Interactive comment on “Modeling the Zeeman effect in high altitude SSMIS channels for numerical weather prediction profiles: comparing a fast model and a line-by-line model” by R. Larsson et al

I have just three minor comments. 1. I did not get a feeling for the significance of the results shown in Fig. 6 which show discrepancies between 3D and 2D magnetic field treatment of up to +/-7K in the tropics. This sounds a lot and makes me wonder if modelling from 2D is in fact useless for these higher channels. A question is how does this compare to variance with respect to climate in observation space at these altitudes / region? If this is less than 7 K then the observation error if processed with a 2D magnetic field is large compared to the climatology error, meaning the information content of the obs even against climatology is small. This would imply to use these data we need to use a 3D field. So its important to get a feel for this. NWP models are pushing higher and the time will come when they beat climatology by some margin. So I would like to understand the relative size of these three errors: climatology, current and potential future NWP, and observation error with 2D and 3D fields. I think such information would further enhance the usefulness of the paper.