the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Nov 2025
Relationships between Thin clouds, Opaque Clouds, and the Tropical Easterly Jet over the Indian Region observed with Aeolus Spaceborne Doppler Wind Lidar
Abstract. Direct wind observations used to be rare over the Arabian Sea and the Bay of Bengal. Since the launch of ESA's Aeolus Doppler Wind Lidar, profiles of horizontal wind are acquired every day and are perfectly co-located with profiles of thin and opaque clouds. In this study, we show that from June to October 2020, during the South-Asian Summer Monsoon (SASM), high altitude clouds formed over the Bay of Bengal by deep convection in the afternoon are advected westward towards the Arabian Sea in the morning by the fast winds of the Tropical Easterly Jet (TEJ). Consequently, the thin high cloud cover over the Arabian Sea is 14 % at 06 LT on days where westward winds are faster than 23.6 m s-1 between 14 and 17 km of altitude, more than twice as much than on days where westward winds are slower than 23.6 m s-1 (6 %). While the TEJ is primarily driven by the thermal contrast between warm land and cooler Indian Ocean, we observe that the diverging-rotating outflow around deep convective (opaque) clouds can strengthen or weaken the TEJ over the Arabian Sea with respect to the thermal wind, explaining a 3 m s-1 amplitude variation during the SASM. These results suggest that the cirrus cloud cover over the Arabian Sea may decrease in the next decades, as the TEJ and the convection over the Eastern Indian Ocean are expected to decrease in intensity.
- Preprint
(6673 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-5335', Anonymous Referee #1, 02 Feb 2026
-
RC2: 'Comment on egusphere-2025-5335', Anonymous Referee #2, 14 Apr 2026
Summary:
This paper uses a dataset (introduced in a previous paper) that has co-located cloud and wind profiles. The lidar retrievals reveal a relationship between the Tropical Easterly Jet, convection, and cirrus clouds over the Arabian sea during the South-Asian Summer Monsoon (June-October 2020). Cirrus clouds are twice as likely to occur on mornings following a strong TEJ the evening before. The authors hypothesize a few mechanisms for the strengthening of the TEJ, concluding that deep convection is a major source of divergence from the thermal winds derived from theory. The time lag between events suggests a causal relationship on the frequency of occurrence of cirrus clouds.
General comments:
I think there are good results here and I particularly appreciate the schematics for describing the complicated relationships and scenarios. However, the paper could be more concise and clearer with a different organizational structure. I like the proposed hypotheses in Section 4, but they are mentioned in Section 3 as well (but without context). The authors can reorganize Sections 3 and 4 to lay out the theory (hypothesis) first then relate it to the evidence from Aeolus. Section 5 also needs restructuring (moving the Methods subsection (Sec. 5.1) to the Methods and Dataset section (Sec. 2). There is a new idea introduced in the conclusions section (Sec. 6) on the impact of changing climate on this work, which should either be introduced as a theoretical implication of this work as TEJ is expected to decrease with warming or not be included in this paper at all (saved for future work).
Introduction: Again, this section needs to be reorganized – introduce the TEJ and do literature review then state the purpose of this paper as it adds to the literature. There is already a wealth of literature on this, so you really need to be clear in what the novelty of this study is. The novelty is that it focuses on shorter timescales (intraseasonal) but that is also a limitation of this study (cannot generalize or create a climatology). The novelty is also in the use of the instrument data (though it is introduced in a previous paper) and the ability to show a time lag between deep convection over the Bay of Bengal with cirrus cloud fraction over the Arabian Sea 12 hours later.
Statistics of the data: The scarcity of data is a bit concerning for drawing conclusions based on one year of data over a small area. I’m not sure that the conclusions can be generalized to multi-year or climatology or conjecture for changes with warming. Additionally, in lines 343-344, the total number of days does not match the breakdown of fast/slow days even considering the days for which there are no Aeolus measurements; this discrepancy needs to be addressed.
There are some interesting results throughout the paper, but the presentation was a bit confusing. It felt like a data paper and then a theory paper, but it should be combined so that the data supports the theory. In some instances, the authors are present clear evidence and explanation of the phenomena, but other times, the purpose of a certain figure/paragraph is not clear until much later and is distracting from the main results that cirrus clouds over the Arabian sea are influenced by both the TEJ and deep convection to the east.
Specific comments:
Terminology:
- I think you should define cirrus = thin clouds and anvil clouds =thick clouds from Aeolus, then stick to using “cirrus” and “anvil cloud” respectively. This will make everything more easily interpretable.
- Similarly, you could move away from HCFopaque and HCCthin to deep convection/anvil cloud fraction and cirrus cloud cover.
Formatting:
- Percent sign should come directly after the number (1% not 1 %)
Lines 28-35: This paragraph should be moved later in the intro. Make it clear this is a result of your study (“We hypothesize…” etc.). Maybe combine with the paragraph starting with “The aim of this study is thus to investigate the relationship between…”
Lines 38-39: This reminds me of a paper – Holton and Gettleman (2001): Horizontal transport and dehydration of the stratosphere.
Lines 53-54: Ye et al. (2023) focused on June, not July
Line 91: What is the time period and spatial coverage of this dataset? You go into more detail in Section 4 (Figure 4), so maybe move section 4 (Lines 189-201) to this data and methods section.
Line 91: Reference to Titus et al., 2025 should be updated to actual paper, not the preprint
Line 103: Do you ever distinguish between profiles 2 and 3 or are they all binned together? It seems like profile #3 would be thin cirrus with underlying low clouds - perhaps no meaningful difference between these two profiles
Section 3.1 Title might be better described as “Properties of the TEJ and clouds in 2020”
Fig 2: Is this a domain-average over the Indian region?
Line 139-142: Do you see this anticyclonic circulation and upper tropospheric thermal gradient in SASM 2020?
Section 3.2: This section could be more concise. The main take away of the section is that there is a strong diurnal contrast in deep convection, while the contrast in cirrus clouds is not as large. The key point here is the time lag between deep convection (18LT) and cirrus (6LT), which leads into your hypotheses in Section 4.
Line 158-162: This is a confusing statement. Kripalani et al., 2022 said that 2020 was an anomolous "erratic" monsoon year, yet you find that it is in "good agreement" with the typical diurnal cycle of convection? Can you really make this statement with the limited data that you have?
Line 162: The Arabian sea received less attention than the BoB considering deep convection in the literature? And what is your point? The rest of this paragraph is about the Arabian sea, not the Bay of Bengal which would have made more contextual sense after this sentence.
Line 164-166: Do you see evidence of the diurnal cycle of SST driving deep convection in your results? You could look at SST in reanalysis during this time period.
Line 174: Where is your evidence for this conversion from thick to thin clouds overnight?
Line 175-77: How do you reconcile this non-local convective origin with the sentence above that thick cloud are converted to thin cloud during the night... isn't that still a local effect since they are in the same region? Or else, what do you mean by "non-local convective origin" - do you mean in time or space? This is an interesting concept (origin of cirrus clouds) but I don't think you can make this argument here without evidence. Instead, you could propose these two mechanisms for cirrus cloud formation and leave it to future work to quantify/prove which hypothesis is true.
Line 177-179: You have space for one more subplot in Fig 3 so you could show the zonal mean across longitude in this region for each type of cloud (4 lines on one plot). Then you don't have to list the specific values, it will be clear from the plot.
Line 180: Yang et al. (2010) - Is this the right citation? I don't see any mention of longitudinal extent of cirrus in this paper on contrail cirrus...
Line 180: Das et al (2001; Fig. 8) shows a peak in the cirrus cloud occurrence and zonal wind around 90E whereas you are looking at 70E.
Line 184: the ratio of HCCthin and HCCopaque is really hard to tell by eye from the plots. I suggest a supplemental/appendix figure that plots the ratio.
Line 185-6: This hypothesis of the source of water vapor coming from deep convection then advecting westward, should be saved for later when you discuss the hypothesis in Section 4. I think some of this argument could just be moved to that section.
Section 4.1 & Figure 5 could be moved to an appendix/supplemental material to make the narrative clearer.
Section 4.2 title could be “Drivers of cirrus cloud cover” instead, more concise
Section 4.2.1 title could be “Hypothesis 1: Temperature anomalies drive changes in cirrus clouds cover”
Section 4.2.2 title could be “Hypothesis 2: Deep convection drives changes in cirrus cloud cover”
Section 5.1 needs to be moved to the methods section introducing the ERA reanalysis dataset as well as describing the analysis involved here.
Lines 242-245: This could be a bit more complicated since the presence of thin clouds can actually feedback on the temperature and act to warm the TTL (Fu et al., 2018 - doi:10.3390/atmos9100377). So you might not actually see cirrus associated with lower temperatures in the data. Additionally, temperature still has an important role in the formation of cirrus clouds in the TTL, temperature could also be impacted by gravity waves from convection. Finally, what are the uncertainties/variability in the temperature data over Aeolus track? And what are the uncertainties in cloud fraction/winds in Aeolus? This question of uncertainty should be addressed in the Methods and Data section.
Line 252-253: I really like this calculation! Does this mean that this analysis comparing the BoB at 18 LT and AS at 6 LT is only valid for "fast wind" days since slow wind days would take longer than 12 hours to advect that distance?
Line 264: Yes, the source of water vapor in the upper troposphere from convection is important. Have you also considered that thin clouds are directly formed via outflow from convection and persist for 12 hours or more? Cirrus clouds may have long lifetimes in favorable conditions.
Line 284: “HCFthin(AS, 06 LT) with respect to HCFopaque(BoB, 18 LT)” a (you defined alpha for a reason, so use it to simplify your explanations)
Figure 9: I like this schematic – could you somehow include the transport of water vapor as well as the simple convective outflow of clouds?
Line 334: the median value is -2.1 m/s but how sensitive are your results to this value – I assume it would be different for a different year of data. What is the standard deviation? mean? Maybe a scatter plot of u_allsky vs u_thermal would be informative (or a joint histogram). The method for calculating this cloud also be moved to the methods section.
Lines 343-344: Something is not adding up… 153 days total - 7 days without Aeolus obs = 146 days with data - (69+69 days strengthen/weakened) = 8 days unaccounted for. Are these days the exact same value as u_thermal? What is the prevision with which you can determine equality?
Line 345-346: Is there previous work that supports your analysis that deviations from the thermal winds are due to deep convection? Please add some citations here.
Figure 12: There is a small difference in the u_thermal for each scenario – is this difference significant? Does this suggest that the deep convection is feeding back on surface temperatures which impacts the thermal winds (an indirect effect of convection)?
Section 6 (Conclusions):
This feels like it is just a set up for your next paper which will use these results as a baseline for your CMIP model results. If that is the case, this paper could be reduced to the relevant figures and kept as a separate paper, or if could be modified to suit your needs for the next paper.
However, I do feel like the schematics in particular could be a nice contribution to literature from this paper and that the observations support these "hypotheses" that you’ve presented.Line 409-410: I would reframe this; cirrus clouds act to warm their local environment which may complicate these results...
Lines 412-414: Earlier you also mentioned water vapor brought to the UTLS by convection then advected (forming cirrus later) which is different than cirrus detrained directly from convection and advecting hundreds of kilometers. This separation is not clear here.
Line 424: Lemburg et al. (2019) and Diongue et al. (2002), right?
Lines 424-425: Did you actually show that u_allsky(BoB 18 LT, 12 hours before) did not have an anomaly associated the TEJ at 18 LT? Like Fig 11 but for u_allsky(BoB 18 LT, 12 hours before) rather than HCCopaque(BoB 18 LT, 12 hours before).
Lines 430-445: This whole paragraph should have its own section in the results "Implications for future warming". This is a nice back of the envelope calculation but it should not be only introduced and discussed in the conclusion section. This is a huge pivot from the focus of the rest of the paper, so it should be introduced earlier.
Lines 446-448: This would be a cool paper but this and the previous paragraph could go in the intro to that paper, but the conclusion section to this paper should focus on your actual results, relevance/novelty to existing literature, and future work that could build on your work or support your hypothesis. For example, it could be relevant to note that a dataset with higher temporal resolution could better show the evolution in time from triggering of deep convection over the BoB as it moves towards the Arabian sea.
Technical Corrections:
Lines 42, 44, etc.: Capitalize all instances of East-West/North-South to be consistent. Check throughout the paper.
Lines 54-55: This sentence doesn't make sense grammatically. Suggested edit:
Liu et al (2024) further showed that this longitudinal variability also has a latitudinal component, particularly in July, that changes with convective activity in the Arabian sea. They demonstrated that a strengthening of deep convection in the Arabian Sea and a suppression of convection in the Bay of Bengal drive the northwest shift in the core of the TEJ.Line 75: delete “perfectly” as it is an extraneous adjective
Line 81: replace “shortly” with “briefly”
Line 90: delete comma after winds
Line 103: delete “entire single” – extraneous
Line 109: “resp” Change this and next few paragraphs to use this layout:
gridded opaque (thin) cloud fraction profiles, CFopaque (CFthin).... among "opaque" ("thin") profiles... etc. Keep order of thick (thin) consistent throughout.Line 130: delete “comprised”
Line 170: “about” “above”
Line 184: Would make more sense as "Indeed, strong winds push further into the Arabian sea at 6 LT (Fig. 3g & h)"
Figure 7d: the caption says alpha but the subplot legend says f. Keep it consistent or explain the difference between alpha and f
Line 261: It is “associated with” not “associated to”
Line 262: Add reference to Fig. 7c.
Lines 278-280: You don’t need to explain this since it is evident from the equation.
Line 285: “and then seems to form a plateau” “and alpha seems to plateau”
Line 366: I think you are refer to panel c in Fig. 11. Change reference to Fig. 11c
Fig. 11: I think the schematics to the left of panel a and b should also be explained in Figure caption. They are nice intuition for the concepts here.
Figure 381: “counter acted” counteracted. One word.
Line 396: typo in number format. Should be period, not comma: -1.7 ms-1.
Line 419: “the temperatures, and these days were associated to” “their temperature gradients, which were associated with”
Figure 12 caption:
(1) Add “proposed” before mechanisms in the first line.
(2) the “average” situation might be called something else instead. If most days have u_thermal not equal to u_allsky (69 days strengthened and 69 weakened), then this “average” state is not so average. Perhaps this scenario is "in thermal wind balance"?
(3) The shading is gray, not brown.Lines 389-390: Maybe AN, AS and BN, BS would make more sense here (points A and B or Arabian sea and Bay of Bengal are easier to remember than P and Q).
Line 426-end of paragraph: Rephrase from “Additionally…” “Liu et al. (2024) showed that this dipole in deep convection explained the interannual longitudinal variability of the TEJ core; however, here we show that this dipole modulates the TEJ on much shorter time scales (on the order of days).”
Citation: https://doi.org/10.5194/egusphere-2025-5335-RC2
Data sets
Data/TEJ/* Zacharie Titus https://dx.doi.org/10.25326/746
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 331 | 114 | 34 | 479 | 48 | 54 |
- HTML: 331
- PDF: 114
- XML: 34
- Total: 479
- BibTeX: 48
- EndNote: 54
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General comment:
The manuscript by Titus et al. investigates the relationship between thin clouds, deep convective clouds and upper tropospheric wind structures over the Indian monsoon region using a synergistic cloud and wind profile dataset from Aeolus during the summer monsoon 2020 time period. They conclude that a higher occurrence of thin clouds over the Arabian Sea is associated with stronger westerly winds combined with higher amount of water vapor injected in the upper troposphere by deep convection over India and the Bay of Bengal.
They further show that the strength of the TEJ is mainly consistent with thermal-wind balance, but that the short-term regional variability of the TEJ is modulated by the outflow and divergence, vorticity perturbations caused by deep convection.
While Aeolus’s repeat cycle is usually 7-days, the authors benefit from the orbit configuration to have 12 hours between an ascending and descending orbit with spatial distances of 1150 km, which provides them the opportunity to study advection processes. This is unique for a single observing system.
The dataset is valuable and the topic is very relevant. The study addresses several interesting scientific implications, which are explained in detail. My only concerns are regarding the structural clarify in sections 3-5 plus a few minor comments.
Detailed comments:
The manuscript is clear and well written, the abstract, introduction and conclusion read very well and communicate the motivation and relevance of the study. However, I have several concerns regarding the structure.
While section 2 (Dataset and method) is very descriptive and easy to follow, sections 3, 4 and 5 present a mixture of observational findings (what does the data show us directly without interpretation), interpretations from these findings and hypothesis. Also, the subsection “method” re-appears in section 5 for the individual hypothesis about deep convection and the strength of the TEJ. This causes that the narrative reads as a sequence of results, observations, impacts and hypothesis with no clear logical hierarchy. I recommend separating the observational results, interpretations, discussion and conclusions, either globally, or if you like to keep the hypothesis-driven sectioning, also per section.
Section 5: Deep convection strengthens … This is more like a conclusion or hypothesis. Maybe rephrase title to “Impact of deep convection on TEJ strength over Arabian Sea”?
Section 4: I actually like the idea to already provide the conclusions with the title, but later, with 4.2.1 you refer to the role of temperature and 4.2.2. the role of opaque clouds. Maybe better call section 4: “Drivers affecting the occurrence of thin high clouds over the Arabian Sea” Then you go to 4.1. wind, 4.2. temperature and 4.3. opaque high clouds over India ? This is just a suggestion to make the order clearer.
The abstract is clear and well written, but you have some room for further details. You mention the Aeolus dataset, some results, numbers and implications, which is a nice structure for the abstract. One sentence about the methodology (disentangling the effects) would make it rounder.
The Sub-grid scale variability below the native resolution of Aeolus is not resolved by the merged and re-gridded cloud-wind dataset. You should add some sentence that this high-resolution dataset nicely compares to the resolution of CALIOP, but does not provide higher level of details compared to the native Aeolus resolution, which brings some uncertainties, especially with respect to the cloud top levels.
The manuscript reports convective perturbation wind components (u_conv) on the order of −1.7 to +1.3 m/s at the end of Section 5. However, no uncertainty estimate is provided for these values, nor is their magnitude discussed relative to the wind measurement uncertainty and derived thermal-wind uncertainty described in Titus et al. (2025). Without this context, it is difficult to assess whether the reported perturbations are robust and statistically distinguishable from dataset uncertainty. I recommend providing uncertainty estimates or at least add a reference of typical random errors of Aeolus Mie-cloudy wind and to compare them to the observed amplitude.
Minor technical comments: