Analyzing the turbulent planetary boundary layer by remote sensing systems: the Doppler wind lidar, aerosol elastic lidar  and microwave radiometer

de Arruda Moreira, Gregori; Guerrero-Rascado, Juan Luis; Benavent-Oltra, Jose A.; Ortiz-Amezcua, Pablo; Román, Roberto; E. Bedoya-Velásquez, Andrés; Bravo-Aranda, Juan Antonio; Olmo Reyes, Francisco Jose; Landulfo, Eduardo; Alados-Arboledas, Lucas

doi:https://doi.org/10.5194/acp-19-1263-2019

Articles | Volume 19, issue 2

https://doi.org/10.5194/acp-19-1263-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Special issue:

EARLINET aerosol profiling: contributions to atmospheric and...

https://doi.org/10.5194/acp-19-1263-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 19, issue 2

Research article

|

31 Jan 2019

Research article |

| 31 Jan 2019

Analyzing the turbulent planetary boundary layer by remote sensing systems: the Doppler wind lidar, aerosol elastic lidar and microwave radiometer

Gregori de Arruda Moreira, Juan Luis Guerrero-Rascado, Jose A. Benavent-Oltra, Pablo Ortiz-Amezcua, Roberto Román, Andrés E. Bedoya-Velásquez, Juan Antonio Bravo-Aranda, Francisco Jose Olmo Reyes, Eduardo Landulfo, and Lucas Alados-Arboledas

Download

Final revised paper (published on 31 Jan 2019)
Supplement to the final revised paper
Preprint (discussion started on 16 Apr 2018)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review', Anonymous Referee #2, 25 May 2018
- AC1: 'Answers Referee 2', Gregori Moreira, 14 Aug 2018
RC2: 'Review for ACP-2018-276', Anonymous Referee #1, 10 Jul 2018
- AC2: 'Answers Referee 1', Gregori Moreira, 14 Aug 2018

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Gregori Moreira on behalf of the Authors (14 Aug 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (16 Aug 2018) by Albert Ansmann

RR by Anonymous Referee #2 (28 Aug 2018)

RR by Anonymous Referee #1 (03 Sep 2018)

Suggestions for revision or reasons for rejection

2nd Review for

Analyzing the turbulence in the Planetary Boundary Layer by the synergic use of remote sensing systems: Doppler wind lidar and aerosol elastic lidar

by

G. Moreira et al

ACP-2018-276

General comments

This manuscript presents results from the SLOPE campaign in Granada, Spain, in which the objective was to obtain closure between remote sensing and in-situ measurements. For this manuscript, the focus is on characterising the planetary boundary layer using a Doppler lidar, multi-wavelength lidar (MULHACEN), and a profiling microwave radiometer, all operating at high temporal resolution (2 seconds). The authors investigate the use of fluctuations in aerosol number density from the elastic system (EL), vertical velocity fluctuations obtained from the Doppler lidar (DL), and potential temperature profiles retrieved from the microwave radiometer (MWR), to identify the boundary layer height (PBLH).

As stated in the first review, some of the methodology is relevant, and the influence of random error introducing extra noise in higher-order moments is explored using suitable techniques, but a major issue was that the manuscript did not have a concrete focus and conclusion, and without these, did not present anything new.

The authors now state that the focus of the paper is 'the synergetic combination of information from different remote sensing systems that are sensitive to different tracers'. This would present something new and useful to the scientific community but there is minimal and insufficient discussion presented in the manuscript in its current state. The comments outlined in the first review have not yet been adequately addressed, and so the manuscript is not yet ready for publication.

Major comments

The authors state that the focus of the paper is on the synergistic combination of different methods to determine PBLH, but there is still insufficient discussion in the main text. There is little suggestion on how the various retrieved parameters could be combined, or how the relative magnitudes of their uncertainties could influence the combination.

"The EL and DL parameters are calculated over 1-hour periods. Is this 1-hour timescale suitable during rapidly varying conditions such as during the morning growth of the boundary layer?" This question is asking whether a 1 hour timescale is suitable when, during the morning growth, a particular region may have been calm for 30 minutes, and then strongly turbulent for 30 minutes. If changing the timescales has an impact, then this is important information to include in the manuscript, e.g. how does the noise reduce the integral time scale, and is this SNR-dependent? The abstract states that there is "low influence of noise", so how can both od these statements be correct?

"What is the impact if you change the averaging period, and why was 1-hour chosen when the MWR data are averaged over 30 minutes?". The authors state that using a different timescale for MWR parameters does not interfere in the analysis, yet the focus is the synergistic combination of the various retrievals? The MWR retrieval is much smoother in time than the lidar retrievals, so what is the likely impact when combining them?

The manuscript requires a much more rigorous but short description of the processes driving turbulent mixing in the boundary layer.
The response from the authors does not satisfy this requirement, is far too long, and contains many factual errors. For instance, air temperature is not directly related to RN and Sw'. The abstract states (lines 22-24) "Furthermore, we show how some meteorological variables such as air temperature, aerosol number density, vertical wind speed, relative humidity and net radiation might influence the turbulent PBL dynamic" but the main text does not discuss how any of these parameters influence the PBL, and in any case it is not clear to me how, in most situations, the aerosol number density would influence the turbulent PBL dynamic.

Minor comments

DL analysis: The time-height plots provided in the supplementary material are not satisfactory. The upper panel displays vertical velocity, not wind speed and appears to have been drastically smoothed or averaged compared to the lower panel, which is not backscatter but potentially attenuated backscatter. From the system configuration information provided by the authors (telescope focus) it is unlikely that the 'attenuated backscatter' field has been corrected for the telescope focus, hence the request to provide the signal (SNR+1) instead. Please plot both the vertical velocity and signal (SNR+1) at the original resolution without averaging. If you plot the SNR, you can then see that all velocity data above about 2 km is likely to be noise.

"EL analysis: 'Is it safe to assume the two-way transmittance is negligible?" and "what are the typical molecular, aerosol and total extinction values for the cases shown here?" The two-way transmittance at a wavelength 532 nm is less than 0.98 by 2 km above sea level from molecular extinction alone. The AOD is also > 0.1 for 19th May 2016, and over 0.3 for 8th July 2016 (at a wavelength of 500 nm, from AERONET data). These values are not negligible. They may not be sufficient to impact the results, but whether this is the case should be discussed. Does this attenuation impact the noise characteristics and the integral time scale? The new figures 10 and 14 do not aid the interpretation and are not necessary for the manuscript.

Supplementary material, Figs. 5 and 6, do not show time-height plots, only two profiles, so it is still not possible to evaluate whether these parameters provide a reliable guide to the boundary layer development.

Doppler lidar and Elastic lidar analysis: Since you make some effort to quantify the influence of noise on the statistical parameters derived from these two systems, it would be beneficial to discuss how this impacts your interpretation, e.g include time-height plots of the correction factor or relative correction, relative importance in determining PBLH, how much temporal averaging is required to obtain good results. You state that your objective is 'to approach a synergetic combination', hence discussing how the influence of noise impacts your interpretation is vital, otherwise there is nothing new presented in this manuscript.

Case study 2: Did you try cloud-screening EL data before calculating EL parameters? The PBLH from EL would agree much better with PBLH from MWR in Figure 13, and maybe Figure 14 (it is hard to tell with the scales used). Clouds should also be visible in DL data. The authors response is "No, any cloud-screening method was not applied before calculating the EL parameters". What happens if you do attempt a simple cloud screening procedure. This is simple to apply and would presumably be used in any synergetic combination?

If the DL telescope focus is set to 800 m then what method do you use to obtain attenuated backscatter profiles from the (SNR + 1) profile? Therefore, in the supplementary material it would be more appropriate to present the time-height plots of vertical velocity and (SNR + 1), which was what was originally requested.

Line 32: How do the variables interfere in the process?

Lines 40-42: Please check and reformulate these sentences. Surface heating is still unlikely to directly impact the upper troposphere. The convective boundary layer does not reach the upper troposphere.

Lines 226-229: The methodologies are not used synergistically, even though this is the focus of the paper, and it is not shown or discussed how each variable 'influences' the turbulent PBL behaviour.

Many of the figures still have very short captions without enough information. Please include the instrument name, date and the location in the caption.
Figure 4: Is this autocovariance from DL?
Figures 5,7: Profiles from which instrument, and from which location? At what time, and on what day? What height is the surface? Please include this information in the caption
Figures 5, 7-10,12-14: The captions do not state which instrument the data comes from. Please include the instrument names and the location in the caption. Where applicable, state which data comes from which instrument.
Figures 11, 15, 16: The caption states 'elastic lidar data'. Please include the instrument name and the location in the caption.

Hide

ED: Reconsider after major revisions (10 Sep 2018) by Albert Ansmann

AR by Gregori Moreira on behalf of the Authors (29 Oct 2018) Manuscript

ED: Referee Nomination & Report Request started (04 Nov 2018) by Albert Ansmann

RR by Anonymous Referee #2 (13 Nov 2018)

RR by Anonymous Referee #1 (11 Dec 2018)

Suggestions for revision or reasons for rejection

3rd Review for

Analyzing the turbulent Planetary Boundary Layer by remote sensing systems: Doppler wind lidar, aerosol elastic lidar and microwave radiometer

by

G. Moreira et al.

ACP-2018-276

General comments

This manuscript presents results from the SLOPE campaign in Granada, Spain, in which the objective was to obtain closure between remote sensing and in-situ measurements. For this manuscript, the focus is on characterising the planetary boundary layer using a Doppler lidar, multi-wavelength lidar (MULHACEN), and a profiling microwave radiometer, all operating at high temporal resolution (2 seconds). The authors investigate the use of fluctuations in aerosol number density from the elastic system (EL), vertical velocity fluctuations obtained from the Doppler lidar (DL), and potential temperature profiles retrieved from the microwave radiometer (MWR), to identify the boundary layer height (PBLH). As stated in the first and second review, the methodology is relevant, and the influence of random error introducing extra noise in higher-order moments has merit and is explored using suitable techniques.

The manuscript has now been improved significantly, with a clearer focus, and explores the impact of applying the elastic lidar methodology at different wavelengths. New Figures 9 and 10 now show that, although backscattering coefficients are wavelength-dependent (molecular and aerosol), the methodology and correction procedure can account for these differences. It is now clear that the higher moments exhibit more correction at 532 nm but the methodology used to derive PBLH from elastic lidar is not unduly sensitive to the wavelength used, at least.

The new supplementary figures are much better, and clearly display where reliable data may be obtained. However, there are still a few issues for the authors to address before the manuscript is suitable for publication.

This question was asked previously: "The EL and DL parameters are calculated over 1-hour periods. Is this 1-hour timescale suitable during rapidly varying conditions such as during the morning growth of the boundary layer?" This question is asking whether a 1 hour timescale is suitable when, during the morning growth, a particular region may have been calm for 30 minutes, and then strongly turbulent for 30 minutes. What happens to the integral timescale for both EL and DL properties when you include atmospheric regions with very different turbulent attributes (e.g calm and convective) within the same averaging period? With one hour averaging, you will always miss the rapid growth of the CBL. I understand you want to capture the integral time scale, but how can you do this when the turbulence characteristics themselves are changing?

Minor comments

Lines 42-47. Please check and reformulate these sentences.

Line 78: Replace 'turbulenc' with 'turbulence'.

Line 86: The convective PBL is the CBL.

Line 96: Replace 'realibility' with 'reliability'. Explain why measurements at 532 nm should be more reliable, or suggest removing this sentence.

Lines 129-130: Suggest replacing 'This system record the backscattered signal with 300 gates, being the range gate length 30 m, with the first gate at 60 m' with 'This system records the backscattered signal with a range resolution of 30 m in 300 range gates with the first range gate starting at 60 m from the instrument.'

Line 131: Suggest stating 'The instrument was operated in vertical stare mode with a temporal resolution of 2 s'. Using the phrase 'with respect to the ground surface' could mean that, on a sloping surface, you imply you are pointing normal (90 degrees) to the surface. Pointing vertically is unambiguous and doesn't require the qualifier.

Line 168: Replace 'performed campaign of comparison' with 'intercomparison campaign'.

Line 173: Replace 'allow to estimate the CBL height' with 'allows the estimation of the CBL height'.

Lines 179, 182, 186: σw2, σRCS, q', are used before being defined.

Line 179: Do you mean the PBLH_Doppler is attributed to the height where σw2 drops below a pre-determined threshold?

Line 191: Replace 'isthe' with 'is the'.

Line 222: Replace 'depends' with 'depend'.

Lines 242-244: Suggest rewriting this phrase as it is unclear. You could use ' in this study we evaluate using RCS532 fluctuations to determine turbulence following the procedure described in Figure 3. This EL methodology is very similar to that described earlier for DL.'

Line 257 and elsewhere: Suggest using 'below the PBLH' rather than 'under the PBLH'.

Line 259: Replace 'relation the other' with 'relation to the other'.

Lines 263-265. This statement is only true in high SNR conditions - figure 5 only shows data within 1200 m of the instrument.

Lines 272-273: Suggest replacing 'its spread use in observation network with higher reliability than 1064 nm' with 'its widespread use in observation networks'.

Lines 274-275: Suggest replacing 'As expected, in both cases the increase of height produces the increase of ε,' with 'As expected, ε increases with range'.

Lines 278-283: The difference in noise levels between the two wavelengths depends on the SNR at each wavelength, which is more likely to be determined by the laser output power, filters, and detectors used, at the two wavelengths. Higher molecular extinction at 532 nm can then reduce SNR relative to the 1064 nm wavelength, as does separation of the molecular and aerosol backscattering. Figure 8 is not necessary for the manuscript.

Line 299: SNR is reduced, the noise doesn't increase.

Lines 303-304: Suggest rewriting this phrase as it is unclear.

Lines 315-318: It is not clear that the integral time scale can be retrieved throughout the whole PBL, especially if PBLH_MWR is taken as a reference. It is not necessarily true that the grey areas are where T_w' is lower than the DL acquisition time, just that the DL sensitivity is not high enough to measure T_w'. This can be seen in the supplementary material, where SNR is low above 1500 m and the upper portion of the CBL may not be captured during daytime. I would expect T_w' to be just as large here.

Lines 319-320: This sentence can be removed.

Lines 327-328: I suggest removing this sentence.

Lines 329-320: Not correct, skewness describes the distribution of the turbulent velocities - positive skewness implies strong but narrow updrafts surrounded by weaker but more widespread downdrafts, and vice versa for negative skewness.

Line 336: Do you mean that that the MWR method is now selecting for SBLH?

Lines 344-353: Suggest removing these paragraphs. Some of these statements could replace phrases in lines 329-337.

Lines 365-367: The PBLH values in Figure 13 don't agree with Figure 12, where there is a large difference between PBLH_MWR and PBLH_Elastic.

Line 372: Define FT here.

Line 392: Do you refer to the correct figure here?

Line 399. Suggest removing the second phrase of this sentence.

Line 400: Not true according to the figure.

Lines 411-414: No clouds are observed in the DL data until 1400 UTC, and it is difficult to prove that the negative skewness extends to 3 km in altitude. Are you sure that all of the white regions are cloud before 1400? One at 1230 and one at 1330 maybe.

Lines 415-420. As above, it is clear that the Saharan dust layer is having an impact, but Rn alone is probably not sufficient to attribute negative skewness directly to clouds.

Line 423: See above comment. Not all of the high RCS values (white regions) before 1300 can be attributed to clouds, and there is very little attenuation seen in the profile before 1230.

Line 463: This is only true at high SNR.

Line 467: Replace 'lower two ways' with 'reduced two-way'.

Figure 12: Replace 'blues stars' with 'blue stars'.

Hide

ED: Publish subject to minor revisions (review by editor) (16 Dec 2018) by Albert Ansmann

AR by Gregori Moreira on behalf of the Authors (24 Dec 2018) Author's response Manuscript

ED: Publish as is (03 Jan 2019) by Albert Ansmann

AR by Gregori Moreira on behalf of the Authors (10 Jan 2019)

Download

Article (13847 KB)
Full-text XML

Short summary

In this study we show the capabilities of combining different remote sensing systems (microwave radiometer – MWR, Doppler lidar – DL – and elastic lidar – EL) for retrieving a detailed picture of the PBL turbulent features. Concerning EL, in addition to analyzing the influence of noise, we explore the use of different wavelengths, which usually includes EL systems operated in extended networks, like EARLINET, LALINET, MPLNET or SKYNET.