|I think the paper is improved, particularly length aliens of figures. But I still strongly disagree with their size distribution data and this needs to be clearly laid out before publication. My primary complaint is not with the shadow based instruments like the CIP which have measurement capability above 15 um (although I would be careful about interpretation at th elwo end of the spectrum). If they find giant particles with the CIP I believe them. Rather I take exception with the forward scattering components of the CDP probe. The way they have worded the comparison is misleading, and fails to directly address the probe degeneracy issues. Currently, they simply say the differences they found between AERONET and the airborne probes have also been reported the literature. |
“Previous work (Müller et al., 2012; Müller et al., 2010; McConnell et al., 2008) has found relative
864 disagreement between aircraft and AERosol Robotic NETwork (AERONET) size distribution retrievals
865 for dust.”
Well, so did we in Reid et al., 2003 and we explained it. They should flat out state what the problem is and why many of us think it is a fundamental problem with the probes. They also fail to mention that while “Müller et al., 2012; Müller et al., 2010; McConnell et al., 2008” saw differences between wing mounted probes and AERONET that, Reid et al., (2006) and Reid et al., (2008) actually both showed good agreement between the AERONET retrievals and aerodynamic particle counters for coarse mode aerosol particles-and I was looking hard for a problem with the retrievals. Neither did they mention that in the Reid et al., (2003) intercomparisons papers AERONET compared well to the APS well too, as did the field measurements in Izania by Maring et al., (2003). In all of these cases with OPCs, mass closure was demonstrated in the APS and not in the wing mounted probes, which as you would expect with oversizing overestimated particle mass. Now, while the authors may object that these APS comparisons are against ground instruments, the comparisons were the same for when the aerosol particle loadings were dominantly in the boundary layer. Once they fully lay out the issues, only then can they ethically punt and say “we will investigate this further.” On these grounds, even though it is a minor change to the text, I am going to continue to suggest major revisions because this has important downstream implications throughout the paper.
There are remaining interpretive issues with the comparison that should also be noted. As for their one case where AERONET suggest particle VMD larger than 10 um, “However, one retrieval only shows a peak volume concentration at 13 μm which appears to agree much more closely with the shape of the size distribution from the aircraft measurements.” this could be cirrus contamination-which is a problem with AERONET. If there is convection in the area, it could very easily be cirrus as much as local generated (I honestly don’t know which). They should contact the AERONET team for an opinion if they believe it. Similarly, as I mentioned before, looking at Figure 11 in a linear scale gives you quite a different perspective. Finally, I went through the Rosenberg paper in detail which they use has reference for calibration. I would point out to the editor that if you look at Figure 1 (b) which shows the CIP calibration, you can just barely see the Mie inflection point in the 5-10 um range. But based on this scale, you would not notice it if you were not looking for it. Then, if you look at figure 3 left hand size where they calibrated, the smallest particle size they calibrated on is 15.9 um, well above the Mie inflection. I have seen some folks get it close to “right” using mono-disperse sizes around 4 and 8 um. But if you give the distribution any width at all, the 10 um peak returns. Sorry, I have done this many times myself. It then follows that I object to their size distribution methods section and final bulletized conclusion as they fail to mention these issues in their comparison to AERONET.
I am very sorry to have to be a hard ass on this. But this is a major issue with how one interprets airborne datasets and is a recurring problem in the field. It goes way beyond the AERONET comparison. Somebody is going to think these size distributions are real, and do some radiative or mass flux calculations. This does not mean that the probes are useless. I find good use for them in every campaign, and the overall vetical profiles if converted to extinction space are likely very good. One just needs to be careful about how one interprets the data. Please declare it.
Maring, H., D. L. Savoie, M. A. Izaguirre, L. Custals, and J. S. Reid (2003), Mineral dust aerosol size distribution change during atmospheric transport, J. Geophys. Res., 108(D19), 8592, doi:10.1029/2002JD002536, 2003.
Reid, J. S., E. Reid, A. Walker, S. J. Piketh, S. S. Cliff, A. Mandoos, S. Tsay, and T. F. Eck (2008), Dynamics of Southwest Asian dust particle size characteristics with implications for global dust research, J. Geophys. Res., 113, D14212, doi:10.1029/2007JD009752.
Reid, J. S., B. Brooks, K. K. Crahan, D. A. Hegg, T. F. Eck, N. O'Neill, G. de Leeuw, E. A. Reid, and K. D. Anderson (2006), Reconciliation of coarse mode sea-salt aerosol particle size measurements and parameterizations at a subtropical ocean receptor site, J. Geophys. Res., 111, D02202, doi:10.1029/2005JD006200.