The comments were posted as a supplement file (PDF).

This study aims to identify meteorological variables affecting precipitation scavenging of atmospheric aerosols using a combination of aircraft, satellite, and reanalysis data augmented by trajectory modeling to account for air mass history. In literature, key variables controlling below-cloud aerosol scavenging have been well documented, although existing parametrizations still have large uncertainties (especially for particles in the submicron size range). This study addresses the former (key variables), but added little knowledge on the latter (how to reduce the uncertainties in parametrizations). The analysis approach is also subject to large uncertainties. I only provided a few comments related to the science of precipitation scavenging for the authors to consider in improving the quality of the manuscript.

Lines 25-28: Most existing parameterizations for precipitation scavenging of atmospheric aerosols have considered precipitation intensity, precipitation amount, and raindrop and aerosol size distributions, etc. It is not clear what (additional) variables this study proposes (after reading the whole abstract) for better parametrizing wet scavenging of atmospheric aerosols. The authors are encouraged to provide an explicit recommendation instead of a general statement.  This comment also applies to the Conclusions section (the last paragraph) since no clear message of their recommendations is provided.  

Section 2.2: If BC and CO have the same sources but different sinks (one is wet scavenged and the other is not), then using this ratio approach is reasonable in tracking the wet scavenging of BC. However, if they have different sources along the air mass trajectory (which is likely the case), this approach would cause very larger uncertainties.

Lines 78-92: In Chemical transport models, time splitting approach is used for calculating each physical and chemical process (emission, transport/diffusion, chemical transformation, deposition (in-cloud nucleation, below-cloud scavenging, and dry deposition)). Each process needs to be calculated or parametrized as accurate as possible. To improve the parameterization of below-cloud aerosol scavenging, collecting field data through measuring aerosol concentrations before and after precipitation events covering different precipitation types and intensities would be the best approach, in my opinion. The approach used in this study involves both in-cloud and below-cloud scavenging contributions as well as additional entrainment of pollutants along the air mass trajectory, and is not possible to quantify precipitation scavenging. I have difficulties in finding the true scientific value in such an analysis for improving our understanding in below-cloud aerosol scavenging (which seems to the major goal of this study).

Line 11 and line 31:  This sentence implies that wet scavenging is the only dominant sink for aerosol particles. I would consider both dry deposition and wet scavenging as dominant sinks. Maybe change to “As one of the dominant sinks”.

Lines 43-44: You should also include “precipitation intensity, amount, frequency, type”. Precipitation type includes liquid and solid (snow) precipitation.

Lines 41-50: I would like to draw your attention of a series of studies on below-cloud aerosol scavenging conducted by a group in Environment Canada to enhance the discuss presented in this paragraph. They not only systematically assessed important parameters affecting below-cloud aerosol scavenging (Wang et al., 2010, ACP, 10, 5685-5705; Wang et al., 2011, ACP 11, 11859-11866; Zhang et al., 2013, ACP 13, 10005-10025), but also developed a new set of semi-empirical parameterizations (Wang et al., 2014a, GMD, 7, 799–819; 2014b, JAMES, 6, 1301-1310).   

Line 97: Was the experiment cover one single site or one big area?

Lines 229 and below: RH is identified as a key variable, but it is likely because its directly link with precipitation. In this case, there is no need to include this additional variable in precipitation scavenging parameterization. If the effect of the high RH is through hydroscopic growth of particles, then this needs to be mentioned.