the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 22 Jul 2026
Overhaul of the stratospheric lidar at Dumont d’Urville, Antarctica: three-wavelength setup, processing and observations
Abstract. The stratospheric cloud and aerosol lidar instrument Liraan (LIdar pour la Recherche Antarctique Aérosol et Nuages) has been operated at the French Antarctic station Dumont d'Urville since 1989. Following a major upgrade conducted between 2022 and 2024, this paper presents the instrument configuration and its scientific capabilities. Liraan now enables the retrieval of backscatter coefficients at 355 nm, 532 nm and 1064 nm and depolarization ratios at 355 nm and 532 nm. The new configuration is described in detail, along with an improved data inversion algorithm, which is further used as input to a dedicated microphysical retrieval methodology. A unified 532 nm lidar time series spanning 1991–2025 at Dumont d'Urville is also presented for the first time. Finally, 2025 multispectral lidar measurements are used in a case study of volcanic aerosol size distribution retrieval.
- Preprint
(5880 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-2958', Anonymous Referee #1, 19 Aug 2026
-
AC1: 'Reply on RC1', Florent Tencé, 18 Sep 2026
The authors would like to thank the reviewer for their comments and their appreciation of our study. Below, we answer to the different comments and questions sent by the reviewer.
1. Major point
We agree with the reviewer that it is valuable to show a PSC detection in this paper, and a PSC measurement profile acquired at DDU on July 9th, 2025 has been added to the manuscript as Figure 4.
We had initially decided to focus on aerosol measurements as we sounded volcanic aerosols plumes since the setup change in DDU, and this was fitting with the aerosol size distribution retrieval method presented in this paper. We detected few PSC observations with the new lidar setup, but we considered having not enough background on multi-wavelength PSC observation to exploit it adequately.
Building a PSC classification based on multi-wavelength optical parameters retrieved from DDU lidar measurements is one of the objective of this new system. However, we believe it is too early to use such measurements to help distinguish PSC types, as not many PSC observations at 355 – 532 – 1064 nm have been made yet in DDU. To our knowledge, the only 355 – 532 – 1064 nm system able to detect PSC is located in ALOMAR in Norway, and has not been used yet to produce a PSC classification built on multispectral signatures.
Reichardt et al., 2004 presented a PSC case study with measurements at 355 and 532 nm in Esrange, Sweden, but this study focused on a single PSC detection. The authors did suggest that the addition of 1064 nm would be useful for the detection of large NAT particle and ice crystals.
Other studies such as Dörnbrack et al., 2002, Hu et al., 2002 have sounded PSC with multi-wavelengths systems, but were based on airborne campaigns in the Arctic and did not aim at producing a multi-wavelength PSC classification scheme. We believe that it also highlights the relevance of the new DDU lidar configuration presented in this study : it will provide a long-term multi-wavelength PSC observation record. Such a record is still very rare, if not unique in Antarctica.
Figure 4 shows a PSC measurement obtained at DDU on the 9th of July, 2025. The UV, visible and infrared channels exhibit a consistent dynamic over the stratospheric range, and the layer extending from 19 to 21.3 km display scattering properties of a PSC. At 20 km, the scattering ratio peak at 1.4, 1.86 and 6.2 at 355 nm, 532 nm and 1064 nm, respectively. The depolarization ratios peak at 3.6 % at 355 nm and 13.3 % at 532 nm.
Figure 4 is a clear example of a nitric acid trihydrate (NAT) PSC detection: the non-zero depolarization ratio values indicate that the observed cloud cannot be a supercooled ternary solution (STS), which would only contain liquid droplets. The temperature at 20 km is approximately 5 K above the TICE threshold. An ice PSC would also be expected to display significantly higher scattering ratio values (Achtert et al, 2014, Pitts et al., 2018). TNAT, TSTS and TICE formation threshold temperatures have been calculated with H2O concentration of 4.3 ppm and HNO3 concentration of 8.6 ppb taken from MLS measurements above DDU on the 9th of July, 2025 (Tencé et al., 2023).
This example did not need the multi-wavelength capacity of the lidar to determine the detected PSC type. However, the 355 nm and 1064 nm channels could help resolve less obvious PSC observations, as Reichard et al (2004) suggest.
2) It is true that we did not make explicit how the transmission factor used to correct for ozone absorption was applied to the lidar signal. The lidar signal Pλ has to be multiplied by the transmission factor CO3 to take in account the ozone absorption. A sentence was added at the end of section 3.1 to make that explicit.
We are however not sure to understand the reviewer question « Further, for what do you use the total lidar signal? ». We assume the reviewer asks why we combine the parallel and perpendicular channels to form the total lidar signal. The total lidar signal is the main input of the inversion procedure described in section 3.2. In case this was not explicit, another sentence was added at the end of section 3.1 to explain this point.
3) The altitudes that we call znc are all the potential candidates for zref ; i.e. altitudes between 20 and 30 km that are not included in a detected layer. Each of these altitudes are tested as zref : for each of the resulting backscatter ratio profiles of these inversions, we compute the criteria we call « Deviation of Rλ from unity ». The znc value which minimizes this criteria is kept as our optimized zref value.
This is what we explained in the lines just above the ones quoted by the reviewer : we wrote that « each lidar profile is evaluated over a wide range of lidar ratios LR and reference altitudes zref. The resulting backscatter ratio profiles are then compared using two independent criteria presented hereafter, and the optimal set of (LR, zref) values is retained ».
4) The altitude range in Figure 3 is indeed too small to be read : it spans from 8 to 28 km. We now included it both in the Figure caption and in the text.
5) We agree with the reviewer that after the long transport experienced by the aerosol plume, it is expected to be aged and spatially dispersed when observed above DDU. The « variability » observed on Figure 4 (now Figure 5) is misleading and is not in contradiction with the previous statement.
There are two reasons to this : one being intrumental and one physical. First, as specified in the Figure caption « The time axis is not linear and depends on the number of available measurements. Each individual profile is integrated over 15 min. » This choice was made to be able to observe the fine dynamics of aerosol layers, which would have not been possible with a daily-averaged time series for example. This explains some discontinuities observed in the time series. For example, the reviewer mentions the dates around the 1st of April. During that period the lidar could be operated on the 26th, 27th and 28th of March, then on the 4th and 6th of April, and the following one was only the 24th of April. The white lines – and their varying width – indicate these time discontinuities, which are due to local conditions such as storm or thick tropospheric cloud cover which often prevent stratospheric lidar measurements.
The second reason is due to stratospheric dynamics. The lidar, by construction, observe air masses as a function of altitude, where as these air masses are not transported along constant altitude levels but rather along constant potential temperature lines. Additionnally, even though it is minor in comparison to horizontal transport, stratospheric air masses still experience some degree of vertical transport.
Finally, even if the main part of the stratospheric plume was located 12 and 18 km, some measurements, for example in January and February or around the 31st of May, show clear aerosol detection as high as 20 km.
We believe these reasons, combined, explain the variability observed in the aerosol layer in the first half of 2025.
References :
Achtert,P. And Tesche, M.: Assessing lidar-based classification schemes for polar stratospheric clouds based on 16 years of measurements at Esrange, Sweden, Journal of Geophysical Research: Atmospheres, 119, 1386–1405, https://doi.org/https://doi.org/10.1002/2013JD020355, 2014
Dörnbrack, A., T. Birner, A. Fix, H. Flentje, A. Meister, H. Schmid, E. V. Browell, and M. J. Mahoney, Evidence for inertia gravity waves forming polar stratospheric clouds over Scandinavia, J. Geophys. Res., 107(D20), 8287, doi:10.1029/2001JD000452, 2002.
Hu, R.-M., K. S. Carslaw, C. Hostetler, L. R. Poole, B. Luo, T. Peter, S. Füeglistaler, T. J. McGee, and J. F. Burris, Microphysical properties of wave polar stratospheric clouds retrieved from lidar measurements during SOLVE/THESEO 2000, J. Geophys. Res., 107(D20), 8294, doi:10.1029/2001JD001125, 2002.
Pitts, M. C., Poole, L. R., and Gonzalez, R.: Polar stratospheric cloud climatology based on CALIPSO spaceborne lidar measurements from2006 to 2017, Atmospheric Chemistry and Physics, 18, 10 881–10 913, https://doi.org/10.5194/acp-18-10881-2018, 2018
Reichardt, J., Dörnbrack, A., Reichardt, S., Yang, P., and McGee, T. J.: Mountain wave PSC dynamics and microphysics from ground-based lidar measurements and meteorological modeling, Atmospheric Chemistry and Physics, 4, 1149–1165, https://doi.org/10.5194/acp-4-1149-2004, 2004.
Tencé, F., Jumelet, J., Bouillon, M., Cugnet, D., Bekki, S., Safieddine, S., Keckhut, P., and Sarkissian, A.: 14 years of lidar measurements of polar stratospheric clouds at the French Antarctic station Dumont d’Urville, Atmospheric Chemistry and Physics, 23, 431–451, https://doi.org/10.5194/acp-23-431-2023, 2023.
Citation: https://doi.org/10.5194/egusphere-2026-2958-AC1 - AC3: 'Reply on AC1', Florent Tencé, 18 Sep 2026
-
AC1: 'Reply on RC1', Florent Tencé, 18 Sep 2026
-
RC2: 'Comment on egusphere-2026-2958', Anonymous Referee #2, 24 Aug 2026
General comments
This preprint presents the current technical configuration and data processing procedures of the stratospheric lidar operated at Dumont d’Urville (Antarctica). This instrument has been operating since 1989 in a particularly remote region, under harsh meteorological conditions and with major logistical constraints. Long-term observations of this type are of great importance for the scientific community, in particular for monitoring the Antarctic stratosphere and documenting the occurrence and evolution of stratospheric aerosol and polar stratospheric cloud events.
In this context, I consider it very valuable that the current technical configuration of the instrument, as well as the associated data processing and inversion procedures, are comprehensively documented and made available to the community. Such documentation is particularly important for long-term observational infrastructures, where successive technical developments and changes in instrumentation need to be properly documented to ensure the long-term usability and interpretation of the data.
Overall, I find the manuscript well motivated and generally well presented. I am therefore in favor of publication in AMT, after the comments and questions raised below are addressed in a revised version.
Specific comments
1) Line 20, the sentence “Similar systems have since been installed in McMurdo, Rothera and Davis stations (Adriani et al., 1992, 2004; Innis and Klekociuk, 2006; Simpson et al., 2005), with some instruments relocated between stations to maintain operations (Snels et al., 2019, 2021; Cairo et al., 2023).” is not entirely clear to me. It is not clear to me what is meant by “some instruments relocated between stations”?
2) The introduction provides a good description of the scientific motivation and historical context of the Dumont d’Urville lidar, and link with satellite missions (CALIPSO/EARTHCARE). However, I think that the organisational and networking context could be described in greater detail.
The Data availability section (line 425) indicates that the instrument operates within the framework of NDACC. I suggest that this should also be introduced earlier in the Introduction. In particular, it would be useful to explain what operating within NDACC implies for the observations presented in this paper, for example in terms of quality assurance and quality control, data processing procedures, calibration, and/or harmonisation of the products with observations from other NDACC lidar stations.
This information would also help the reader to better understand the role of the Dumont d’Urville lidar within the broader international network and the extent to which the processing described in the manuscript follows established NDACC recommendations or procedures specific to this instrument.
3) The data-processing section is clearly presented and provides useful information on the inversion procedures and uncertainty propagation. However, I would appreciate some additional information concerning the status and origin of the inversion and error-propagation codes.
Are these codes research codes specifically developed by the authors for the Dumont d’Urville lidar? Or are they community-developed codes that are also used for other similar or related lidar instruments?
If the codes are based on existing community tools, it would be useful to provide the relevant references and to explain which parts have been specifically adapted or developed for the Dumont d’Urville system. Conversely, if the codes have been developed specifically for this instrument, could the authors indicate whether they are available to the scientific community, for example through an open-source repository?
Given the long-term importance of the dataset and the increasing emphasis on reproducibility of measurement processing, information on the accessibility and reusability of these codes would significantly strengthen the manuscript.
Technical corrections
1) Figure 1
The small text and labels in Figure 1 are readable in the digital version when zooming to 300%, but they are too small to be comfortably read in a printed version.
I suggest either increasing the font size or, preferably, removing the detailed textual information from the schematic and reporting these details in one or more tables. This would make the figure easier to read and would improve its usefulness in both electronic and printed versions.
2) Figure 3
The positioning of the years on the x-axis appears irregular. For example, the distance between 2004 and 2005 is smaller than that between 2005 and 2006. This makes it difficult to accurately identify the dates corresponding to the beginning and end of the episodes indicated by the white brackets.
This can be problematic when the figure is used to relate the observed lidar signatures to known events. For example, the Calbuco eruption occurred in April 2015, whereas the current positioning of the time series gives the impression that the corresponding signatures may start as early as autumn 2014.Citation: https://doi.org/10.5194/egusphere-2026-2958-RC2 -
AC2: 'Reply on RC2', Florent Tencé, 18 Sep 2026
The authors would like to thank the reviewer for their comments and their appreciation of our study. We reply below to the different comments and questions raised by the reviewer.
Specific comments
1) We were referring to the fact that the lidar system initally installed in McMurdo was relocated in 2014 to Dome C, Concordia station, where it still running today. The initial sentence was misleading and has been rephrased to be more explicit.
2 and 3) We address comments 2) and 3) together as they are both about the NDACC, and the position of DDU lidar system in this network.
DDU lidar is indeed, as initially stated in the acknowledgements section, part of the NDACC network, and consequently part of the French NDACC-Fr branch. We agree that it should have been made more explicit in the introduction of the article. It is now included in the introduction (lines 75-77).
The inversion algorithm described in this paper was indeed developped by the authors. We believed it was explicit in the way the algorithm is throughly described in the study, as noted by the reviewer. In the revised version, it was made more explicit that the code was developped by the authors (line 149).
DDU lidar is part of the NDACC / LWG and has been since the creation of this network. The validations conditions and quality control protocols of the NDACC / LWG are described here : https://ndacc.larc.nasa.gov/about/protocols/appendix-iii-lidar
As stated by the protocols and mentioned by the reviewer, the Antarctic context is very difficult and implies major logistical constraints. If one takes a closer look at the 6 ways listed in the recommandations for the validation of lidar instruments in the LWG, it is quick to realize that most of them cannot be applied to an Antarctic station such as DDU. All methods involving comparison with another NDACC instrument on-site, another neighbouring lidar, or the use of a NDACC mobile lidar, are not applicable. The closest NDACC lidar system is in Concordia, about 1100 km away, in a very different atmospheric context.
The point 2, i.e. comparison with a satellite borne instrument, is to us the most adapted method in our situation. As stated in the article, we have started this intercomparison with ATLID, launched in 2024 and operated in the UV. An article dedicated to DDU – ATLID intercomparison is in preparation. We believe ATLID will become the reference of spaceborne lidar for the decade to come, and will therefore provide a solid basis for intercomparison of lidar systems through a common reference.
Finally, a metadata document describing the instrument, data acquisition procedures, processing algorithms and method for uncertainty analysis is available on the NDACC website and the current article will be used to update these documents and data following the recent instrument rejuvenation.
As a research code still susceptible to evolve, for example to include Raman channels processing as mentionned in the paper, we are not in favour of publishing the algorithms in the online repository for now. However, we believe the current paper describe these procedures in a quite transparent and exhaustive way.
A side note to this comment on networks, we unfortunately forgot to mention that the update of DDU lidar was also partly supported by the Aerosol, Clouds and Trace Gases Research Infrastructure (ACTRIS). We forgot to acknowledge the ACTRIS support in our submitted manuscript : it was a mistake and they are now properly featured in the acknowledgements section.
Technical comments
1) It is true that the labels of Figure 1 are too small to be read in the original version. It would be difficult to increase all the text labels as the figure would become too loaded. We decided to keep the figure as it is, but include a new table below (Table 1), which features the main parameters of DDU lidar system (emission, reception, acquisition). We kept Figure 1 as such so that technical details about the optics, detectors or acquisition parameters can still be found.
2) The reviewer is right that the positioning of the years on the x-axis is irregular. This is a direct consequence of the methodology used to build Figure 3. As mentioned in the Figure caption : « The time scale is not linear and depends on the number of available measurements. Each individual profile is integrated over 15 minutes. » This choice was made to improve the readabilty of the figure and enable to observe the dynamic of aerosol layers. Changing this figure into a linear x-axis with daily-averaged lidar measurements would make it very difficult to fit a 1991 – 2025 in a single figure. The « length » of a year of Figure 3 therefore directly reflects the number of available lidar measurements.
To improve the readability of this Figure, we included vertical red dashed lines to separate different years, as stated in the caption : « Dashed red vertical lines delimit the different years ». This is why we do not understand the last comment of the reviewer about the Calbuco eruption, as the white bracket marking the beginning of the Calbuco aerosol event is clearly within the year 2015.
We had to make choices when designing a figure covering such a long period, and we believe its objective is to highlight the long-term record of DDU lidar, the evolution of measurements operations, as well as the numerous stratospheric perturbations observed in recent years. It is however not designed for the user to read precise dates nor optical properties.
Citation: https://doi.org/10.5194/egusphere-2026-2958-AC2 - AC4: 'Reply on AC2', Florent Tencé, 18 Sep 2026
-
AC2: 'Reply on RC2', Florent Tencé, 18 Sep 2026
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 223 | 66 | 28 | 317 | 31 | 41 |
- HTML: 223
- PDF: 66
- XML: 28
- Total: 317
- BibTeX: 31
- EndNote: 41
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The manuscript entitled “Overhaul of the stratospheric lidar at Dumont d’Urville, Antarctica: three-wavelength setup, processing and observations” by Tence et al. presents a substantial upgrade of the stratospheric lidar at Dumont d’Urville, including the implementation of a three-wavelength configuration and new approaches for data processing and analysis. The enhanced capabilities of the lidar are particularly relevant for future studies of stratospheric aerosols and polar stratospheric clouds (PSCs).
Overall, the manuscript is well structured, scientifically relevant, and well suited for publication in Atmospheric Measurement Techniques. However, I believe that the manuscript would benefit from a more comprehensive demonstration of the capabilities offered by the new three-wavelength configuration. I therefore recommend that the authors address the following major point:
The manuscript would benefit from additional examples demonstrating the scientific benefit of this new configuration. At present, the full potential of the three-wavelength measurements is not sufficiently illustrated.
According to the authors, the upgraded lidar is particularly well suited for observations of polar stratospheric clouds and aerosol-perturbed stratospheric conditions. In this context, I strongly encourage the authors to include an example of a PSC observation obtained with the new lidar system. Ideally, this example could also demonstrate how the available measurements can be used to distinguish or classify different PSC types. Such an example would provide a valuable demonstration of the capabilities of the upgraded system and would strengthen the relevance of the manuscript for future PSC studies.
Minor points
Section 3.1 Pre-processing steps are not clear and further explanation are needed. For example, how is the transmission factor CO3, λ later used in the lidar signal Pλ. Further, for what do you use the total lidar signal?
Line 219 to 225: It is not clear how znc is used to optimise zref.
Figure 3: Could you add within the text what altitude range is shown in the plot?
Line 347: The Ruang plume is not decreasing smoothly in Figure 4, there are “jumps” of around 4km between different observations times (e.g. before 1st of April). I am somewhat surprised by the large variability observed within the stratospheric layer attributed to volcanic plume. Considering the remote location of Dumont d’Urville and the transport time required for volcanic material to reach the Antarctic stratosphere, I would have expected the volcanic plum to be relatively well aged and spatially dispersed by the time it is observed over the station. Could the authors provide some discussion of the processes that may explain the pronounced variability observed in this layer?