Review of “Vertical profiles of global tropospheric nitrogen dioxide (NO2) obtained by cloud-slicing TROPOMI” by Horner et al.

8/16/24

Summary:

This paper describes an extension of previous efforts in total column NO2 cloud slicing in order to obtain vertical profiles of NO2 amount. Building on the work of Marais et al. [2021], who derived NO2 mixing ratios in the upper troposphere from a single year of TROPOMI data, the authors refine the earlier technique in several ways and extend the analysis to the surface, based on 4 years of TROPOMI using a self-consistent retrieval throughout the period. The new cloud-sliced climatology is compared with both aircraft in situ data and GEOS-Chem model output. The analysis presented is very thorough and thoughtfully presented.

Comments (minor):

• In section 2.1, the cloud-slicing technique is nicely explained in the text. The addition of an equation(s), would make it easier to understand how the authors go from a total column amount to mixing ratios in individual layers of the troposphere.
• Also in section 2.1, a better of explanation of what “informed by thresholds used by Choi et al.” (line 131) means is needed.
• Please explain what “NO2 mixing ratios within each layer are relatively well mixed” (lines 149–150) means. Do you mean NO2 is well mixed or the mixing ratio is constant within the layer?
• Better rationales for the thresholds used in the analysis are needed. For example, comparisons between cloud-sliced and GEOS-Chem grids are made only when at least 10 cloud-sliced data points are within the cloud-sliced grid (lines 248–249). Why 10 and not 5 or 15?
• Figures 1 and 2: Recommend adding “n=” to the insets in each panel, indicating the number grids used. It will make the figures easier to interpret without reading the text.
• It is hard to derive much quantitative information about the cloud-sliced vs. aircraft NO2 from Figure 4. The maps are valuable in defining the geographic regions analyzed subsequently, and the larger circles show where the aircraft data are nicely. No change is requested—just a comment.
• “The greater variability in the DC-8 data in each layer (larger interquartile ranges), is because DC-8 are single year measurements, whereas cloud-sliced NO2 are multiyear means” (lines 353–354): I do not understand why the longer time period would necessarily reduce the IQR. In fact, I might have expected the opposite, as multi-year variability might increase the range of values sampled. Could this instead be due to a difference in the representativeness of the measurement, seeing as the TROPOMI footprints are larger than the DC-8 sampling and thus average out smaller-scale features? In any case, more explanation is needed of this quoted statement.