Dear authors,
I has not been easy to come to a decision about this paper because the reviews were so divergent. I've gone through the comments, your replies and the revised manuscript and came to the conclusion that you have adequately addressed most of the reviewers' concerns and that the paper is useful and technically sound. I think it can be published after some final polishing. My specific comments are listed below. I hope they are helpful.
Best regards,
Thomas

All page and line numbers refer to the paper version attached to the rebuttal letter.
p3 l15-16: Here I think you intended to condense too much into one sentence, leading to the expression "unususlly tropical characteristics". I guess you mean something like "... air with tropical characteristics which is not usually seen over Lauder..." I suggest to `deconvolve' this expression and, if need be, split the sentence into two.
p4 l4: "and a discussion is presented in Section 5" seems to be a somewhat empty statement. I would prefer if you could be a little more specific here. e.g. "...we summarize the most prominent anomalies in our data set along with our suggested explanations" or something like that. By the way: I miss a conclusions Section. Since the discussion takes already place in Sections 3 and 4, one could think about changing the current "Discussion" section into a "Conclusions" section.
p4 l.23: Would "spectral opacity of the troposphere" clearer? I am not an expert in microwave measurements but the term "tropospheric opacity" seems to imply a scalar value and thus should only be used if it is meant this way.
p5 l.5: "selected spectra chosen" sounds a bit circular. Do you mean "...continuously, and spectra associated with large or highly variable tropospheric opacity are discarded from further analysis..."
p5 l.9: Here it would be interesting to describe how long it takes to make one single measurement. Later in the paper you mention the term `integration time' but here you just talk about 6-hour data periods and the reader does not know if you average all the measurements in a 6-hours bin or if one measurement takes six hours. You could introduce and explicate the term `integration time' already here.
p.5. l. 13: "value is a bit vague, because the agreement depends probably on the specific definition of the quantity, and particularly the vertical resolution. Write "concentration" or "mixing ratio" or whatever the correct quantity is instead of value, and mention what layer thickness or altitude resolution the quantity refers to.
p.5. l22/23: "above the horizon" seems obsolete to me, because this is already said with the term `elevation angle'.
p.6 l.3: Wouldn't "difference spectra" be clearer here? I am not sure if I understood the text correctly. One way to understand it is that the y-vector in the retrieval is actually the difference of two spectra. lines 21 and 22 of page 5 seem to suggest this. The other way to understand it is that the y-vector in the retrieval is a concatenation of the two spectra recorded under different angles. In this case some information is obtained from the difference but formally it is not the difference which is analyzed. According to what understanding is correct, please insert either 'difference' in l. 3 or rewrite the first part of this paragraph accordingly.
p6 l 5: I do not think that the term 'independent variable' is adequate here. In the RTE, the concentrations are the independent variable and the radiance the dependent variables, and in an inversion context the radiances are the independent variables and the concentrations the dependent variables. I suggest to replace this by 'altitude coordinate'.
p6 l11/12: It is not quite intuitive why you mention already here that the averaging kernels are not used. I would prefer that you mention this when the actual MOPI1-Satellite comparison is made. I think this is more adequate because there is no such comparison over a couple of pages, and the first one shown does use the averaging kernels.
the page where Sect 5 begins, l 5: 'ending' typo double 'd'
same page l5 ff: The age of air trends by Stiller, 2012, 'Observed temporal evolution of global mean age of stratospheric air for the 2002 to 2010 period', ACP 12 3311-3331 seem to be consistent with this (see their Fig 10 top panel).
Fig 7 Caption: "shown 4-times annually": Here it is not quite clear to me what is meant. "for four datas sets"?
Fig 7: The left axis caption should not only include the unit but also the quantity, e.g. "ozone anomaly / ppmv"; The lower axis should be "year"
Fig 8: The numbers of the isolines are hard to read. Could you make them larger or provide a colour scale?
Figures, general: I have just learned that, according to the relevant iso-norm, the units in plots should not be written in parentheses but like "quantity / unit". I will not force you to make all these changes for this paper but in the long run it may be better to convert the plotting routines towards the iso norm.

Dear Gerald and colleagues,
your paper is now ready for publication in ACP. Thanks a lot for considering my recommendations. The publisher will inform you about the necessary next steps (file upload etc).
Best wishes,
Thomas