|There are improvements to this manuscript and there is valuable content that should be published. The most useful parts of the paper are validation, the comparison with ACE, discussion of the high HCFC22 values filling the subtropics at 16 km and how it affects zonal means (Figs. 10-12), the global distributions and temporal evolution (Fig. 13-16), and comparison with surface networks. These sections alone would make a solid paper. I recommend publishing a paper of reduced scope, but not the current version.|
The primary problem with this version remains the writing. The authors’ response to my original comments on the writing placed the burden of writing issues on the journal:
‘We have tried to improve the language of the manuscript and the paper has been proofread by native English speakers. However, language editing is now routinely done by professional language editors of ACP as part of the journal publishing process. Thus, such issues should not affect the acceptance of the paper.’
The problems are beyond proofreading. The problems include syntax, odd word choice, misuse of terminology, lack of clarity, poorly organized explanations, and failure to put new findings in the abstract and summary. The writing quality is such that it does affect acceptance. This is my primary reason for not accepting the manuscript in its current form. There are also a few science issues – detailed below. Directly below are some examples of the writing problems; they are not meant as a comprehensive list.
Example of writing Issues
Section 5.5 is a single paragraph of nearly 60 lines. Poorly organized writing is a symptom of poorly organized thoughts.
The authors did not like my comment about use of the word ‘sounded’ with respect to measurements. Their response (‘In the manuscript we have never written that measurements are sounded.’) was not to search the document for the problem areas but to deny they exist. Please see line 122 (“MIPAS sounded the atmosphere…”) and lines 393 (“…is caused by difference in the air masses sounded by each instrument…”). One could say that MIPAS remotely senses the atmosphere or that it measures thermal emissions in the limb. A passive measurement (MIPAS) does not ‘sound’ the atmosphere.
An example of confusing statements is found lines 153-161. First it is stated that the retrieval uses components relevant to non-LTE, and the next sentence says those components aren’t relevant and are not used. So which way there the measurements retrieved?
Lines 602-606. So hard to extract the meaning (about description of a feature in Figs 10 and 11). After reading the paragraphs that follow, I think I understood what you were getting at. If you discussed Figure 12 (before Fig. 10 and 11) then you could more easily explain the features in Figs. 10 and 11.
Avoid subjective adverbs ‘nicely’ and ‘perfectly’.
Section 5.2, lines 684-701. There are several occurrences of the phrase ‘numerical mixing’ used to refer to the effect of taking zonally asymmetric features and viewing them in the zonal mean. This is not what numerical mixing means! First, there is no actual mixing going on, and second, ‘numerical mixing’ is a widely used expression that refers to subgrid scale mixing in models. Similarly, the phrase ‘tropical pipe’ is erroneously used in the discussion of vertical transport in the tropical troposphere. The ‘tropical pipe’ refers to the isolation of the tropical stratosphere.
Line 690. Wave activity does affect the polar latitudes but the waves originate in the midlatitudes. They are not referred to as polar stratospheric waves.
Line 718. While the QBO has phases, the Brewer Dobson circulation does not. It has a seasonal cycle.
Line 802. The ‘semi-annual’ variation in northern midlats at 30 km. What variation is this? The SAO was previously mentioned, are you implying that’s what you see here? The SAO is a tropical upper stratospheric phenomenon and has a very weak effect outside the tropics at this altitude.
Line 920. “This reason is roughly this.” Try phrases such as ‘this can be explained by…’ or just delete.
The authors’ response stated that they had 3 new findings in this paper:
“The fact that the mixing of monsoon air into the tropics is instantaneous and not restricted to the time when the monsoon breaks down is to our best knowledge a new finding.
The fact that the temporal development of HCFC-22 in the stratosphere cannot be explained by the trends measured at the surface and the known age of air induced time lag is also a new finding.
The fact that Asian HCFC-22 emission have become the major source of global upper tropospheric HCFC-22 has been plausible but our data are to our best knowledge the first empirical evidence in support of this hypothesis.”
If you have new findings they belong in the abstract and in the summary, not buried in the paper. They don’t appear in the abstract and only 1 appears in the summary.
Fig. 6. While I do see two difference colors of points, I do not see that they clearly fall into two clusters with respect to bias. The differences are small.
I agree that Section 4.3 (comparison with MkIV) has some value in the validation exercise, but Section 4.2 does not. The authors’ justification is that the cryosampler comparisons with other MIPAS trace gases are published therefore this comparison should be too. The papers referenced are in fact not published but are under review. The criteria used for the cryosampler data comparison (1000 km co-location) is too broad to be of use. This problem is demonstrated by the near total lack of agreement shown in Fig. 8. This section adds no value to the validation.
(You might be able to use the cryosampler data and avoid the co-location problem by comparing the tracer-tracer correlations found in MIPAS and the cryosampler. The Eckert and Laeng manuscripts use CFC11, CFC12, and CH4 observations, yes? If a goal of these comparisons is to (line 568) identify a bias due to spectroscopic information, you might be able to use the CFC11 & HCFC22 (or CH4 and HFC22) correlations for each instrument to show the bias.)
Table 3 trend comparisons. The surface growth rate of HCFC22 was fairly constant from 1992-2005 then increased. See http://www.esrl.noaa.gov/gmd/hats/graphs/graphs.html. Stratospheric air is typically between 2 and 5 years age. Because of this you cannot expect that a MIPAS trend between 2005-2012 will be the same as trends published from any stratospheric date pre-2007. Also, given the unknown bias (as reported in this paper), a trend of 6.15 +/- 0.24 (for example) shows at least 1 too many significant digits.
Section 5.5. ‘unexplained trends’. Consider these three facts in the context of the claim that this 7 year data set shows ‘an unambiguous sign that the circulation must have changed over the MIPAS period.’ 1. The changes in stratospheric HCFC22 from 2005-2012 are dominated by large increases in emissions. 2. You have measurements that you think are not biased, although you report there are biases at some altitudes with respect to other measurements. In fact, you aren’t sure of the accuracy of these measurements. 3. Studies of mean age using much longer data sets and a more careful assessment of the uncertainties find no indication of a change in stratospheric circulation (except a very slow one in the lowermost stratosphere) (Boenisch et al, ACP 2011, Engel et al., Nature 2009). Regardless of the fit you get to the time series, it is not credible that a 7 year time series of a rapidly growing trace gas could identify a change in the stratospheric circulation. It is extremely unlikely that your method is accounting for all processes affecting stratospheric variability, for the actual growth rates during this period, and that what remains must be an indication of circulation change.