the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 11 Aug 2025
Trade wind regimes during the Great Barrier Reef coral bleaching season
Abstract. The trade winds over the Great Barrier Reef (GBR) dominate the local weather in the region, bringing cooler and drier air over the Reef, which promotes ocean cooling. The absence of the trade winds is often marked by periods of weaker winds and higher humidity, known as the doldrums, which cause ocean temperatures to spike and can develop into marine heatwaves that lead to coral bleaching. As the shallow waters of the GBR are strongly tied to the local meteorology, studying the evolution and structure of the trade winds during the austral warmer months is essential for understanding the development of thermal bleaching events. Through a K-Means cluster analysis on reanalysis soundings from Davies Reef from December–April 1996–2024, we find the formation of the doldrums is linked to the passage of a Rossby-wave train over eastern Australia. Years with mass thermal bleaching are correlated with more doldrums days, but also less of the strong trade wind days in December and April which can promote early-summer warming and allow warmer temperatures to persist later into the season.
- Preprint
(10549 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-3639', Alexander Sen Gupta, 01 Oct 2025
-
RC2: 'Comment on egusphere-2025-3639', Anonymous Referee #2, 15 Oct 2025
Trade wind regimes during the Great Barrier Reef coral bleaching Season, by Lara S. Richards et al.
Main recommendation:
This paper seeks an explanation for coral bleaching events in certain weather/synoptic regimes and uses a K-Clustering technique to find those regimes. It addresses the character of those regimes, and links them to statistics on bleaching. The paper is overall well-written and clear, yet very descriptive and lengthier than need be to highlight the main findings. My recommendation would be to considerably shorten by drawing out what really matters. While some material is nice-to-have, it might better go into an appendix or supplementary material.
The paper is for example lengthy on the transitions between clusters, which relates to larger-scale dynamics and feedbacks between the ITCZ, tropical deep convection and the surface, which, while intriguing, is not the key focus of this paper or would require substantially different analyses. The paper ends with a conceptual figure 10, which is appreciated, but how clusters transition is not key to the bleaching events per se, and figure 10 tries to explain how they occur sequentially. What seems a key results, but which gets lost in the text, is that sustained days in the non-trade clusters – as well as a lack of early season cooling – have a relation to CBEs. An autocorrelation or persistence statistic is not presented.
I also imagine that what is important is how LHFs progress during the day (along with cloudiness), which is not analyzed. The diurnal cycle in SST and heat fluxes can be substantial in the doldrums, but the paper looks at daily mean fluxes. In other words, the # of sustained days in the non-trade cluster and the minimum LHFS/cooling rates observed during those days seem important to understand CBEs.
The conclusions are short and nice. The target is thus to do a major revision on the rest of the paper to bring out the essentials by removing material that is nice but not key, and to look at the diurnality of fluxes and heating/cooling rates. I describe these points more specifically below.
- Introduction: your objective is in the last paragraph; it would be nice to turn this into a more specific question. What are you specifically interested in? E.g. are all CBE’s preceded by a trade-wind breakdown?
- L52: boundary layer and trade-wind layer in literature are often used to denote the same thing
- L24; the doldrums are areas with subsidence, which would brin down drier air. Why is the humidity high? Or perhaps, do you want to say the humidity is confined to layers close to the surface (in the presence of subsidence).
- L25: boundary layer turbulence – do you mean convection?
- Figure 1: what does N represent? Number of days in each cluster? I don’t think the latter, because the number differs from what is shown as number of days in a later figure. Can you denote its meaning in the caption?
- Section 2.1. A large diurnal cycle in surface heat flux seems to be present over the doldrums from work I have seen. What is your motivation for picking the specific time of 10 LT? Is it a better constrained analysis? Are you avoiding certain sea breeze type of circulations? And why do you not pick the peak cooling/heating rates that are present in a day?
- Trajectories: I am a bit confused about specific days in the northerly wind cluster, of which many appear to have a back trajectory that goes towards the southeast/ Does this not tell us that the day-to-day weather variability/synoptic variability is very strong and that subsequent days fall into one cluster or the other. Also, a cluster does not tell us about persistent weather regimes per se, which seem key.
- Lines 144 – 145: I also see that in Fig 1 there is still that eastern high-pressure ridge, but additionally, the L pressure on the northeastern side of Australia is extended further down, which seems to me one of the main differences between the trades and non-trades. Also, in L376 you highlight this main difference (Low pressure area south of Davies Reef).
- In section 3.1 it appears that some statistical analysis such as autocorrelation would be helpful to address how long a certain cluster is sustained. it would be very interesting to learn about the persistence of one cluster, as it seems that persistence of doldrum conditions has more important on bleaching than if doldrum days persist only shortly and alternate with other clusters. In that sense the number of days in one cluster alone is not telling us the full story. You address the persistance only later in Lines 280-297.
- I find the transition table to be far less interesting than statistics on the number of persistent days during which non-trade events exist.
- I find Figure 10 and lines 391 – 405 to be very descriptive (with Fig 10 complex) and not key to explaining bleaching events, which are not uniquely related to a transition between clusters. In some sense explaining the large-dynamics of the trades itself goes beyond the scope of this work.
- Section 3.2 is overall very descriptive and detailed, so that key information is somewhat hidden, such as that the trades regimes are cooling SST in the GBR region, while the doldrums/northerlies regimes are warming the SST. Could this section better highlight the key information needed to understand the CBE events? Can other information/details go into an appendix?
- Section 3.3, while nice, could also be shortened. There are repeating sentences.
- Section 3.4 MJO could also be shortly summarized by stating that between trades and non-trade clusters you transition from inactive/suppressed to active phases of the MJO.
- In section 5 - last paragraph running from page 19 – 20, there is notable repetition in results discussed, except for the important point of duration and persistence of the doldrums, which I would highlight as one of the key aspects that should not come at the very end of the results hidden in between a paragraph. I would encourage to make this analysis on the persistence stronger.
- In the Discussion/section 6 (1st paragraph) it is discussed that CBEs relate to a persistence in non-trade clusters. Given that those seem particularly important, would it be good to look at what Pacific-basin wide synoptics were present during these persistent periods in particular? (or do you think these are well reflected in the cluster anomalies in Fig 4.
Grammatical errors:
L69: ‘was found to’ - > ‘were found to’
L85: for each day in each cluster or each cluster?
L294: respectively instead of respectfully?
L 372: ‘to form creating local SST heating’ incorrect grammar
L 386: fix citation style (Windmiller, 2024) - > Windmiller (2024)
Citation: https://doi.org/10.5194/egusphere-2025-3639-RC2 - Introduction: your objective is in the last paragraph; it would be nice to turn this into a more specific question. What are you specifically interested in? E.g. are all CBE’s preceded by a trade-wind breakdown?
-
EC1: 'Comment on egusphere-2025-3639', Shira Raveh-Rubin, 16 Oct 2025
Dear Lara Richards and co-authors,
Thank you for submitting this interesting paper to WCD. As you can see, the reviewers have completed their assessment of your manuscript. They value your work assessing the relationships between coral bleaching events and synoptic weather regimes via the induced wind, surface fluxes and related SST anomalies.
The reviewers provide critical comments and valuable suggestions to improve the work and its presentation, especially with regard to focusing the text and grounding it in the figures. With this major revision, please provide a detailed response to each of the comments. I am looking forward to receiving your revised manuscript and response.
Best wishes,
Shira
Citation: https://doi.org/10.5194/egusphere-2025-3639-EC1
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 774 | 54 | 13 | 841 | 43 | 39 |
- HTML: 774
- PDF: 54
- XML: 13
- Total: 841
- BibTeX: 43
- EndNote: 39
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Trade wind regimes during the Great Barrier Reef coral bleaching season
by Richards et al
This paper examines atmospheric conditions over a GBR site to try and establish a link to coral bleaching risk. The paper is generally well written and looks at a scientifically interesting problem. I do however have a number of comments and issues that require clarification or modification before I would recommend publication.
General comments
Except for one analysis related to SSTSA composites, there is no attempt to look at the statistical robustness of results. For example, differences in properties between clusters should be tested.
In figure 4 you show the SSTA associated with different clusters. However your clusters are most closely linked to wind strength conditions and therefore LHF anomalies and warming rate. I suspect that you would get a stronger and more informative plot if you looked at a composites of SSTA tendency (dSSTA/dt).
Unless I have my directions mixed up the vertical velocity composites and the Lagrangian analysis seem to suggest opposite behaviour. Given that most particles remain in the vicinity of the release location, I would expect consistent results.
Section 4 on coral bleaching, should be about coral bleaching risk (as there is no bleaching data used). Moreover given that corals bleach due to accumulated stress, I think this section is more about the risk of temperature extremes.
Figures should be more diligently referred to in the text
Why are you only going back to 1996? For ERA we have reliable information going back to the start of the satellite period. Beyond that it becomes less reliable.
Specific comments
Nice introduction
72: data were extended
Im not clear what you mean by extended. Extended from what?
75: to fully capture the austral summer
Austral summer is December to February. I think you are referring more broadly to the period with warmest ocean temperatures
2.1: Figure 1 shows air temperature and dew point temperature. Its not very if you are clustering on these two variable or on air temperature, dew point temperature, horizontal wind components at 925 hPa, 850 hPa, 700 hPa, 600 hPa, and 500 hPa
67: 0000 UTC (1000 LST)
Is there a particularly reason for using a snapshot rather than a daily average?
70: A clustering analysis was performed at Lizard Island and Heron Island (not shown), where similar weather regime clusters are found.
You might consider explicitly including some of this analysis if it demonstrates that your results apply more broadly – it makes your paper more valuable if its representative of a larger region
80: Day-of-year based anomalies are calculated using a daily mean climatology for the 1996-2024 period to remove the cluster’s seasonal biases
Not clear what you mean. The clusters themselves are calculated based on absolute variables rather than anomalies. Seems like you are just saying that you remove the seasonal cycle
83: and have then undergone a simple one sided t-test where only the anomalies with a p-value < 0.01 are considered
Im also unsure what you mean by this. You might use a t-test to see if the composite of a set of anomalies is different from 0, but Im not sure what you are referring to here
85: The daily back trajectories each start at 0000 UTC running for 72 hours from Davies Reef at 925 hPa
It might help to provide some motivation as to what you are trying to achieve with this analysis.
89: Davies Reef AWS
It would be useful to have a brief description of the site e.g. is this inside a lagoon vs open ocean
096: We note ERA5 was previously found to have a high correlation with the AIMS Davies Reef observations
But ERA5 is an atmospheric reanalysis. Are you now referring to atmospheric variables?
95: total cluster days assigned
Not sure what is meant by this phrase
100: daily averaged mean
Don’t need averaged and mean
110: Here we consider only a 1◦ box centred on Davies Reef taking only the 0000 UTC observations.
Cloud cover can vary rapidly over timescales much less than a day. Why only take a single time stamp if other times are available?
121: be the optimal number
There are objective measures to find optimal cluster numbers. Are you using these or are you just picking manually? I have no problem with the latter, but if you call it optimal, then you should indicate what you are optimising to make it optimal
Fig 1. Hard to see any differences in SST. Might be more helpful to show SSTA. Indeed I would expect low wind conditions to be most strongly related to temperature tendancy
Fig 2 given 18 years of data you could assess whether the changes in cluster proportions are actually significantly different from month to month
125 do you mean: are ordered based on the direction and the strength of the surface winds?
126: The classic trades are scarce
Better to be precise, the frequency of classic trade wind days is small …
127: compares
I think you mean, ‘is similar to’
129: wind shift
Do you mean wind reversal?
130: Are you referring to Fig 3?
130: slightly weaker surface ascent separating the summer trades ON AVERAGE. In general its important to remind the reader that you are talking about composite means not al members of the composite.
131: summer trades are evenly distributed
Here and elsewhere the terminology should be tightened e.g. the monthly proportion of summer trade days remain similar across all months
132: and a larger easterly component
Hard to make this out from the figure. Low level winds look pretty similar across the three clusters
134: The wet trades omega profile follows a similar shape
The wet cluster looks quite different to the others to me. It is associated with ascent throughout most of the column, while the others are associated with decent above 800hPa
136: strong boundary layer temperature inversion
In panel j the temperature decreases monotonically with height. I don’t see any temperature inversion, just a weakening of the lapse rate
137/138. Make sure you always reference your figures
141 the monthly PROPORTION of trade and non-trade cluster DAYS REMAINS ALMOST CONSTANT …
143: this ratio > this proportion
164: with many trajectories extending into the Southern Ocean
The southern ocean is usually defined as south of 60S. From your density plot it looks like very few trajectories would come from there
Fig 3. If negative omega is associated with ascent, its very surprising that clusters associated with the strongest ascent, the Trades and Northerlies are associated with the weakest upward motion in terms of the lagrangian particles (given that the highest proportion of particles is close to the release location). Could there be some issue with the sign convention here?
197: the same > a similar
206: shows warm SST anomalies confined mainly to the southern GBR
Looks like the anomalies primarily sit to the east of the GBR
235: relative humidity and rainfall
Here and elsewhere refer to tables and figures
236: high cloud (38%), deep convection (6%), and alto-stratus (11%)
Is there a particular reason to break up the cloud cover? Why not simply present cloud fraction, this is a far more important metric for you discussion?
Except for the SSTA composite analysis no statistical significance was presented. This is important for looking at whether differences in heat fluxes, MJO phases, cluster frequencies etc are significantly different between different clusters. Otherwise with the presentation of just cluster means its impossible to tell what results are robust.
265: The non-trade clusters show clear bias
Its true that the means are bigger, but are these differences statistically significant?
Section 4. Coral Bleaching
This is looking at bleaching risk not bleaching. Indeed I don’t think its even really looking at bleaching risk its looking at risk of temperature extremes
275: Using the local bleaching threshold (29.8◦C) for Davies Reef (Berkelmans, 2002) and the maximum +1 standard deviation temperature (29.4◦C).
Bleaching thresholds typically relate to cumulative metrics like DHW. The absolute temperature thresholds are just the point when stress starts to accumulate. This is why bleaching often occurs after the periods of maximum seasonal temperatures.
280: it may be reasonable to suggest, bit it wouldn’t be reasonable to conclude
282: cluster trends
Not sure what you mean by trends.
285: the doldrums stand out as the main difference
What does this mean? Are you saying that doldrums have the largest difference in no.of days compared to the other clusters?
Is this difference statistically significant?
It looks like the frequency (which I assume means number of days) goes from about 18 to 21. This seems like a very small difference to explain whether we are getting mass bleaching or not.
302: this does constrain the statistical robustness
With 8 bleaching and 10 non bleaching you can still conduct a statistical analysis.
305: show the highest risk for coral bleaching
As above I don’t think this is sufficiently precise. Moreover I don’t really agree that you are looking a bleaching risk, you are looking at risk of temperature spikes. Bleaching risk would relate to accumulated stress.
307: 2016 (El Niño), 2022 (La Niña), and 2024 (Neutral)
Did bleaching actually occur on this reef in those years?
315: In general, all three bleaching events show the same well-documented meteorological pattern where periods of ocean temperature spikes coincide with a change in wind direction
It is the increase in temperature leading up to a peak, not the peak itself, that should be associated with positive heat fluxes (and typically weak winds). This is why the net heat flux is strongly anti correlated with the wind speed in you figure (not with temperature)
324: The first heating spike
The first heating spike occurs right at the start of the timeseries. Id suggest providing a date
334: The doldrums are not only the most frequent but generally persist the longest with an average duration of 3.9 days with 30% of events ≥6 days
Are you referring to the 3 years or in general across the full timeseries?
This question applies for the whole paragraph, please make it explicit
335: Similarly to the results in cluster climatology
Not sure what you mean, it would help to refer to figures.
337: temperature gradients > temperature trends
337: (0.09◦C/day) (-0.08◦C/day) (29.2◦C)
Needs more explanation.
349: the most pronounced temperature drops tend to occur during the wet trades cluster, likely due to the fact that both doldrums and northerlies most frequently transition into this cluster
I don’t understand the logic of this statement
367: Lee et al. (2010), who attributed a central South Pacific marine heatwave in 2009-2010 to the overlying anticyclones in a Rossby-wave chain
This RW train in Lee’s study was associated with a strong central Pacific warming. Indeed you see a hint of tropical warming in Fig 4g. I wonder if this is the RW source. Would be interesting to extend the domain of that figure a little further to the north.
Figure 10: might be worth indicating the proportion of time spent in each of the clusters (e.g. by scaling the cluster box)
397: Why would low cloud dissipates due to the high humidity and low LHF?
423: with 3.5% of days exceeding the local bleaching threshold
This isn’t a useful number unless put into context.
425: During January-March when ocean temperatures are warmest, years with mass bleaching have on average nine more doldrums days than non-bleaching years
From Fig 8b I see a difference of 3 or 4, Im not sure where the 9 comes from, and you haven’t demonstrated if there is a statistically significant difference.
432: The most prominent mechanism in our analysis is Rossby-wave breaking
‘most prominent’ implies that you looked at other mechanisms