|Review of "Comparison of ground-based and satellite measurements of water vapour vertical profiles over Ellesmere Island, Nunavut" by Weaver et al.|
The manuscript presents an assessment of ACE-FTS and ACE-MAESTRO water-vapour products in the upper troposphere - lower stratosphere with two different reference datasets: in situ measurements from radiosondes and ground-based sounding with upward looking FTIR.
In addition, the study involves other satellite humidity products, from AIRS, MIPAS, MLS, SCHIAMACHY and TES.
The assessment is limited to the Arctic station at Eureka and aims at evaluating the suitability of such products for climate purposes, considering GCOS requirements. The study holds on about 7-years worth of data.
The study is found of great relevance for the scientific community, in characterising the performances and limitations of those products in view of climate applications, process studies...
The manuscript is found very well structured and written, with clear description of the various datasets (satellite and reference) and with good presentation material for the results. QA practices regarding the reference datasets and data selection is sufficiently described or referenced. The methodology for comparing the profiles involves averaging kernels as much as possible, or proxy-kernels where these are not available.
I recommend the publication of the manuscript, with few small comments.
The study makes use of data from one location only, and hence -despite the quality of the exhaustive work- there are limitations to the generalisation of the conclusions wrt objectives: i.e. assess the relevance of the proposed ACE datasets in view of climate applications with GCOS requirements. The authors are invited to expand a little bit on this in their conclusions and provide outlook to possible future study/assessment plans seeking larger coverage in the Northern Polar cap if that is the main region of interest.
Since the point of the study, besides documenting the performances of the various satellite products, is to evaluate the performances wrt GCOS requirements, it is important to quantify the contribution of the collocation uncertainties to the overall budget (satellite - sonde/125HR). The issue is briefly touched on in the manuscript but should be discussed a bit more in my view: the conclusion of the study seems to be that ACE products are not suitable according to GCOS requirements. They might be more relevant than it appears in fact if the collocation uncertainties were considered.
The relevance of the respective satellite products is assessed with bulk departure statistics of satellite wrt reference data. The individual overall figures are then compared to the GCOS 5% requirements. Also important for climate is the ability to monitor the distribution of the moisture load and to track extreme situations. It would be informative to plot the retrieved (satellite) and reference (in situ, 125HR) WV distributions in the studied period and layers of interest, to confirm that dryest and moistest ends are correctly captured in the satellite products. In addition, a stratification of the statistics (bias/stddev...) in different WV-load bins would be instructive too in that perspective.
Page 6.Line 30-33: adding/Recalling figures illustrating MUSICA vertical sensitivity would be informative material to the reader.
P14.L8: "but not THE information..." - missing word
P14.L8: typo/syntax "We use of the width..."
P19.L15: talkig about "cloud-free" retrievals. AIRS products are generated also in cloudy scenes, using the cloud-clearing algorithm. Have such retrievals been included - cloud-clearing accounted for in the stratification/interpretation of the results?
P20.L29-30: "the smoothing operation is not enough..." Isn't it also a possible sign that the DoFs is overestimated? I.e. that the assumed observation error in the OEM are underestimated for instance?
P21.L10: see second general comment
P21.L15-16: see third general comment