Interactive comment on “ Open-path , quantum cascade laser-based sensor for high resolution atmospheric ammonia measurements ”

Response: The current laptop used in this system consumes ∼50 W of electrical power. The laptop computer provides LABVIEW-based data processing (lock-in detection), final retrieved mole fractions and diagnostics data recording. The same laptop has been used to control the NH3 sensor and another open-path, QCL-based sensor [Tao et al., 2012b] simultaneously. However, we are not limited to a standard laptop computer to perform these tasks. The algorithms and data recording can be accomplished with lower power computers (e.g. tablets or single board computers with single Watts power consumption) or compact embedded systems, which include drive electronics, data acquisition and processing with very low (single Watts) power consumption. These improvements make reductions in total system power quite feasible in future configurations. We add the following to the revised manuscript (section 2.1): “The laptop computer power consumption (currently ∼50 W) can be reduced further with alternative, lower power single-board computers or embedded systems.”


Geoscientific Instrumentation
Methods and Data Systems Responses to D. Griffith, Editor: Comment: I agree with the two referees that the paper is well written, clear, and suitable for publication in AMT after minor revisions (including mine below). Please go ahead and address the referees' comments and submit a response to these comments and revised manuscript with changes highlighted.
Response: We are grateful for the constructive comments of the Editor. We have improved the existing manuscript and submitted a revised manuscript based on the C3122 comments of the Editor and two anonymous referees. Below are detailed responses and improvements to originally submitted manuscript in response to all comments and suggestions.
Comment: I would like to add a couple of editor comments to those of the two referees: -Please be consistent in the correct use of "mixing ratio" (more correctly "mole fraction" ) and "concentration". A quantity measured in ppb is the former, the latter is either a generic term in the appropriate context, or specifically an amount per unit volume.
Response: Ammonia is measured in ppbv as mole fraction (uncorrected for water vapor dilution effects). We have corrected the revised manuscript accordingly. As noted in the revised manuscript (section 4), all field measurements presented are corrected for water vapor dilution and expressed as mixing ratios in dry air.
Throughout the typeset manuscript, "double f" appears as a script ff, presumably a typesetting error by the journal (it is correct in the submitted manuscript. I will bring it to their attention and suggest the authors do as well.
Response: We will bring this typesetting error to the attention of the journal.
Comment: In the introduction pages 7007-7008, the authors completely neglect to mention open path measurements of NH3 by FTIR spectroscopy, which has been around for some time in biomass burning and agricultural measurements.  Wooster, et al. (2011). "Absolute accuracy evaluation and sensitivity analysis of OP-FTIR NLS retrievals of CO2, CH4 and CO over concentrations ranging from those of ambient atmospheres to highly polluted plumes." Atmospheric Measurement Techniques 4: 97-116. Griffith, D. W. T. and B. Galle (2000). "Flux measurements of NH3, N2O and CO2 using dual beam FTIR spectroscopy and the flux-gradient technique." Atmospheric Environment 34 (7) We note that Smith et al. (2011) have not reported accuracies for NH3 measurements. As explained in our introduction and references therein, the surface adsorption and partitioning artifacts associated with NH3 calibrations present unique challenges (compared with other trace gases such as CO2, CH4 and CO) that limit accuracy, especially for low mole fractions. It is only calibration if one or the other result is proven and accepted to be an accurate measure of the true mole fraction of NH3 in the sampled air (I add that it should be also then plotted on the X axis as the independent variable). Calibration is the relationship between the sensor response and an independent value of the mole fraction measured by a reference method. In this case the reference method hinges on the 25ppm stock mixture of NH3 in N2, and the quantitative dilution with N2. The sensor response (either direct or 2f) should be plotted and regressed against the mole fractions calculated from this dilution series to derive the calibration equation. The errors in the reference values (starting mixture and propagated dilution accuracy) should be quantified.
Response: We agree with the stated definition of calibration. Our calibration is based on a spectroscopic reference method, which is independent of our WMS retrieval. We note that many gas standard mixing ratios are validated using spectroscopic methods. Direct absorption is an accepted method to measure trace gas concentrations as long as the uncertainties, including absorption lineshapes (HITRAN parameters) and path length, are quantified and acknowledged. The standard NH3 mole fractions for calibration of our WMS retrieval are calculated based on a direct absorption fitting using these calibrated HITRAN parameters. In previous work, we have characterized uncertainties in the HITRAN parameters for the NH3 absorption lines utilized [Sun et al., 2013]. Owen et al. (2013) also studied these NH3 absorption transitions and reported an uncertainty of <10%. Our calibration method has an uncertainty of ± 10%, constrained by spectroscopic calibration uncertainties for the relevant HITRAN parameters. Details are explained in Sun et al. (2013). We add a new section 2.4 to the revised manuscript to describe the calibration method.
The uncertainties associated with the standard dilutions method for NH3 calibration are much more challenging to constrain due to surface adsorption and partitioning artifacts in the dilution system and calibration chamber as well as residual NH3 in the dry air used for dilution. These uncertainties are likely much larger than our method using known uncertainties in HITRAN parameters. The uncertainty increases at lower mole fractions, when dilution of typical NH3 standards by a factor of 1000 or larger is necessary. We acknowledge that for low mole fractions, the direct absorption reference method also presents larger uncertainties. However, we discuss zero calibration and associated uncertainties of the 2f retrieval in our response to Anonymous Referee #1 below and in the revised manuscript in section 3.1.
Finally, we modify Fig. 4 so that the reference mole fractions, based on direct absorption, are plotted on the horizontal axis.
Comment: Why was N2 used as the diluent, rather than air, when it was well known and recognised by the authors that this involves a significant line-width error? It sounds C3126 somewhat weak to say you didn't bother because it was within the accuracy limits of the technique (10%). In general, I find the approach to calibration is rather superficial and could have been improved with little effort.
We made an error in the original manuscript in describing our diluent for calibrations. To clarify, we used dry air (rather than dry N2 as stated originally) as the diluent, consistent with the zero NH3 background experiment described in section 3.3. In addition, the spectroscopic correction due to line-widths in dry N2 is removed, as it is not relevant for our discussion. Section 3.1 is modified as follows: "This 25 ppmv NH3 standard mixing ratio is decreased incrementally by dilution with dry air and each calibration point is recorded in both 2f and direct absorption as shown in Fig. 4. The horizontal axis is the NH3 standard mole fraction derived from direct absorption measurements. The vertical axis displays the retrieved NH3 mole fraction based on 2f spectral fitting with the online spectroscopic calibration method described in section 2.3. We note that the calibration is performed at mole fractions much less than that of the original 25 ppmv NH3 standard. Therefore, we use HITRAN parameters appropriate for NH3 in dry air." Interactive comment on Atmos. Meas. Tech. Discuss., 6, 7005, 2013.