the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 09 Oct 2023
Eccentricity forcing on Tropical Ocean Seasonality
Luc Beaufort
Anta Clarisse Sarr
Abstract. The amount of radiative energy received at the Earth's surface depends on two factors: Earth-Sun distance and sunlight angle. Because of the former factor, high eccentricity cycles can induce the appearance of seasons in the tropical ocean. In this paper, we use the Earth System model IPSL-CM5A2 to investigate the response of the low-latitude oceans to variations in Earth's orbital eccentricity. Sea Surface Temperature (SST) and Primary Production (PP) were simulated under six precession configurations at high eccentricity and two configurations with low eccentricity, representing extreme configurations observed over the past million years. Results show that high eccentricity leads to increased seasonality in SST, with an annual thermal amplitude of approximately 2.2 °C in low latitude ocean surface waters (vs. 0.5 °C at low eccentricity). PP, which already exhibits inherent seasonality under low eccentricity conditions, sees its seasonality largely increased under high eccentricity. As a consequence, we show that on long time scales the intensity of SST seasonality exhibits only the eccentricity frequency, whereas that of PP additionally follows precession dynamics. Furthermore, the seasonal variations in both SST and PP at high eccentricities are influenced by the annual placement of perihelion with its direct impact of radiative energy received in tropical regions. This leads to a gradual and consistent transition of seasons within the calendar. We introduce the concept of "eccentriseasons," referring to distinct annual thermal differences observed in tropical oceans under high eccentricity conditions, which shift gradually throughout the calendar year. These findings have implications for understanding low latitude climate phenomena such as El Niño-Southern Oscillation and monsoons in the past.
- Preprint
(12422 KB) - Metadata XML
- BibTeX
- EndNote
Luc Beaufort and Anta Clarisse Sarr
Status: open (until 29 Dec 2023)
-
RC1: 'Comment on cp-2023-80', Anonymous Referee #1, 12 Nov 2023
reply
The authors examine the seasonality of low latitude sea surface temperature (SST) and primary productivity (PP) in a set of high-eccentricity simulations spanning various longitudes of perihelion. They find that high eccentricity leads to increased seasonality in SST as well as in PP. As a result, they find that the phasing of the SST and PP seasons alter with the timing of perihelion, leading them to introduce the concept of ‘eccentriseasons’ where the seasons do not stay constant with the calendar year.
This study is a timely and useful addition in the literature, and highlights the important role played by eccentricity in the tropical ocean seasonal cycle and how the phasing of the seasons relative to the calendar change as a result of precessional orbital changes. The results with SST are perhaps not as surprising given that it ties directly to the tropical insolation seasonality - Earth’s axial tilt mainly provides a semiannual cycle of insolation near the equator and little in the form of an annual cycle, and so eccentricity provides the larger annual cycle forcing and especially at high eccentricity. I have no expertise with biology, but I surmise that the seasonal control of tropical PP is less well understood and so the results there is presumably novel.
My main critique is that the manuscript is somewhat hard to read because there are many different threads of argument, and that there is a vagueness to some of the physical linkages being made. The writing needs to be improved. The manuscript would benefit from simplification, and that the connections be made more precisely.
Specific comments
- I strongly recommend the authors run a simulation with zero eccentricity if at all possible, and not use the low eccentricity simulations.Using low-eccentricity simulations as a control to compare against the high eccentricity simulations complicates things: you can’t unambiguously separate out the contributions coming from orbital eccentricity and Earth’s axial tilt, and moreover you have to consider the effects of different longitudes of perihelion in the low eccentricity case. A zero-eccentricity simulation solves both these problems and makes the analysis cleaner and simpler.
- To make the connection between insolation and tropical SST seasonality more explicit, I recommend that the authors include a figure of TOA insolation averaged over the tropical latitudes for the high eccentricity and low eccentricity cases (or zero eccentricity case, assuming you do #1), and show the difference between them to reveal the magnitude and phasing of the annual insolation coming from eccentricity.The link between the insolation and tropical SST should be discussed.
- The authors average over the 6 eccentricity cases in their evaluation of the seasonal amplitude for figure 3 and 4 (see lines 118-121). However, Erb et al. (2015) showed that their ‘AE’ simulation (perihelion at autumnal equinox) has a significantly smaller seasonal amplitude in the Pacific cold tongue than the other cases (WS, VE, SS - see their figure 3e-h) and also compared to their zero eccentricity simulation (see figure 6c). Assuming that the IPSL model also shows similar behavior, this example demonstrates a problem with averaging over the 6 high eccentricity cases to evaluate the annual amplitudes, as it hides a lot of regional amplitude variation that can occur when the timing of perihelion is varied. Also, what if the behavior exhibited by the cold tongue occurs in other regional tropical oceans? Related to this, the amplitude of seasonality in PP (Figure 7) clearly has dependence on the longitude of perihelion in various regional oceans, making me question the wisdom of taking the average in the annual amplitude across the 6 eccentricity cases. The assumption of averaging over the 6 eccentricity cases needs more elaboration and justification if you want to keep it, but I’m wondering if the story can be made simpler and more precise if you do away with it.
- Related to point 3, part of the vagueness is that the authors analyze both regional tropical SST and tropical (area) mean SST, but the wording in the manuscript often does not clearly separate the conclusions between the two.This wording needs to be more precise. My understanding is that the conclusions (as stated in the abstract) applies to tropical (area) mean SST, and it should be stated as such. Those conclusions do not necessarily hold for specific regional SST where there can be marked differences in behavior, in particular over the Pacific cold tongue region (see point 3). This comment applies also for the analysis of PP, where there are clearly marked regional differences.
- A claimed novelty of this manuscript is that the simulations done in this study – covering a complete revolution of precession – allows for revealing the gradual transition of the tropical ocean seasons within the calendar year (section 4.2.2). However, both Erb et al. (2015) and Chiang et al. (2022) undertook simulations to cover a complete revolution of precession, and both noted a gradual shift of the seasons relative to the calendar with the Pacific cold tongue, and this gradual shift was a central focus of both these studies. In this aspect, this study and its revelation of gradual seasonal phase changes is not entirely novel, and credit should be given where it is due.
- Most sections appear to consist of only one paragraph, and as a consequence some paragraphs in this manuscript are quite long and cover many points (for example the Introduction).The long paragraphs make the manuscript difficult to read. I would suggest breaking them up into paragraphs, each paragraph covering one major point.
Technical comments
Line 2 – suggest rewriting as “Because of the former, ….”
Line 6 – ‘increased seasonality in SST’ – do you mean tropical mean SST here? Please also check other instances to clearly define the regional scope of the SST
Line 7-8 “PP, which already exhibits inherent seasonality under low eccentricity conditions, sees its seasonality largely increased under high eccentricity. “ Similarly, is the tropical mean PP? Please also check other instances to clearly define the regional scope of the SST
Line 26 – just a note: the tropics experiences seasonality through rainfall, in particular monsoon climates. It is no accident that the word ‘monsoon’ comes from the Arabic word for ‘season’
Line 62 – instead of ‘perihelion position’, suggest using ‘longitude of perihelion’ to be more precise.
Line 64 – suggest remove ‘voluntarily’
Line 95-99 – I note that a calendar correction for orbital changes is not done but I agree that it won’t affect your main conclusions. However, in figures where you report timing (e.g. figure 5), it would be useful to remind the reader of this fact in the caption. Related to this: “the accuracy is scaled at a few days (its inherent precision)”. Can you be more specific about the accuracy, and how is this calculated? Is it 1-2 days, or 10 days? Also, I would rephrase this sentence as “the accuracy with regards to the calendar timing of results is accurate to within +- X days”
Line 118: just to be clear – you first evaluate the annual amplitude for each eccentricity case, and then average, correct?
Line 119: “potential loss of certain interhemispheric dynamics “ I don’t understand what this means – can you elaborate? Also as mentioned in point 4 above, this averaging isn’t valid for the Pacific cold tongue, which is big part of the tropical oceans.
Line 121-122 – how is this correlation done? Are you only comparing annual means?
Figure 3 and section 3.1.1 – there are notable differences between the model annual mean with observed, so saying there is consistency between model and observed is not a convincing statement. However, the main point you are making here is that the annual mean is not much affected by eccentricity. Also, it would be helpful to show a difference plot for the annual mean between the low and high eccentricity cases.
Figure 3 and 4 – I suggest making one figure for annual means only (so the left panels of figure 3) and include a difference plot between low and high eccentricity. Then, make another figure for amplitudes (i.e. right panels of figure 3) and include the difference plot (figure 4) here.
Figure 4 – with the difference plot, can you assess the statistical significance of the difference?
Figure 5 –these are area averages across the longitudes entire tropics 30S-30N, and just SST or both SST and land temp? Please state more clearly in the caption.
Line 150 – since you’re only averaging over SST, it doesn’t really include monsoon regions.
Line 156-158 – just a comment - the seasonal cycle for low eccentricity is dominated by the semiannual cycle derived from axial tilt.
Line 161-163 and figure 6 – again, I see marked differences between the simulated and observed PP means, and the differences are not small in my opinion; but I think the main point you want to make is that the annual means are similar between low and high eccentricity.
Figure 6 – as with the SST plot, I suggest breaking up into two figures, one for the annual mean, one for the annual amplitude; and in both cases, plot the difference between low and high eccentricity. For figure 7, you can just plot the individual eccentricity cases.
Line 190 – “annual wobbling of the Earth’s axis”. Do you mean the effect of precession? Or do you mean the Earth’s axial tilt?
Line 194-195 and figure 10. Since you are comparing the seasonal cycles, should you be removing the annual mean for each timeseries before plotting? As a next step, if you then subtract each high eccentricity seasonal cycle from the low eccentricity (take the mean, say), do you get an annual cycle as the difference? This annual cycle can be attributed to eccentricity, and phase difference to the longitude of perihelion.
Line 204: “We show that the seasonality in the tropical ocean becomes more pronounced”. Again, do you mean for the tropical area-averaged ocean?
Line 220 – I’m a little confused by the equation, and it needs more explanation. In particular, shouldn’t the denominator be (0.053 – 0.005) to account for the eccentricity in the low eccentricity simulation? Also the equation should be in its own line, and numbered.
Citation: https://doi.org/10.5194/cp-2023-80-RC1
Luc Beaufort and Anta Clarisse Sarr
Luc Beaufort and Anta Clarisse Sarr
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 202 | 75 | 4 | 281 | 2 | 2 |
- HTML: 202
- PDF: 75
- XML: 4
- Total: 281
- BibTeX: 2
- EndNote: 2
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1