|Peng et al present a revision of their manuscript combining an AAC and single-particle mass spectrometry to determine chemically-resolved single-particle effective densities. As noted previously by both reviewers, this work is interesting and useful. However, both reviewers pointed to a major manuscript weakness being limited citations to previous, highly relevant work and some explanations needing clarifications. In addition, several previous comments focused on stated uncertainties and errors in the results. Several of these issues have been corrected, but some still remain, as noted below.|
The authors added sentences in the introduction discussing the use of particle light scattering to obtain effective density, and these added sentences on Lines 87-94 are useful. However, most previous and similar work has used a DMA with single-particle mass spectrometry, and both reviewers pointed to the need to discuss this previous work in the introduction to place the current similar work (that combines AAC with single-particle mass spectrometry) in context and make it clearer how this current work builds upon this previous work. However, the authors chose not to do this, with the exception of adding a statement that the derivation of Eq 8 was previously presented in Schneider et al (2006) (lines 82-83). I still believe it is a major weakness of the paper to not discuss previous work combining a DMA with single-particle mass spectrometry (see previous reviews) in this manuscript’s introduction.
Following on the previous comment, the authors did not rephrase the problematic statements that give the impression that this is the first work to measure chemically-resolved particle effective density (previous lines 90, 150-151, 298-299, now lines 98, ). These sentences include: “The aim of the present work is to develop a method to obtain Dve and pe.” (line 98 in current manuscript) “These two properties cannot yet be measured for unknown particles by current techniques.” (when referring to Dve and pe on lines 167-168 in the current manuscript) “…to first characterize the Dve, pe, and chemical compositions of atmospheric particles…” (lines 321-322 of the current manuscript) As previously requested, these misleading statements should be rephrased.
Throughout the manuscript, error should be reported with one significant figure. In addition, while error was added to the PSL sizes on Lines 139-142, it is overstated. For example, 203.0 +/- 5.0 nm shows greater certainty than 203. +/- 5. nm, which would be in line with the manufacturer’s stated uncertainty. Also, in the response, the authors state that they fixed error to be reported with one significant figure in Table 1, but all of the errors shown in this table still are 2-3 significant figures.
The authors state in their response that “Considering the precision of the PSL size is less than that of the instrument of AAC, the discrepancy between the measured value (from AAC-SPAMS) and the true value (the size and density of PSL) is used to represent the measurement uncertainty, which is presented in Figure 2.” However, the error in the AAC size needs to be stated in (added to) to the manuscript for the reader to evaluate this statement. It appears that these measurement uncertainties were not propagated through to the reported deviations in size and density in Figure 2, so the reader cannot evaluate where the major source of error is originating, which is important for the results. The authors at least need to state the errors in the manuscript to justify their method. It is also critical that these uncertainties be clearly explained so that the number of decimal places used in presenting the data in the results can be properly evaluated.
The ambient sampling section of the Methods (Sec 2.3) needs to state the dates of sampling, as previously requested, since the seasonality will impact particle composition, for example.
Lines 304-306: Errors were reported for all particle types in the previous version of the manuscript, but now these are missing for the Amine, EC-N-S, and EC-S particle types for an unknown reason.
Consistency between Methods and Figure 1 caption: It is now stated that the system flow rate is 0.3 lpm, but line 153 states an “overall sampling flow of 3 lpm”. Line 124 states “The exhaust flow from the AAC was about 0.3 lpm.” However, the sampling flow rates of the AAC and SPAMS are not stated in Section 2.1. Please add these to Section 2.1, clarify the flow rates at each part of the system in Figure 1, and clarify the discrepancy between the figure caption and line 153.