Review of “JLH Mark2 – An Improved Opto-Mechanical Approach to Open-Path in situ Water Vapor Measurement in the Upper Troposphere / Lower Stratosphere,” by R.L. Herman, et al., Egusphere-2024-4019

By Darin Toohey, University of Colorado Boulder

This is an important contribution to the scientific literature, and AMT is an appropriate forum for publication. I recommend that the paper be published with minor revisions. The JPL group is uniquely positioned to get this information into the public domain. First, they pioneered this method for water vapor measurements from aircraft. Second, they showed the way forward that others have used. This paper documents a number of important features of the JPL aircraft instrument in various versions that are important for understanding their performance and accuracy. Although my comments are extensive, they are all straightforward for the authors to address, even the “most significant comments,” and I trust them to decide how best to respond without the need for further review. 

My detailed comments are in the attached pdf supplement.

Review of “JLH Mark2 – An Improved Opto-Mechanical Approach to Open-Path in situ Water Vapor Measurement in the Upper Troposphere / Lower Stratosphere,” by R.L. Herman, et al., Egusphere-2024-4019

This paper discusses the improvements to the JLH water vapor instrument and recent flight performance. Overall I found the paper well written and an important contribution to the literature for those using JLH data. The paper also details the new optical design which could be used for other instruments deploying an open path design for a flight instrument. Minor comments are detailed below, but otherwise I find this paper suitable for publication.

Minor Comments:

Line 28: addition should additional

Line 37: Be consistent with UTLS or UT/LS

Lines 83-90: Where are these temperature and pressure measurements compared to JLH. Could you comment on the suitability of these measurements for the water absorption. Do you expect there to be T and P differences between where these measurements are made and the Herriott cell?

Line 107: Can you comment on boundary layer depth of the fuselage.

Line 117: This additional pathlength seems potentially an important source of error. Do you believe this air to be flushed in the same way the air between the mirrors? Have the same temperature and pressure? I realize it’s only 0.8% of the pathlength, but if this volume is “dead air” it will have much higher mixing ratio of water.

Line 119: How is the pathlength calculated? Is this by ray trace or knowing the angle that the light bouncing between the mirrrors?

Line 166: 8 Hz is a little ambiguous here. You mean that you have 8 ramps a second, but the actual rate of data acquisition is much higher. Probably best to have another sentence that specifies the acquisition rate and ramp rate and remove the 8 Hz at the end of the sentence.

Line 168: What is DC mean here, direct current, or did you mean DT?

Line 211: Be good to mention how well this works. What was the flight stability of the laser temperature thermistor.

Line 249: should this be ‘background noise’?

Line 287: I’m assuming the interpolation is linear, but inherently a laser’s tuning rate with current does not have to be linear, so it might be worth noting whether using the methane lines you find it linear over this scan range.

Line 292: It would seem that this last sentence flows more naturally right after the discussion of determining the water line position (ending in the middle of line 287).

Lines 309 – 312: I find this procedure confusing or at least its description. Is the HITRAN database parameters being used to fit the direct absorption spectra? When you say you use the ratio, do you mean for the field data you scale the direct absorption measurements by this scale factor? So ultimately the standard that you are using to calibrate your instrument is the Thunder Scientific.

Line 312: I’d be careful of saying you are canceling the uncertainties in the HITRAN database. There may be many sources of error for fitting the direct absorption line. Line purity for one, which you mention. The choice of which line shape parameters you are using. At higher concentrations or for your stiudies using pure water self broadening can be important and also choice of lineshape (Voigt versus Galatry for example). How well you know your tuning rate and how linear it is. For the purposes of measuring water vapor in the atmosphere using methane lines to derive a tuning rate and position is perfectly adequate, but may miss some small nonlinearities in the laser.

Line 361: Not sure why this section is called Laser stability. ‘Absorption line selection’ perhaps.

Line 362: Given that everywhere else you write ‘water’, I’d change H2O to water. Unless there is some more subtle point you are trying to make that I’ve missed.

Line 466: laser tuning rate here refers to cm-1 versus time or versus current? It should be defined somewhere. You also mention it earlier, so perhaps defining it there.

Lines 466 – 476: The discussion of line widths is a little confusing. On line 472 you say “the Gaussian is a convolution of Doppler-broadened linewidth and laser linewidth, then we calculate an effective instrument linewidth …” Given the first clause of that sentence, what you are calculating is the laser linewidth. You then go on to say that the instrument line width is now a convolution of the true laser linewidth and electronic broadening which would seem to invalidate the presumption you make on line 472. I think it would be simpler to say that the difference between the Gaussian fit and the calculated Doppler broadening is dominated by the laser linewidth and any electronic broadening. You believe (I don’t think you say that you’ve measured the laser linewidth independently) the laser linewidth is smaller and therefore think that the electronics are broadening the line. Usually if the electronics are the issue they act as an RC filter of the data. That would not broaden the line symmetrically around the line center. You can also characterize the electronic time constant by chopping the laser light or cutting the current quickly and looking at the exponential decay on an oscilloscope.

Line 518: Could you explain what you mean by the limited time response of the signal chain?

Figure 10: The y axis label says ‘and Pressure’. I think that is a typo.

Section 6. Performance: In all this discussion, you never state your accuracy. You talk about the calibration procedure in in section 2.3.2, but I expected at some point a statement or graph showing water vapor in your lab sample flow as determined by the Thunder Scientific versus water vapor mixing ratio calculated by JLH as you would do in flight. A plot of the points showing the linearity over a couple order of magnitude of water. A difference plot and statement of accuracy in percent or ppmv. I think this is critical for an instrument paper. Comparison with other instruments is not sufficient.