the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Nov 2025
The role of cyclonic eddies in the detachment and separation of Loop Current eddies
Abstract. The Loop Current (LC) and its associated eddies, known as Loop Current Eddies (LCEs), are key oceanic features in the Gulf of Mexico. Using a statistical analysis of 29 years of satellite altimeter data (1993–2021), we show that more than half of the LCEs that detach from the LC reattach within 30 days and that only 42 % truly separate from the LC and move westward in the Gulf. Our observational analysis also shows that i) before a detachment can occur, the LC needs to extend far enough north in the Gulf to reach the Mississippi fan (∼27.5° N); ii) the ratio of separations to reattachments depends on latitude, with detachments being more prone to reattach if they occur south of 25° N and to separate if they occur north of 25° N; and iii) cyclonic eddies are consistently present during the detachment process, with one cyclonic eddy on the eastern side of the LC if the LCE is to reattach and one cyclonic eddy on each side of the LC if the LCE is to separate. In the latter case, the co-occurrence of eastern and western cyclonic eddies in the LC bottleneck zone forms a large cyclonic structure. This large cyclonic structure is often observed during separation events, and when it is absent, LCEs are more likely to reattach, indicating a potential role in modulating the LC extension into the Gulf of Mexico. Sometimes, it can restrict LC growth on time scales of several months. The observed associations between cyclonic eddies and LCE detachments provide a statistical framework that could help anticipate separation events.
- Preprint
(10989 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-5574', Anonymous Referee #1, 13 Jan 2026
-
RC2: 'Comment on egusphere-2025-5574', Kathleen Donohue, 15 Jan 2026
This is a very interesting paper that reveals patterns of meso-scale variability in the Gulf during Loop Current Eddy formation. Through composites of satellite SSH, the authors show that the "necking down" preceding Loop Current eddy separation is associated with two cyclonic features. On the eastern side, a meander along the periphery of the Loop Current grows through some combination of baroclinic instability triggered by vortex stretching due to Loop Current interaction with the Mississippi Fan. On the western side, a cyclonic eddy with origins in the Caribbean and/or Yucatan Channel. This pattern is clearly present for southern separations. In other words, while detachments occur with just the eastern cyclone, separations require a source of cyclonic vorticity along the eastern edge. This paper will have significant impact and represents an important step toward understanding and predicting Loop Current Eddy separations. I clicked the 'major revision' button but I think its closer to minor revision. well done.
Major Comment
The discussion could be substantially strengthened. As it stands, the results lean toward kinematic rather than dynamic interpretation. I encourage you to place these results in a more dynamical context even if somewhat speculative or perhaps by interpreting your results in the context of previous work. I believe this work will serve as a launching point for future studies.
For example, modeling studies suggest barotropic instability is important for detachments. How do the cyclonic eddies interact with this process? Could either the western or eastern eddy strengthen the horizontal shear, thereby enabling barotropic instability? Additionally, are the western eddies truly "fundamental"? Is their role dynamical, or do they simply narrow the Loop Current width? Is there a relationship between the eddies? Is it appropriate to say that the eastern cyclone gets established and as soon as a cyclonic eddy propagates into the Gulf, separation occurs?
Is this worth exploring? The deep, nearly depth-independent cyclone in the channel just north of Yucatan (see Perez-Brunius et al. 2018, Fig. 4) may be key. Perhaps the western cyclones squeeze the Loop Current, reducing its width, while the eastern cyclones and the deep channel cyclone then accomplish the separation. I think that the large cyclonic feature may just be the surface expression of the depth-independent gyre (24.5N, 85.5W) rather than the merging of the western and eastern cyclone – but this is a hypothesis.
Additional Comments
The paper could benefit from increasing the clarity by homing in on the key results. I appreciate the comprehensive nature of the work – it’s impressive yet perhaps not all needs to be in the main body of the paper. Some figures are quite small and may not be essential to the main narrative (see specific suggestions below). Given the comprehensive nature of this work, consider moving some material to the appendix to streamline the main text.
To be clear, the following are suggestions and I’ll leave it to the authors to decide whether they benefit the manuscript.
Kathy Donohue
Abstract:
Line 8: ‘latter’ I don’t know what ‘latter’ refers to here. The preceding sentence is quite long and so latter is not clear.
Line 12: This is an interesting comment that is not discussed (much) in the summary and discussion section.
Introduction
Line 42: why is cyclonic capitalized?
Line 50: Worth citing Oey?
Oey, L. Y., H. C. Lee & W. J. Schmitz (2003) — Effects of winds and Caribbean eddies on the frequency of Loop Current eddy shedding: A numerical model study, Journal of Geophysical Research: Oceans, 108(C10).
https://doi.org/10.1029/2002JC001698
Line 40: Barotropic/baroclinic energy conversions. Donohue did not even consider the barotropic eddy conversion so the studies are not at odds with one another. I don’t see how an observational study that shows the baroclinic conversion modeled by Hurlburt and Thomson, Oey and others is controversial. Meanders along the eastern side, deepen, these cyclonic features often have high Rossby numbers – they may be existing features that have propagated in from the west. The deepening is a baroclinic intensification process. All models show this.
Yang et al. (2023) show that during Loop Current Eddy detachment, the eddy kinetic energy increase is primarily fueled by barotropic instability, with barotropic energy conversions dominating the horizontal energy budget relative to buoyancy forcing during detachment. I would argue that the energy conversion in the MITGCM might be underestimated based on the low deep eddy kinetic energy relative to observations (see Morey’s paper).
Line 42: I would change the topic sentence to ‘Omnipresent in both the GOM and the Caribbean.’ Most of this paragraph is about Caribbean vorticity entering the Gulf. This is an opportunity to discuss both cyclonic and anticyclonic vorticity into the Gulf.
Line 55: I don’t love ‘debate’. Reads more adversarial than necessary. Consider something like -- reflecting the complex and not yet fully resolved role of Caribbean cyclonic eddies in the shedding process.” Or reflecting our incomplete understanding of Caribbean cyclonic eddies, perhaps due to the difficulty in both observing and modeling these features. (This last sentence would set the stage for your section about how the altimeter barely resolves these cyclones.)
Line 62: as mentioned before, I think this large cyclonic feature is really that deep cyclone sitting at 24.5N observed by the Rafos floats.
Figure 1 – nice figure! The caption needs a rewrite. Maybe just delete that first a) and write Figure 1. Detachment locations. Odd to put panel d in the lower left – you don’t label the other figures in this clockwise manner. Could you make the height of panel d smaller – I spent misguided time trying to align the latitudes in histogram plot with the maps.
Line 111: Detachments that lead to reattachments maybe say Detachments with subsequent reattachments. Lead sounded odd. Would one say detachments that lead to separations?
Line 133: I read this over and over and I’m still not sure what the authors are communicating. Please consider a rewrite.
Figures 3, 4, 5, 6, 7, 8:
-why this colorbar for the vorticity? Why not make negative vorticity red and positive vorticity blue. In that way, the anticyclonic features are red in the SSH anomaly plots and red in the vorticity plots.
Are all the panels in 4-8 necessary. The text hardly talks about the Ro# or occurrence of cyclonic eddies. Could these be in the appendix or a supplement. In that way, you might be able to condense to 4-8 to 1 or 2 plots. I think that would really help the reader see the patterns at a glance.
The titles should be changed because they are consistent with the text. For example, ‘below’ should ‘south of’ and above should be ‘north of’ Perhaps you were trying to keep the titles short.
Figure captions generally: Please take a careful look at the captions.
Lines 182-189: This is very confusing. Please try to rewrite.
Line 189: Self-serving sure, but you say frontal eddy, I say meander and while there may be eddies merging I think it was clear from Donohue that substantial meander growth happens through baroclinic instability and that it’s the trough (cyclonic vorticity) that grows.
Figure 10 uses both term barrier and blocking eddy. Please clarify if they are different. How is barrier eddy defined in the analysis leading to Figure 10?
Figure 11: typo in caption for the bottom panels. It should be EMF, I think.
Figure 13: last sentence. ‘that is not visible in AVISO’. I’ll argue that the cyclone is visible -- it is just weak and displaced relative to SWOT. This displacement is likely do to sampling of the ‘standard’ altimeters. These are nice figures!
Line 236: SLA not SALCitation: https://doi.org/10.5194/egusphere-2025-5574-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 166 | 144 | 22 | 332 | 15 | 14 |
- HTML: 166
- PDF: 144
- XML: 22
- Total: 332
- BibTeX: 15
- EndNote: 14
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General comments
In this manuscript, the authors use satellite altimetry observations of the Loop Current (LC) and of ocean eddies to analyze the role of cyclonic eddies in the detachment and separation of LC Eddies. They find that temporary detachments are more frequent as they happen south of the Gulf, as opposed to final separations that are more frequent in the north. They also find that cyclonic eddies tend to be present to the east of the LC in temporary detachment cases, and on both sides of the LC in final separation cases. Finally, they find that the merging of these two cyclonic eddies leads to a large cyclonic eddy that can block the northward extension of the LC.
This is a very nice study, and I enjoyed reading the manuscript. It is generally well written and presents very interesting results that provide a very solid background for analyzing future LC detachment/separation cases. I think the manuscript would benefit from discussing some of its results in relation to previously published results that I believe are relevant to the present study. I also think some aspects need to be clarified. Please see the specific comments below for more details.
Specific comments
Some of the results presented here nicely complement previously published results. For example, Le Henaff et al. (2014) found that cyclonic eddies along the Campeche Bank tend to be observed immediately after a detachment or separation of an LC Eddy, which suggested that, since such eddies are involved in the pinching off of LC Eddies (Zavala-Hidalgo et al., 2003), they could only do so after the LC Eddy reattaches to the retracted LC. This is consistent with the present study, which demonstrates that temporary detachments are dominant in the southern part of the basin, meaning that LC Eddies frequently re-attach to the LC there. Since the present study also demonstrates that such eddies along the Campeche Bank tend to be involved in the final separation of a LC Eddy, this suggests that there is a typical sequence starting with the initial, temporary detachment of an LC Eddy, under the influence of a cyclonic eddy on the eastern side of the LC (based on the present study), followed by the formation of a cyclonic eddy on the western side of the LC, which can later be involved in the final separation of the LC Eddy. Such a sequence is consistent with the modeling results from Yang et al. (2023), who found that, during a 2010 LC Eddy shedding sequence, a cyclonic eddy along the Campeche Bank forms only in simulations in which the pre-existing cyclonic eddy on the eastern side of the LC is sufficiently intense to lead to an LC Eddy detachment (and independently from the boundary conditions from the Caribbean Sea). Results from the present study thus suggest that such a teleconnection between the initial presence of a cyclonic eddy on the eastern side of the LC and the later formation of a cyclonic eddy along the Campeche Bank might be quite common. I think it would be nice that these ideas are discussed in the manuscript.
In this context, regarding the role of cyclonic Caribbean eddies discussed in the manuscript, which is also important as illustrated in the study, it is possible that the timing of such eddies matters: a cyclonic anomaly propagating from the Caribbean Sea to the southern Gulf at the time of the initial LC Eddy detachment might find favorable conditions for intensification or growth (allowing it to be detected by current observing platforms), whereas it might not intensify otherwise.
Regarding Figures 4, 5, 6, 7, and 8: the captions mention that these composites are for separation cases, but the manuscript mentions that these figures include both reattachment and separation events (l. 165). This needs to be clarified, as some analyses in the text would not be possible if these figures merge both types of events (l. 175-177, l. 205-206).
In the initial analysis of Figures 1 and 2, point 3 (l. 133-134): I am not sure I fully understand the reasoning here. Can the authors better explain this result? Also, the equivalence between an 1800 km length and the distance to the Mississippi Fan is only mentioned in the summary and discussion, later in the manuscript.
Technical corrections
- l. 23: Shay et al. (2000) could also be cited here.
- l. 62: I suggest “tends to form” instead of “forms”. Otherwise, there would not be cases in which this does not happen, mentioned in the following sentence.
- l. 97: Tables in the Appendix are named A1, A2 etc., not S1, S2. Also, tables A3 and A4 do not seem to be used in the manuscript. If they are not, they should be removed.
- Figure 1 caption: There is no MF on the figure, and there are no continuous gray contours.
- l. 167: In the appendix it is Figure A1, not S1.
- l. 178 and 180: I believe it is “south of 24oN” instead of “south of 25oN”.
- l. 180: I believe it is Figure 4 instead of Figure 5.
- l. 180 and 182: the locations of the cyclonic eddy with respect to the LC are hard to identify, since the shape of the LC changes. In Figure 6 the mean LC appears to be detached at -30 days, so it is hard to tell what North, East or West of the LC means, compared to Figure 5 for example.
- l. 216-218: Based on Figure 10d (not Figure 10a-d), it seems that barrier eddies persist for more than 5 months (not 4) northwest of the LC after separations that occurred below 24◦N, and 4 months (not 3) after separations between 24 and 25◦N.
- Figure 10 caption: There is no subplot e, so the text “The cumulative persistence of blocking cyclonic structures over the 29-year period is shown in e. The cumulative persistence is computed by multiplying the mean persistence of blocking cyclonic structures by the number of separations per region” should be removed. Also, what are the 3 types of LC contours visible on Figure 10a-c?
- l. 239: SLA, not SAL.
- Figure 11 caption, I suggest: “Hovmöller composites of SLA (first column) and the occurrence of cyclonic eddies from the AVISO Atlas (second column) during separations occurring below 24oN etc.”
- Figure 12 caption seems to have many mistakes. I suggest: “Hovmöller composites of SLA (first column) and the occurrence of cyclonic eddies from the AVISO Atlas (second column) during reattachments occurring below 24◦N (first row), between 24-25◦N (second row), and between 25-26◦N (third row). Thick continuous-black contours etc.” There are only 3 rows, not 5.
- Figure 13 caption: In the second column, it seems that SWOT data are blended with CMEMS data outside the SWOT swath, is that correct? If so, this should be mentioned.
- Figure 14: Based on the description of the thick black line and the thin black segmented line, I believe these lines should be the same on the 1st and 2nd rows. Can the authors explain?
- l. 279: I suggest starting a new paragraph after “from the LC.”
References:
- Le Hénaff, M., Kourafalou, V. H., Dussurget, R., and Lumpkin, R. (2014). Cyclonic activity in the eastern Gulf of Mexico: Characterization from along-track altimetry and in situ drifter trajectories. Prog. Oceanography 120, 120–138. doi: 10.1016/j.pocean.2013.08.002
- Yang, X, Le Hénaff, M., Mapes, B., and Iskandarani, M. (2023). Dynamical interactions between Loop Current and Loop Current Frontal Eddies in a HYCOM ensemble of the circulation in the Gulf of Mexico. Front. Mar. Sci. 10:1048780. doi: 10.3389/fmars.2023.1048780
- Shay, L.K., Goni, G.J. and Black, P.G. (2000). Effects of a warm oceanic feature on Hurricane Opal. Monthly Weather Review, 128(5), pp.1366-1383.