Dear colleagues,

we received two assigned reviews and two additional voluntary comments. All reviews were positive asking for minor revisions and clarifications. Therefore, I am pleased to accept your paper subject to minor revisions (review by editor).

Best regards

Lars Kaleschke

Dear colleagues,

thank you for carefully considering the comments and questions of the referees. I have some additional comments and suggestions for a minor revision.

The neural network method works well if the training data are from similar conditions. Therefore, the method works well for the springtime central Arctic where aircraft data are available. However, it is questionable if this is the case for different seasons and regions (i.e. for ice formed from low salinity water on the Siberian shelf). For SMOS we have investigated the sensitivity of brightness temperature to several parameters (see Maaß et al. (2013) and her PhD thesis). Knowing the snow (or ice) surface temperature is of great importance for the snow thickness retrieval from 1.4 GHz. This information does not go (directly) into your neural network and therefore causes uncertainties. I therefore doubt that the current neural network approach is the best choice for CIMR and instead suggest to ask for further research on methods. You may write sentences like the following in a more neutral way and consider physical-based radiative transfer techniques as a perhaps/likely superior choice for the retrieval of sea ice parameters from CIMR.

"An adapted version of the AMSR2 + SMOS snow depth on sea ice neural network retrieval would be extremely valuable"

Table 1+2): What is the meaning of negative R²? Please check!

Tables should have headings and not captions.

References are not complete, all refs should include DOIs where available.

Incorrect use of \citet, please check.

Best regards

Lars Kaleschke

Dear authors,

congratulation, I am pleased to accept your revised manuscript for publication in The Cryosphere! I am looking forward to the application of this method for the Copernicus radiometer which will hopefully be launched in a few years.

Best wishes
Lars Kaleschke