the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Jul 2026
Precipitation processes in an Antarctic moist air intrusion: insights from multi-frequency radar observations over a 1100-km transect
Abstract. In Antarctica, intrusions of coastal moist and warm air onto the high plateau play an important role in the mass balance of the ice sheet, due to their significant contribution to annual snowfall accumulation. The synoptic drivers of warm intrusions are well-established, in particular for extreme cases linked to atmospheric rivers. However, a lack of suitable observations means the micro-physical cloud and precipitation processes within intrusions remain uncertain. In the context of the Atmospheric WAter Cycle over Antarctica (AWACA) project, we investigate an intrusion associated with a coastal cyclone in East Antarctica in February 2025. Lagrangian trajectory analysis confirms that air masses within the intrusion pass over a 1100-km observational transect from the coast to the plateau, allowing precipitation properties to be tracked as the intrusion moves inland. At four sites along the transect, a multi-frequency, polarimetric, spectral radar dataset is used to investigate micro-physical processes. The reflectivity and dual frequency ratios indicate a decrease in particle size as the intrusion moves inland. Near the coast, high fall-velocities and spectral signatures point to riming of snowflakes, fuelled by ascending air masses above the coastal slope and the availability of supercooled water. On the plateau, dry and cold conditions lead to smaller particles, for which variation in the radar signal appears to arise from primary ice crystal habits. The case study illustrates the potential of multi-frequency radar data from an autonomous observational transect to investigate precipitation processes between the Antarctic coast and plateau.
- Preprint
(24139 KB) - Metadata XML
-
Supplement
(1180 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-3964', Anonymous Referee #1, 10 Aug 2026
-
RC2: 'Comment on egusphere-2026-3964', Anonymous Referee #2, 11 Sep 2026
The manuscript of Corden et al., 2026 shows the change of microphysical properties of the atmosphere during a warm air intrusion event into inner Antarctica. Measurements are taken by a unique set of Operational Platform Units along a transect from Dumont d'Urville station and Concordia station. Each Operational Platform Unit is equipped with three vertically pointed radars which operate on different frequencies (among other instruments). Those stations span different altitudes and distances from the coast and therefore different environmental conditions. The observed microphysical properties of the atmosphere above each station is discussed in detail. Hydrometeor size was found to decrease towards central Antarctica, in concurrence with more primary ice habits. Riming happens close to the coast due to ascending air masses and available supercooled water.
This study shows a unique dataset of microphysical development during precipitation events in Antarctica. In my opinion, this manuscript should be accepted after a couple of major and minor improvements.
## Major comments ##
Especially the descriptions of each station's microphysical properties is hard to follow, because plots mentioned in the sections are positioned in different parts of the paper. The link between different measurements is therefore hard to understand.The introduction should be improved. Even though the abstract states, that the research gap is the missing knowledge on microphysical development during WAI-like events or AR events in Antarctica, it is not motivated in the introduction. In general, the research gap has to be pointed out stronger, especially a detailled analysis of riming and ice crystal shapes is presented in the following sections. Please state more clearly: How are those properties influencing snowfall, precipitation rates and what does this mean for Antarctica? What do you hypothesize, what does literature say about it?. E. g. use the difference between observations at DDU and PE you mention, as a start.
L. 142 - 170: The analysis of causes of ascend are interesting but I do not see how it matters for the research question, if the ascend onto the plateau is forced? - Please clarify and shorten.
L. 179: Please show both the forward trajectories from D17 and backward ones from DMC somewhere: it is the main condition for the whole analysis that air masses cross all stations or at least close to the stations?
L. 187: Please motivate, why this Matrosov-Z-R relation paper was used instead of the most recent one for Arctic snowfall using the MOSAiC dataset: https://online.ucpress.edu/elementa/article/10/1/00101/124572/High-temporal-resolution-estimates-of-Arctic
Fig. 5: As one of the most prominent features, panel a) shows a radar bright-band at low-levels on Feb 17 after 12 UTC. Please discuss possible reasons for it. Also panel e) shows an increase in MDV at lower levels at around the same time. - Are those coinciding with the Ze-increase?
Fig. 7: To confirm the presence of rimed particles determined via the Kneifel and Moisseev 2020 method, riming can be identified from peakTree moments of the subpeaks: If there are (at least) two peaks present and one is falling > 1.5 m/s faster than the slow one, the fast falling peak is caused by rimed particles. I suggest applying this peakTree-based identification of rimed particles and overlay it with the K&M 2020 derived results. The advantage of the peakTree-based approach is that it does not depend on absolute MDV values (but contrasts the difference in MDV of two peaks), so it is vertical-air motion independent.
Section 4,5,6 are very long. Please shorten by making it more concise and concentrating on the research questions.
L. 248 - 249: As this relation is dependent on the absolute values of MDV, I wonder how well it performs under the updraft region at the coastal slope? Was the convection index kappa of K&M 2020 determined and applied?
L. 332: I would disagree with that "zig-zag" assessment here. In Fig. 10a, there are peaks at around 6 km and 3.75 km, and Fig 7e at 5.75 km and 3.5 km (so just 250 m lower). I see a "zic-zac" pattern here.
Fig. 15: Looking at these differences in sounding data and ERA-5, I think it is necessary to discuss the trustworthiness of ERA-5 data for the other sites (D17 etc.) more in detail.
Fig. 16: There is strong vertical shear evident in the time-height-figures of reflectivity and LDR. Spectrograms should be presented along these fall streaks instead of along the height axis.
L. 477 - 480: The lidar and the radiometer show supercooled liquid. As they are direct observations onsite and given the difference in profiles from radiosondes and ERA-5 are significant (Fig. 15), I would trust the measurements more than ERA-5 and phrase that accordingly. Also, please include the figures of the depol lidar (att. BSC + depol) and radiometer LWP in the appendix as they are valuable synergistic data.
L. 492 - 493: I disagree. Please check the literature for radar Doppler-spectrum-based analysis of midlatitudinal cirrus. Additionally, how is this paragraph related to your research question?
I really miss a detailed answer or summary in "Discussion and Conclusion". Along a track, that spans from near ocean to 3000 m a.s.l. and temperatures from near zero to - 50 °C, microphysical conditions will differ greatly. Please highlight the novelty of this paper more clearly.
## Minor comments ##
L. 1: In this case, better refer to "intrusions of coastal moist and warm air" as "warm air intrusion". Even though your event does not classify as one, the general point of this sentence stays true.
L. 4 - 9: I would recommend to restructure a bit here. Now it sounds a somewhat, as if the instruments were set up in a hasty operation during the WAI-event. Maybe start "To tackle the uncertainty, we placed instruments along a transect..." and then later "In Feb 2025 a WAI event passed over our instruments, giving us the opportunity..."
L. 8: Please state the frequencies here.
L. 9: reflectivities instead of reflectivity
L. 10: Is the MDV corrected for vertical air motion? If not, refer to it as high MDV not as fall velocities.L. 25: It sounds odd, please rephrase for clarity
L. 43: Always cite multiple references in chronological order.
L. 51: Split the sentence into two and elaborate on what happens when "the larger particles exit the Rayleigh regime". Mention the Mie scattering regime.
L. 61: The full name is Neumayer Station III.
L. 66: Maybe rename it into "WAI-like event".
L. 72: Check if subsections are more suitable?L. 76: "At DDU..." Remove that, since it is not relevant for the study.
L. 85: What decided on exactly those positions?
L. 86: Replace transect by OPUs?
L. 89: Add the manufacturer of the radars.
Fig. 1: Enlarge the map, so orange and blue dots are distinguishable.
Fig. 2: This figure shows a lidar. Add it to the auxiliary instrumentation description and explain why/why not you have used its data (e. g. for identification of supercooled liquid droplets)
Tab. 2: Check MRR-Pros first range gate.
L. 116: Add temporal interpolation.
L. 117 - 121: Please rephrase this paragraph. Start with explaining that an offset from the vertical was found for the BASTA radar.
L. 122: Please use one or two sentences describing the principle of peakTree and why you need to identify regions of multi-modality in Doppler spectra. Add which settings you used for peakTreeFig. 3: The fronts in 3a may not be right. It looks like the occlusion starts at 63° S. Please add fronts at 3b too. Add units to the TCWB anomaly colorbar.
Fig. 4: Remove unnecessary information here. Without discussion neither potential temperature nor specific humidity give any additional value here. The blue dots are badly visible.
L. 182: Describe the amount of decrease in reflectivity.
L. 186: Motivate your choice for the chosen Ze-S-relations
L. 188: Are there wind speed measurements of those stations? Very low wind speeds would exclude the option of BLSN
L. 188: What is the threshold of spectral width for the turbulent layer?
L. 191: Replace "parameters" with "Ze-S-relation"
L. 197: Remove "tall"
L. 197: Better show your evidence for katabatic winds here, as you mention it here and not later in L. 203.
L. 199 - 200: Please show your BLSN detection and what sensor it is based on.
L. 218: In this paragraph, highlight more clearly, which aspect is considered in which secton.Fig. 7: d) Please overlay the 0.4 rime mass fraction as in a); x-axis labels in d) should be simplified to indicate date (Feb 15 - 17) and time (as HH instead of HH:MM); Additionally, motivate the choice of times for which profile comparisons are made. It would really help to see Z, MDV and spectral width again here. Also, add an overlying grid for all plots shown in this manuscript. E. g. Fig. 5 is too small to see at which height exactly MDV decreases.
L. 225: Either use MDV as acronym or write mean Doppler velocity.
L. 228: what is "high reflectivity" exactly?
L. 231: In Fig. 5 there is also a layer of enhanced spectral width, may newly formed particles explain the reduction in Doppler speed and increase in reflectivity? Additionally, there is also a region of more frequently detected double peaks in Fig 7a
L. 256: Please explain, how saturation of the DFR signal can happen?
L. 257: This is hard to see in Fig 7 b+c). I suggest adding contour lines of elevated DFR_K-Ka in panel b) to make the vertical positioning more comparable. Please also add single profiles of DWR_Ka-W and DWR_K-Ka in one figure to show more clearly at which altitude they increase, respectively.
L. 258: Add a comma between the subscript e and K, Ka, etc. at every other occurrence.
L. 262: Start where? Colloquial, please rephrase.
L. 265: How is advection into the radar beam consistent with strong vertical air velocities (and also, why are you not using "wind speed"?)
L. 265: Does that mean, that you expect no growth of particles smaller 4 mm in the whole radar beam, despite strong updraft?
L. 274: Please add the height to the time span. Also show given time intervals as shading or polygon in your plots.
L. 280: Why is that presence not shown in Fig. 8d)?
L. 284: But in Fig. 8, there is DFR K-Ka shown
Fig. 8 c): Why is there no area of close-to-zero DFR K-Ka as in Fig. 7 c)?
L. 290: Please add a figure with reflectivity and DFR gradients. It is difficult to estimate from the plot with absolute values.
L. 291: That seems not correct. The updraft region shown in Fig. 8 d) extends up to 6 - 8 km, while the height of the isothermal region is 3 km. Please clarify.
L. 298: Please provide a comparison of the reflectivity value of the slow-falling peak at D47 and D17. Also provide information on the LDR of the slow peak.Fig. 9: Put it on top of page 17 and add which radar is used for Reflectivity, MDV and LDR in this plot.
L. 305: Add the relevant height ranges.
L. 310: Replace "velocity" with MDV
L. 316: At which height are there three modes?
L. 325: The third mode is again at 0 ms-1, so why couldn't it be more supercooled liquid, as you mentioned above? It is unclear which mode you mean with "third" or "other". Also it appears at 2km, so already at around -10 °C, which would mean it is not in the rime-splintering T-range. Please check and clarify.
L. 333: Which other spectrograms?
L. 334: This points to that ERA5 data should be taken with a grain of salt. Please clarify.
L. 336: That is difficult to see for Ze and MDV. Please mark those time frames in your plots
L. 337: What is "some reaction"? Colloquial, please rephrase.
L. 338: What would you classify as "strong reaction"? What as "weak reaction"?
L. 339 - 340: Check time periods of when LDR changes happen and add altitude ranges. From which height is the LDR increasing/decreasing?
L. 341: Why 06.45 - a time when the cloud layers are non-continuous in height and Fig 11d) shows less coherent number of peak structures chosen?Section 7: Link radar cloud top height to the column average LDR. That could explain temperature dependence of particle shape and thus LDR.
L. 361: Please show the rime mass fraction to make the figures more comparable.
L. 362 - 365: This is unclear. Is there supercooled liquid water (or not) and thus (light) riming happening at D85 (or not)?
L. 391: How much is "moderate"? Please add spectral MDV value.
L. 402: Sounds unclear, please rephraseFig. 15: Change x-axis label to "Wind Speed [ms-1]"
L. 416: By what percentage do columns dominate?
Fig. 16: Why don't you show the Doppler spectra at the times of the ICE-CAMERA images?
Please put each box's description into a separate paragraph.Particle types at D85 and DMC are a major part in the paper but completely missing here. Why?
L. 504: Please rephrase "not far-removed". That is cool, that conditions are similar to mid-latitude winter, but how does it answer anything of your research gap? That point shouldn't end without further explanation.
L. 505 - 513: That point is also cool, but please set it into context with your own work.
L. 516: Prefactor instead of multiplierFig. A1: Why do you use two different colors for the area of the best linear fit? Why not shading it?
L. 591 - 592: Please check, if liquid layers can be identified by the lidars in the OPUs.
L. 627: I disagree: There seem to be considerable differences in minimum detectable reflectivity of the MRR at the four sites. Please correct.
L. 639: Actually, there is a decrease of DFR-Ka-W between 4-6 mm particle size
Citation: https://doi.org/10.5194/egusphere-2026-3964-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 158 | 94 | 33 | 285 | 35 | 33 | 24 |
- HTML: 158
- PDF: 94
- XML: 33
- Total: 285
- Supplement: 35
- BibTeX: 33
- EndNote: 24
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The manuscript by Corden et al. presents a detailed observational study on cloud and precipitation microphysics during a warm air intrusion reaching the interior of Antarctica. Based on an unprecedented dataset of radar observations along a 1100km transect, they investigate changes in microphysical properties along the slope of the East Antarctic Ice Sheet. Those observations are crucial to advance our understanding of precipitation impacts on the ice sheets surface mass balance. The manuscript covers the general concept of the data collection with three autonomous station and first case studies. It fits well into the scope of ACP, especially considering that this manuscript likely opens up the stage for subsequent, more detailed studies on a unique dataset.
A few minor points might be considered before final publication: