Interactive comment on “Global HCFC-22 measurements with MIPAS: retrieval, validation, climatologies and trends”

This manuscript describes the retrieval of MIPAS HCFC-22 proﬁles from 2005-2012, validates them, and presents some ﬁgures showing its morphology. It concludes with a discussion of stratospheric trends from this short data set. It is long and poorly written, especially the retrieval section. The validation section underused (maybe misused) the ACE data, which are a valuable tool for the validation. Two balloon proﬁle data sets are also used but are not appropriate for validating the MIPAS data. There is little analysis of the MIPAS data. Figures are presented and described but not analyzed in any quantitative way. There is no HCFC-22 climatology in the Climatology section. No meteorological data are used to support descriptions of HCFC22 behavior or statements about processes affecting the distributions. Many statements about HCFC22 strato-C4691

Introduction p. 14786, Line 10. The most current determination of the atmospheric lifetime of HCFC22 can be found in the 2013 SPARC report (a WMO publication). The SPARC lifetime is the same as reported here but it's a more current assessment of the lifetime.

Section 2, MIPAS data
No mention is made of the MIPAS sampling pattern. What is its latitude range? Is the range covered daily? Does is sample more at some latitudes than others? Does it measure all latitudes in all seasons?

Section 3, Retrievals
This section is not well organized and it reads like a series of unconnected details regarding the retrievals. For example, the last paragraph of the section is on information C4692 Interactive Comment By examining a limited latitude region and sorting by season, Figure 5 provides the most useful comparison of this section. The two data sets agree within their uncertainties below 20 km all the time while seasonal differences are revealed above 20 km. Why, then, do you combine all latitudes and seasons in the other figures? So much information is lost. In fact, Figure 6 gives the opposite impression as Figure 5: it shows continuous, nearly 1:1 agreement below ∼25 km or so. (Thus I cannot understand the statement that the points fall into 2 clusters, p. 14794, line 2). In general I find that the words written in this section do not align with what is shown in the figures. For example on p. 14793, line 11, 'The bias is significant at all altitude levels.' It clearly isn't! In Figure 7, there is a bimodal shape in the lower right histogram, but as global data are combined in this figure, who knows why this shape occurs? (But a different comparison would probably reveal the answer.) Overall, the ACE comparison is inadequate. It C4693 ACPD 15, C4691-C4698, 2015 Interactive Comment

Full Screen / Esc
Printer-friendly Version

Interactive Discussion
Discussion Paper could be a more valuable and useful part of this paper with comparisons that examine specific seasons and latitude ranges.
Section 4.2, Cryosamplers p. 14795, lines 3-6. These sentences are poorly worded. They serve as the introduction to the topic of the next paragraph and should be combined with it. The coincidence criteria is so broad (1000 km and 24 h) that when the profiles do not agree you really can't know why. If you use some meteorological analyses to show the profiles are from similar environment, then you would know whether it made sense to compare them. As it is, Figure 8 shows mostly a lot of disagreement with the balloon profiles but the reason is probably geophysical variation -thus there is no point to these comparisons! And why calculate a 2005-2011 mean profile at all? This gas is increasing rapidlyat least 25% over this time period. The multi-year mean is meaningless, and should a balloon profile match it, that is meaningless too. Unless you can demonstrate that in spite of the broad coincidence criteria that it makes geophysical sense to compare with MIPAS, these comparisons could be eliminated. the ACE data comparisons as noted above. After a more thorough validation using the ACE data, I think you will be able to state much more definitively where the data sets agree and where there is bias (and how much).
Section 5. Climatology The data are presented in various ways (e.g., latitude v. time, altitude v. latitude, etc) but there is actually no climatology here. In general, the 'analyses' in the subsections are only descriptive (i.e., descriptions of what is already known) or speculative, and do not present any quantitative analyses. No meteorological data are used in support of speculative statements about processes that might be indicated in the data. This section would be improved by including a climatology (i.e., mean distributions as a function of month/season, mean cycles, etc.) and by adding meteorological analyses to give support to the processes you describe. Sections 5.1 and 5.2 have lots of qualitative discussion but there is no actual MIPAS data analysis that demonstrates any of the processes discussed; e.g., p. 14798, line 22: "The following scenario is suggested. . .". Analyses, not suggestions, are required for publication. Why not bring in meteorological data to support your ideas? Also, Figure 15, which is used to show something about monsoon transport, crams 6 years of data on 5 surfaces into a very tiny space. Most air enters the stratosphere through the tropical tropopause, so interhemispheric (IH) differences in long-lived trace gases found in the troposphere are usually not found in the stratosphere. (CO2 has some but they disappear quickly with height.) If you speculate about IH differences in the stratosphere, you'll need a supporting analysis to demonstrate that tropospheric IH source differences are the cause.
p. 14800, line 14. 'Interestingly, the breakup of the vortex seems to take place at all altitudes at almost the same time for the northern polar region.' The Arctic vortex final warming ('breakup') occurs in March or April and shows considerable interannual variability in how it breaks up (e.g., wave 1 or wave 2 warming). Its variability is much greater than the Antarctic vortex, so your statement doesn't make sense. The sharp discontinuities found each year in your figure appear too early in winter to be the breakup -they are probably midwinter sudden warmings, not the final warming (breakup) that occurs in March or April. Please check your figures and interpretation.
The lowermost stratosphere is generally below 16, not 20 km. It is below 380 K (below Hoskins' 'overworld' p. 14805, lines 21-24. Again, this is more speculative discussion that is not supported by meteorological analyses or other trace gas measurements. This section could be very interesting if you examined the HCFC22 data as a function of altitude in the lowermost stratosphere (e.g., 12-16 km) in each hemisphere to identify seasonal transport processes in the lowermost stratosphere. Interhemispheric differences between the results may reveal process important to the cycles in each hemisphere. p. 14806. I see that you recognize that chemistry, emissions, and transport are all important to understanding HCFC22 behavior. But this means that it can't be understood as simply as discussed here. An atmospheric model is required to adequately interpret the behavior.
Section 5.5. Stratospheric trends This paper examines a 7-year data set. Seven years is less than 3 QBO cycles. Each QBO cycle is different in length and its seasonal timing. It is not completely accounted for in the regression analysis by considering terms at 2 pressure levels. Because the regression analysis cannot adequately remove the QBO effect, the residuals (Figs. 17 C4697