Thank you for providing a revised manuscript. In reviewing it, I have found it quite dense and disjointed (even for a model description paper). Hopefully we can improve it prior to publication, so it will be accessible and helpful to readers. To that end I have the following suggestions for technical corrections prior to publication:

-Improve English and conciseness of Abstract and Introduction
-better differentiation of COSMO-CLM and Community Land Model (CLM) acronyms. Quite confusing for the reader
-A more comprehensive ‘tie-together’ of all the models used. Currently, to the fist-time reader it appears as somewhat of a laundry list. For example:
—better high-level justification of reasoning for tying together so many disparate models
—Justification for the couplings to different ocean models, depending on basin
—A better model-specific justification for why each model was chosen
-is the the full MPI-ESM coupled model is used, or just the atmosphere? The first time reader is left confused as to the explicit connections between this coarse global model, and COSMO-CLM, which are only revealed as 2-way coupled, deep in the technical descriptions (unless I missed something)
-"An overview of the coupled models selected for coupling with COSMO-CLM (CCLM) is given in table 3": the term coupled models implies they already all coupled. It is thus confusing to read that they will be coupled *again* to something else.
-confusing: "atmosphere-atmosphere coupling" or “ocean-ocean coupling”. Need to be more clear that it is global->regional one/two way coupling (if I understand correctly)
-It is confusing that some models being coupled to COSMOS are component models (e.g. CLM), while MPI-ESM is a full coupled model. A clearer up-front description of this differentiation (if correct) is necessary
-“The climatological means of fresh-water inflow of 33 rivers to the North Sea and the Baltic Sea are collected from Wikipedia.” A more robust source is needed here.
-Check that all acronyms are first defined in the text, and then the acronyms are only used in the future. For example, COSMO-CLM is defined as ‘CCLM’, but then ‘COSMO-CLM’ is used again in full, later.
-There are two sections, titled ‘OASIS3-MCT coupling method and performance optimization’, followed by ‘OASIS3-MCT coupling method’. These titles are redundant, and I suggest a careful merge and re-write of these sections into one coherent section.
-It’s not clear to me early in the manuscript, whether MPI-ESM is used only as generators of boundary conditions for the regional model, or if two-way coupled has been implemented between these. Please clarify early on, so the readers are prepared for the later, more detailed descriptions.
-“Here send_fld ends”: what is send_fld?
-Has the two-way-coupled CCLM+MPI-ESM system actually been prognostically run (as opposed to using a prescribed COSMO-CLM solution)?
-In Conclusions: perhaps make a more ‘tied together’ set of conclusions, from the many paragraphs that start with the words: “The optimum configuration of the coupling…”. Or alternatively, make bullet points. In either case, the optimal coupling strategies should be more accessible to readers, who are potentially interested in each individual coupling.
-In Conclusions: I think it is necessary in the concluding section, to return to the initial scientific reasons for constructing COSMO-CLM couplings in the first place. This will remind the reader why this technical work is relevant to a broader research efforts, and perhaps motivate the reader to work with these couplings themselves.