Parameter Study on Ultraviolet Rayleigh–Brillouin Doppler Lidar with Dual-Pass Dual Fabry–Perot Interferometer for Accurately Measuring Near-Surface to Lower Stratospheric Wind Field

Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe author describe a theoretical study about a dual edge direct detection wind lidar system. The author should address following questions and suggestions:
1, Line 13-14 "To suppress the influence of aerosols and achieve high-accuracy measurement of the wind field throughout the troposphere to the lower stratosphere, an ultraviolet 355nm Rayleigh-Brillouin..."
--> Usually the Rayleigh Scattering is used not to suppress the influence of aerosols but not to rely on the existence of aerosol scattering. Coherent wind LIDAR using Mie scattering of aerosols has no significant thermal broadening of the scattered laser light and therefore a higher resolution in terms of frequency measurement - but it depends on the existence of aerosols. Rayleigh-Brillouin scattering, a molecular scattering process will be subject to thermal broadening an result in GHZ spectra... This is exactly what the authors write in lines 44-47. In the following lines (47ff.) it becomes more clear what the author wants to express in the abstract.
I suggest to clarify the abstract. In the introduction the authors classifies the direct detection systems in "edge technique" and "fringe imaging technique" systems. So he can use this classification in the abstract and the title.
e.g. "Parameter study on ultraviolet Rayleigh-Brillouin Doppler Lidar with an double edge detection realized by a dual-pass dual Fabry-Perot interferometer for accurately measuring near-surface to lower stratospheric wind field"
"To suppress the influence of aerosol scattering on the double edge detection technique and achieve ..."
2. Line 143-145: "FPI-L is used to measure and lock the frequency of emitted laser. Before each period of radial wind speed measurement, the emitted laser frequency is locked at the right half waist of FPI-L."
This is generally a good idea, but it should be described in more detail. Usually commercial injection seeded lasers use build up time minimizing and cavity dithering for the cavity length control https://doi.org/10.1364/AO.24.000940. Therefore, the laser itself differs in frequency to the injection seeder slightly and cavity dithering might result in frequency deviations of +-1/10 th of the lasers FSR (e.g. +-20MHz or 60MHz for the tripled laser). So controlling the seed laser with this technique might work - but to achieve frequency stability of the injection seeded laser also the cavity has to be stabilized with a method yielding the needed frequency stability. (Ramp and Fire stabilization scheme might be more accurate for this purpose https://doi.org/10.1364/AO.40.003046 )
3. System structure design:
Please describe the PMT channels. Line 132 states "PMT-1 and PMT-2 are always in analog mode, ..." what is meant by "analog mode" - photo-multipliers tend to be very analog electronics :). You probably mean linear mode or photon-counting, or Geiger mode (saturated mode) operation of the PMTs. E.g. PMT-1 is the intensity channel used to normalize the signal of PMT-2 to recover the transmission of the etalon....
What is the use of PMT-3? This will give a combined rayleigh-mie signal (back scattered power). How is the signal used in the signal recovery?
4. Line 198: "a_i (i=1, 2) is the system calibration constant" shoud be described. Citation [18] the ALADIN (ADM-AEolus) Receiver e.g. has a single pass double edge fabry-perot detector and a separate mie receiver, which does not loose the photons, which are not transmitted by the filter "A" Etalon. These are reflected to the Channel "B" Etalon. The BS4 will reduce the efficiency by 50% (my guess) the Beam Splitters are not described in the model and text. Please describe the effect of the splitting ration influence on the sensitivity of the instrument. What is the optimal splitting ratio for BS-3 and BS-4? BS-4 I can guess 50%. BS-3 <xx% maybe 1-5% ??
5. Line 220: "First, the wind speed measurement error caused by random noise is considered..." : What are the assumption t the "random noise" usually Photodetectors have regarding their operational mode thermal noise, excess noise, shot noise (which is kind of random :) ). The authors should describe the assumed noise model.
6. Figure 6. Why do you have for the performance simulations Transmission in (a.u.)? You have an input power and a transceiver efficiency (including the fiber)? You should have assumption regarding the beam splitters...
So 80% transmission should not be achieved by the system because BS3 splits the received photon between the channels. --> In Line 311 you do make assumptions about the beam splitters!
Overall it is a nice work.
Author Response
The author describe a theoretical study about a dual edge direct detection wind lidar system. The author should address following questions and suggestions:
Response: Thanks a lot for your carefully review and giving a lot of high quality and professional suggestion. The manuscript has been greatly improved with your comments.
- Line 13-14 "To suppress the influence of aerosols and achieve high-accuracy measurement of the wind field throughout the troposphere to the lower stratosphere, an ultraviolet 355nm Rayleigh-Brillouin..."Usually the Rayleigh Scattering is used not to suppress the influence of aerosols but not to rely on the existence of aerosol scattering. Coherent wind LIDAR using Mie scattering of aerosols has no significant thermal broadening of the scattered laser light and therefore a higher resolution in terms of frequency measurement - but it depends on the existence of aerosols. Rayleigh-Brillouin scattering, a molecular scattering process will be subject to thermal broadening an result in GHZ spectra... This is exactly what the authors write in lines 44-47. In the following lines (47ff.) it becomes more clear what the author wants to express in the abstract.I suggest to clarify the abstract. In the introduction the authors classifies the direct detection systems in "edge technique" and "fringe imaging technique" systems. So he can use this classification in the abstract and the title.e.g. "Parameter study on ultraviolet Rayleigh-Brillouin Doppler Lidar with an double edge detection realized by a dual-pass dual Fabry-Perot interferometer for accurately measuring near-surface to lower stratospheric wind field""To suppress the influence of aerosol scattering on the double edge detection technique and achieve ..."
Response: Thanks very much for your reminding. Wind lidar systems can be divided into direct detection systems and coherent detection systems. Coherent detection systems are only using in aerosol regions, while direct detection systems can utilize atmospheric molecular scattering to conduct wind field measurements. Therefore, when utilizing molecular scattering signals in direct detection techniques, it is necessary to take the influence of aerosols into account. So, "To suppress the influence of aerosol scattering on the double edge detection technique and achieve ...", the addition of "scattering on the double edge detection" express the research problem of this manuscript more rigorously.
The presence of "dual Fabry-Perot interferometer" in the title indicates that the double edge detection technique is adopted in this manuscript, and the "fringe imaging technique" mostly employs a Fabry-Perot etalon. In addition, it would be more rigorous to add "double edge detection" to the title, but then the title would appear too long.
Change: Line 13-14,‘To suppress the influence of aerosols scattering on the double edge detection technique and achieve high-accuracy measurement of the wind field throughout the troposphere to the lower stratosphere,’
- Line 143-145: "FPI-L is used to measure and lock the frequency of emitted laser. Before each period of radial wind speed measurement, the emitted laser frequency is locked at the right half waist of FPI-L."This is generally a good idea, but it should be described in more detail. Usually commercial injection seeded lasers use build up time minimizing and cavity dithering for the cavity length control https://doi.org/10.1364/AO.24.000940. Therefore, the laser itself differs in frequency to the injection seeder slightly and cavity dithering might result in frequency deviations of +-1/10 th of the lasers FSR (e.g. +-20MHz or 60MHz for the tripled laser). So controlling the seed laser with this technique might work - but to achieve frequency stability of the injection seeded laser also the cavity has to be stabilized with a method yielding the needed frequency stability. (Ramp and Fire stabilization scheme might be more accurate for this purpose https://doi.org/10.1364/AO.40.003046 )
Response: Thank you very much for this professional suggestion. The frequency stability of the laser should be considered in the design. The influence of laser frequency fluctuations can be reduced by reducing the fluctuations in the operating temperature of the laser , keeping the working temperature stable and measuring the wavelength of the emitted laser with a wavemeter.
Change: Line 149-151,‘In addition, the influence of laser frequency fluctuations can be reduced by keeping the laser working temperature stable or measuring the wavelength of the emitted laser with a wavemeter.’ was added to the manuscript.
- System structure design: Please describe the PMT channels. Line 132 states "PMT-1 and PMT-2 are always in analog mode, ." what is meant by "analog mode" - photo-multipliers tend to be very analog electronics :). You probably mean linear mode or photon-counting, or Geiger mode (saturated mode) operation of the PMTs. E.g. PMT-1 is the intensity channel used to normalize the signal of PMT-2 to recover the transmission of the etalon.... What is the use of PMT-3? This will give a combined rayleigh-mie signal (back scattered power). How is the signal used in the signal recovery?
Response: Thank you very much for this professional suggestion. The Photomultiplier Tube (PMT) is a photoelectric device that utilizes the external photoelectric effect of light and is mainly composed of a photocathode and dynodes. The Photomultiplier Tube (PMT) has two working modes: single - photon counting mode and analog mode. The photon counting mode has the characteristics of high sensitivity, high signal-to-noise ratio and good stability, and is mainly used for detecting weak signals. The analog mode, the pulse signal is amplified through an analog circuit and then converted into a digital signal. The advantage is that it is suitable for the measurement of high- intensity signals and has a relatively low cost.
The signal detected by PMT-3 is used to calculate the signal intensity of the atmospheric echo signal before it enters the dual- FPI
- Line 198: "a_i (i=1, 2) is the system calibration constant" shoud be described. Citation [18] the ALADIN (ADM-AEolus) Receiver e.g. has a single pass double edge fabry-perot detector and a separate mie receiver, which does not loose the photons, which are not transmitted by the filter "A" Etalon. These are reflected to the Channel "B" Etalon. The BS4 will reduce the efficiency by 50% (my guess) the Beam Splitters are not described in the model and text. Please describe the effect of the splitting ration influence on the sensitivity of the instrument. What is the optimal splitting ratio for BS-3 and BS-4? BS-4 I can guess 50%. BS-3 <xx% maybe 1-5% ??
Response: Thank you very much for this professional suggestion. The T/R (Transmission/Reflection) of BS-2, BS-3, and BS-4 optimal splitting ratio are supplemented as 30/70, 5/95, and 50/50, respectively.
Change: Line 127-128,’ The T/R (Transmission/Reflection) optimal splitting ratio of BS-2, BS-3, and BS-4 are 30/70, 5/95, and 50/50, respectively.’ was added to the manuscript.
- Line 220: "First, the wind speed measurement error caused by random noise is considered..." : What are the assumption t the "random noise" usually Photodetectors have regarding their operational mode thermal noise, excess noise, shot noise (which is kind of random :) ). The authors should describe the assumed noise model.
Response: Thank you very much for this professional suggestion. The "random noise" here refers to the signal "shot noise".
Change: Line225,’ Random noise’change to’ shot noise of the signal’
- Figure 6. Why do you have for the performance simulations Transmission in (a.u.)? You have an input power and a transceiver efficiency (including the fiber)? You should have assumption regarding the beam splitters...So 80% transmission should not be achieved by the system because BS3 splits the received photon between the channels. --> In Line 311 you do make assumptions about the beam splitters!
Response: Thank you very much for this professional suggestion. The Transmission of the dual-FPI can be be calculated using the following formula. The coefficients such as a1, a2, a3, etc. need to be taken into account when calculating the actual transmittance. The a1, a2, a3 are related to the TR splitting ratios of BS-2, BS-3, and BS- 4.
G1a(v)=a3N1/a1N3 G2a(v)=a3N2/a2N3
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsSee the attachment.
Comments for author File: Comments.pdf
Author Response
Comment 1 : This work is an innovative work to simulate the better inhibition of aerosols of ultra-violet RayleighBrillouin Doppler wind lidar measurement by using a dual-pass dual-FPI method. The influence of aerosols has been well suppressed judging from the simulation results. Nevertheless, there are several issues that require clarification in this work. It is recommended to publish after dealing with the following minor issues:
Response1: Thanks a lot for your carefully review and giving a lot of high quality and professional suggestion. The manuscript has been greatly improved with your comments.
Comments 2: Figure. 1
The two branch ports of FC-2 output the laser, one branch port of which is collimated by Col.-1 as
indicated by a black solid line. However, the role of the red dashed line is not described.
Response2: Thank you very much for this professional suggestion. the role of the red dashed line is described in the following text, ‘During the transmittance calibration, the multimode fiber is replaced with the other branch port of FC-2 to connect to the input port of Col-2, which uses the SMA interface for easy insertion and removal.’
Comments 3: L 133
In this work, PMT1 and PMT2 are always in analog mode, while PMT3, PMT4, and PMT5 can switch to photon counting mode. How to suppress system errors when both analog mode and photon counting mode PMTs are working simultaneously?
Response3: Thank you very much for this professional suggestion. The detectors PMT1~5 will be calibrated before the system starts to work. In addition, both PMT1 and PMT2 work in the analog mode and are used to calculate the transmittance of the laser locking channel., The PMT3 - PMT5 working in the photon counting mode has no impact on locking channel.
Comments 4: L 204
QV G1m(v , y) / G1m( v, y)→ G1m(v , y) / G2m( v, y) ?
Response4: Thank you very much for this professional suggestion. It was an oversight by the author. The Reviewer is right.
Change: Line 208’ QV≈G1m(n, y)/G1m(n, y)’ change to ‘QV≈G1m(n, y)/G2m(n, y)’
Comments 5: Table. 1
In this work, a 355 nm seed laser is used for the laser system, but the divergence angle of the laser
beam is not provided.
Response:5 Thank you very much for this professional suggestion. Powerlite DLS 9050 has an output divergence angle of 0.5 mrad. Full angle for 86% (1/e²).
Change: Line325 ‘Divergence angle 0.5 mrad’ was added to the table.
Comments 6: How do the authors ensure the reliability of error analysis in a purely simulated process without actual measurement data. After completing the simulation error analysis of the dual-pass dual FPI, have the authors considered comparing the analysis results with the error of the actual measured single-pass dual FPI? Could the author also consider comparing the error analysis of the dual-pass dual FPI and the single-pass dual FPI on simulation?
Response6: Thank you very much for this professional suggestion. Since it would cost a lot to build such a system, negotiations are currently underway with commercial companies. Verification can be carried out after the system is established in the future. So, we will consider comparing the error analysis of the dual-pass dual FPI and the single-pass dual FPI on simulation. The comparison results of the dual-pass dual FPI and the single-pass dual FPI on simulation are presented in Figure10.
Figure10. Profile of radial wind speed measurement bias varies with altitude
Change: Line301-302 ‘The Figure10(a) is radial wind speed measurement bias with proposed technique’ and line 307-316 ‘Figure 10 (b) is radial wind speed measurement bias with the traditional single-pass dual FPI double-edge technique, using FPI parameters described in Reference [20]. As shown in the figure, at the altitude of 0-20km within the radial wind speed dynamic range of ±50m/s, the radial wind speed measurement bias exceeds 1.7m/s in the cloudless region, and there is a significant deviation throughout the entire range; within the radial wind speed dynamic range of ±30m/s, the biases exceed 2.4m/s and 1.1m/s in the regions of light cumulus cloud and cirrus cloud, respectively. This indicates that the single-pass FPI technique has a weak ability to suppress aerosols, and there will be large deviations in wind field measurements when dealing with the lower atmosphere where the aerosol concentration is relatively high or when encountering clouds.’ were added to the manuscript.
Author Response File: Author Response.pdf