Dear authors,

The paper is improved and most of the reviewers' comments have been addressed. However, one comment remains to be addressed and this is related to the uncertainties of the microphysical retrieval and the DRF impact of Canadian smoke.

Specifically, the fact that the inversion provides reasonable estimates for dust, does not support its applicability to smoke particles which have different refractive index and shape. Even more so, the algorithm used (i.e. Veselovskii et al. 2010) provides reasonable estimates of volume and effective radius of dust, by reproducing the depolarization measurements along with the backscatter and extinction. This is not the case presented here, since the measured depolarization values are not reproduced.

It is expected that the uncertainties are larger than the ones reported in the paper and that this uncertainties propagate to the DRF estimation as well. This is a serious concern that has to be addressed, otherwise the paper results are not sufficiently supported.

I propose that the authors will clearly state the level of uncertainties in all aspects discussed (from the inversion to DRF). This statement should be present also in Abstract and Conclusions section. The paper should stick on the EARLINET observations presented and how these can be utilized for radiative transfer applications, stating though that this is just a demonstration of future potential application of EARLINET data and techniques on DRF studies, while we are still working on the uncertainties.

Looking forward to your response,
Vassilis