that the bias and all other open questions have been addressed in a scientifically sound way. The manuscript now contains enough information that allows potential users to evaluate the potential (and alsl shortcomings) of this product for their particular application by themselves.