Minor comment: I do not think that the surface statistics presented in Table 1 and 2 suggest “the importance of including the contributions of this source for improving the spatial and temporal variability in model predictions”. They show a very marginal gain. It is not surprising because the base model is “tuned” (input parameters and schemes adjusted) to match as well as possible existing observations. Therefore, even by making the model more realistic in terms of emissions or physics, one may degrade initially the model performances. Overall, regarding the importance of LNO, results are much more convincing for the free troposphere. I do not think you need to add “suggesting…predictions” here. The point is made more naturally in the conclusion.