|Review of the revised manuscript ‘SEALDH-II – a calibration-free transfer standard for airborne water vapor measurements: Pressure dependent absolute validation from 5 – 1200 ppmv at a metrological humidity generator’ by B. Buchholz and V. Ebert. |
The discussion as a result of the two reviews reconfirmed my impression that there is very little to comment on the technical work of this manuscript. This part is very detail oriented and shows that the authors built and properly validated a very high quality hygrometer.
However, my concerns regarding the novelty and importance have been reconfirmed. I particular, the connection with the earlier Hygrometer for Airborne Investigations (HAI) has not been adequately described. Furthermore, there apparently was an earlier version, SEALDH-I, which had the same features as SEALDH-II. In their replies the authors stated that “Maybe we didn’t elaborate clearly enough on the connection between the instruments”. That is somewhat of an understatement, since neither of the other two instruments were even mentioned. This was quite surprising, since both instruments come from the authors themselves. It is rare to see that authors ignore their own work. The authors also confirm my suspicion that these instruments are borne out of the same concept, I quote, “to ensure a permanent, highly defined, multi parameter instrument control for each individual data point”. Therefore, the authors’ claim that SEALDH-II is a novel concept is simply not correct and should not be described as such. This concept has already been implemented in HAI as atmospheric instrument and SEALDH-I as proof of concept. I fully appreciate the differences between HAI and SEALDH-II and the differing motivations for both instruments. HAI is more appropriate for atmospheric observations, while SEALDH-II was built with metrological traceability in mind. Nevertheless, these instruments are based on the same concept, measure the same atmospheric parameter using the same measurement principle, claiming to be calibration-free. Therefore, statement such as, I quote, “implementing a new holistic concept to achieve higher accuracy levels”, made in the abstract, are simply not correct. The authors may claim novelty for HAI or possibly SEALDH-I, not for SEALDH-II. Monitoring many more instrument parameters than in HAI may improve the confidence in the observations and may allow a better characterization of the uncertainty budget. Their instrument is certainly of very high quality, but the authors cannot claim a ‘new holistic’ concept, since that has already been done. Instead the authors should include an extensive discussion of the relation between these instruments in the manuscript, not just the replies to the authors.
A key part of this manuscript is the metrological validation of SEALDH-II. Given that HAI is much more suited to atmospheric measurements, the authors should elaborate, why they chose to build an entirely new instrument, instead of validating the first instrument they built for atmospheric measurements. Some of the discussion in their replies should be included into the manuscript as well as their plans for the validation of HAI.
The second major issue I have with the comments is their relation to stratospheric measurements. The title includes “airborne water vapor measurements” and “5 – 1200 ppmv”. Throughout the paper, the authors refer to a measurement range of 65 hPa – 950 hPa and that “the instrument is built for a concentration range from 3 - 40000 ppmv”. The lower end of this pressure and mixing ratio range clearly includes stratospheric conditions. The authors motivate their work through significant differences between measurements, most of which are UTLS and stratospheric measurements. They discuss the AquiVIT campaigns, which were conducted largely for UTLS instruments to understand the difficulties of measuring stratospheric water vapor.
All of this gives the impression that their instrument covers UTLS and stratospheric measurements, which in fact they do not. The change in wording in the revised manuscript is insufficient and the authors should add a discussion about what their results mean for the low end of the mixing ratio and pressure range for that instrument. After all they still claim their instrument to be an instrument for airborne measurements.
Given the experience of the authors I fully believe that they could build an instrument optimized for UTLS work as they explained in their replies. However, the manuscript is about the instrument they have, not about an instrument they could potentially build. Since the current instrument is not suitable for UTLS work, it would be much more appropriate if the authors motivated their work with the appropriate instruments, i.e. exclusively tropospheric instruments and better discuss, which part of the atmospheric water vapor range the current instrument can measure and which it cannot. In the discussion it would be appropriate to then discuss that the features of this instrument could be implemented into a new instrument dedicated to upper tropospheric and stratospheric water vapor measurements.
The authors have now also included a reference to SEALDH-I and its validation with the German national standard. Therefore, the validation exercise of SEALDH-II is not the first effort of that type.
This criticism is not against the value of their work, which I certainly recognize and appreciate. My criticism is that the authors do not put their work into the proper context. AquaVIT and the discussion of the accuracy of stratospheric and upper tropospheric measurements is not the proper context.
Please note that the line numbers refer to the final manuscript version without changes markup.
Lines 71ff: Statements such as “... those often show results which are not sufficient for validation of atmospheric models in terms of the required absolute accuracy, precision, temporal resolution, long-term stability, comparability, etc.” are in that generality overstated. However, they were certainly true in the past for some sets of stratospheric measurements.
Lines 80-81: Another major issue limiting the comparison of hygrometers in flight is the issue of how ambient air is moved to the instrument. Issues such as contamination and sampling line problems are additional issues, which have plagues in situ comparisons in the past. This needs to be discussed as this is also a limiting factor for SEALDH-II in atmospheric measurements. The authors nicely point out that contamination and line issues are handled carefully in laboratory calibrations by allowing instruments and air samples to equilibrate over long times. This is not possible in atmospheric measurements, the consequences of which need to be discussed.
Line 149: “…The most holistic approach …” This wording should be changed. Better wording is “…most extensive approach…”
Lines 152 ff.: Delete lines 152 – 154 and better write: “SEALDH-II is described in detail in Buchholz et al., (2016)”
Line 174: Here the authors should explicitly state that the uncertainty of the lower end of the mixing ratio measurement range at 3ppm is 100% due to the offset component. Here the authors need to be clear that this instrument is not suitable for lower stratospheric measurements and only somewhat useful in the upper troposphere compared to other instruments, which have a better claimed (although not validated) uncertainty.
Line 183: At least except for the two other instruments, the authors already built. This sentence should be deleted to avoid confusion.
Lines 280f.: “…over a range which is particularly interesting for instruments on airborne platforms operating from troposphere to lower stratosphere where SEALDH-II’s uncertainty (4.3% ± 3 ppmv) is suitable”. Here the authors imply explicitly that the uncertainty of SEALDH-II is suitable for lower stratospheric measurements, where in fact the numbers in that same sentence clearly say the opposite. An uncertainty of 3 ppmv is not suitable for lower stratospheric measurements.
Lines 311-320: As said above, the core instruments of AquaVIT were built for upper tropospheric and stratospheric measurements in mind. In this sentence the uncertainties at the comfortable range for SEALDH-II are again compared to the challenging regime explored during AquiVIT, where SEALDH-II has itself a very large uncertainty. This is not appropriate and these sentences should be deleted.
Lines 375-378: This has been pointed out before and should be deleted here.
Lines 389ff.: In the evaluation it is not acceptable to ignore the 65 and 125 hPa levels. The title of the paper clearly refers to airborne measurements; therefore, the validation effort needs to consider the atmospheric distribution of water vapor and not discuss how the validation result might look like without these levels.
Line 403: Delete “To prevent further interpretations”
Lines 405 ff: better “This is an essential difference between calibration free instruments and …”
Line 407: better “SEALDH-II tries to guarantee …”
The authors can never guarantee that ALL effects have been considered. They can only guarantee that all effects they know about have been considered.
Line 428 f: Delete “novel” and “holistic”
Lines 451 ff: I would hope that the authors will as well validate HAI, which has been built by them and which is much more suitable for atmospheric measurements. Given the use of that instrument, validating it would be more than useful.
Figure 6-8: I maintain my earlier comment that the abscissa should be log(P), since the authors write about an instrument for airborne use, not laboratory use. For many atmospheric measurements, in particular aircraft based measurements, it is much easier to relate to altitude which varies nearly linear with log(P) than with pressure itself. This emphasizes the entire tropospheric distribution better than linear pressure.