Point-to-point reply to reviewer A
The manuscript is essentially broken down into three separate chapters: (i)
Pacific, (ii) Atlantic and (iii) Indian. It seems quite clear to the reader that three
different authors took primary responsibility for writing these chapters. This is a
particularly glaring problem with respect to the third chapter. Here the authors
seem to be trying to "sell" the importance of the India Ocean Dipole mode (or
zonal mode) without fairly acknowledging any controversy or open questions.
Given the scope of the task, we deliberately divided the writing geographically
into three tropical ocean basins, so that the progress in each of these oceans
can be discussed in more details. In the revision, we try to make different
sections more connected and address the issue concerning the India Ocean
Recommendation 1: Please insert a little more balance regarding the ongoing
debate associated with the Indian Ocean Dipole mode. There are several
references missing in this regard. There is nothing wrong with highlighting the
debate; this is a new area of research and we do not have all the answers.
We believe that the previous version of our manuscript provided a
comprehensive list of unique studies on IOD by taking care of all significant
aspects of the topic. However following referee’s general remark and recent
rapid advances in the IOD studies (e.g., CLIVAR P1 session of IAPSO), we
have included a few more references and modified our discussions further.
Recommendation 2: All three chapters are rather lengthy - this is not a
problem, it merely reflects the wealth of literature. However, it would be helpful
to the reader if each chapter ended with some bullet points that highlight (i) the
new beyond-TOGA understanding and (ii) the key remaining questions or
We added a short summary at the end of each of the three main chapters to
highlight the major progress and the remaining issues.
Recommendation 3: The entire manuscript focuses on variability and there is
little discussion regarding prediction and predictability. It is too much to ask to
have a major discussion of these issues; nevertheless, the conclusion should
have some hints regarding the prospects of using this new understanding to
improve prediction and assess predictability.
The reviewer’s point is well taken. However, considering that there are other
papers (for example, Kurshnir et al. ) in the special volume that directly
address the predictability issues and the fact that the length of this paper is
becoming a real problem, we decide not to further expand the discussion of
prediction and predictability issues in this paper except for a few sentences
referring to the state of the art .
Recommendation 4: There needs another thorough read by one of the authors
to try to smooth out the different writing styles in each chapter.
We tried our best.
Recommendation 5: I have given the corresponding author marked copy of the
manuscript for language.
Point-to-point reply to reviewer B
Overall, this paper is a worthwhile endeavor and should be published. A
general criticism regards the need for this combined summary - a very long
paper - of the tropical variations in all oceans. In the current manuscript, the
regional chapters can basically stand alone, and could be published as three
separate papers without too much modification. The combined paper could be
strengthened by a more comparisons and contrasting of the basins, to
elucidate their common features and differences that lead to the distinct
characteristics of their variations.
We added some more comparisons and contrasting among the three basins.
For example, in the discussion of the Atlantic meridional mode, we point out
that a similar mode has been identified in the tropical Pacific. We also point out
that the seasonal variation of the Atlantic Ocean differs from the Pacific in that
the former exhibits a much stronger variation in its equatorial thermocline than
the latter. Furthermore, we contrasted the meridional heat transport between
the Atlantic and Pacific.
In the description of IOD, the authors go through great length convincing the
reader that the IOD is a climate mode independent from ENSO. However, the
authors never define what they mean by a 'mode' or a 'physical mode'. It would
help the discussion and appreciation of the physical nature of the IOD, if the
arguments could distinguish between atmospheric forcing of the IOD from
those that indicate a truly coupled mode. For example, the association of the
central Indian Ocean wind stress with changes of SST in the Indian Ocean and
anomalies of the thermocline dept could simply indicate that the anomalous
atmospheric state force the ocean. In contrast, an atmospheric response to
ocean anomalies and non-trivial predictability is indicative of a truly coupled
mode and should be emphasized.
Previous studies (e.g. Yamagata et al. 2003; Behera et al. 2003; Saji and
Yamagata 2003) have demonstrated the ocean-atmosphere coupled nature of
the IOD based on correlations among several oceanic and atmospheric
variables. Moreover, several recent studies documented the nature of coupled
Rossby waves during IOD in the equatorial Indian Ocean (e.g. Xie et al. 2003;
Yamagata et al. 2004; Rao and Behera 2005). The atmospheric response to
the SST anomalies related to IOD is also discussed in several modeling studies
(e.g. Ansell et al. 2000; Ashok et al. 2003, 2004; Guan et al. 2003). As pointed
out by the reviewer, we find a good predictability of the East African short rains
and also the equatorial wind anomalies based on the indices derived from SST
and heat content anomalies (e.g. Behera et al. 2005). These are further
emphasized in the revised manuscript to support the IOD as a physical mode
just like the Pacific ENSO.
Below is a list of questions and recommendations. None of the issues are
major. I recommend publication after the definition of the IOD as a 'mode' and
the clarifications listed below are addressed or refuted.
1. Page 9, 11, 'first order': I recommend using different words. 'First order'
implies an expansion in some small parameter. Its use should be accompanied
by the explicit discussion of this parameter.
2. page 10, center paragraph: Please explain how 'an increase in the trades in
the far western Pacific one to two seasons prior to the onset of ENSO' is
consistent with the Bjerknes hypothesis.
This is because it takes a thermocline response to change in the trades 1-2
months to propagate from western to eastern equatorial Pacific and then it
takes another 1-2 months for the thermocline change to affect the SST.
3. page 10/11: This is a very long paragraph that is difficult to read. Simply
breaking the text into paragraphs after ' that is so important to ENSO' (page 10),
'.. primarily resides in the ocean' (page 11), ' ... to the dynamical adjustment
time of the Pacific.', and '.. oscillations in the ENSO paradigm.'.
The long paragraph has been broken into four new paragraphs:
ENSO events … resides in the ocean.
The free oceanic Kelvin … adjustment time of the tropical Pacific Ocean.
The evolution of the coupled mode … in the ENSO paradim.
Mathematically, … a viable oscillating system.
4. page 12, paragraph 2: please discuss the 'free parameter of the system' that
determines the time scale of the recharge oscillator. Is it the residence time
(volume/flux) of the warm water sphere?
The free parameter determines the time it takes to recharge the off-equatorial
heat content in response to changes in off-equatorial wind-stress curl.
5. page 12/13: Please clarify the 'alternative views of ... theories of ENSO'.
Neither the advection mechanism nor the westerly wind burst seek to
determine the time scale of ENSO or its recurrence. Instead, these are
important theories regarding the dynamics, or forcing of the anomalies.
We now mention explicitly that these alternative views do not seek to determine
the time scale of ENSO or its recurrence.
6. Section 2b: It might help to introduce subheadings to help clarify the different
hypotheses for ENSO decadal variability such as stochastic forcing, changes of
the mean state, anthropogenic forcing, ENSO nonlinearity etc.
Subheadings have been added.
7. page 18, discussion of Kirtman et al. (2004): What should the reader
conclude from the finding that in certain areas stochastic theory can not explain
the variability? Does this imply that globally stochastic forcing can be
excluded? Please clarify.
Kirtman et al’s result implies that in western tropical Pacific the ocean-
atmosphere feedback is so strong that nonlinearity must be taken into
8. page 19: Please explain the 'pure kinematic argument' of Schopf (2004).
In Schopf and Burgman’s mechanism, if El Nino can be viewed as an oscillating
front between warm and cold water, an increase in the excursion of the front
(the "strengthening" of El Nino) results in a change to the Eulerian time mean.
But at any single instant, the state is characterized by the same front separating
the same warm and cold water.
9. Page 19: Introduce new heading such as 'Observational evidence'
10. page 22: Rothstein et al. 1998 is not an observational study of the STC.
Please correct citation.
Correction is made. Now Rothstein et al. (1998) is cited in the complex ocean
circulation model studies along with Harper, 2000, Huang and Liu, 1999 …
11. page 25: Please explain how Boccaletti et al. (2004) framed the problem,
i.e. what do they do that is 'different'.
Boccaletti et al. (2004) frame the problem differently by asking the question
how the heat transport by STC can work as a constraint for the equatorial
12. page 26: editorial: the details and mechanism ... >are< not well understood.
13. Page 26: The discussion of the tilting nature of the western boundary
current is not clearly related to the earlier paragraphs and the theme of the
section on Pacific ENSO and decadal changes of ENSO. Please clarify or
While this point may seem a detail in our understanding of the boundary current,
the theory of McCreary and Lu relies upon separation of equatorward and
poleward flow at the western boundary. Since the subsurface limb of the STC
is the important component of the flow for this theory, and since this boundary
can change by 5° of latitude, a more complete understanding of this behavior
seems in order.
Tropical Atlantic Variability
14. The chapter on the Atlantic variability could benefit from more subheadings.
Subheadings are added.
15. page 28: please clarify the 'dipolar pattern across the thermal equator' by
the location of the centers of action. Fig. 6 suggests that the negative
anomalies at the far southern border of the plot form the negative pole. Is this
correct? If so, why not expand the plot further south to better show the
southward extension of the negative pole?
Here, the dipolar pattern is referred to the precip. anomaly (black contours).
The negative anomaly at the far southern border is the SST anomaly regressed
onto the PC time series of the precip. Therefore, the location of the centers of
action for the precip. dipole is within the deep tropics.
16. page 29: The anomaly pattern of precipitation depicted in Fig 7 (right) has
maximum anomalies in the western part of the Atlantic off Brazil. In contrast the
text states that the anomaly patten is max. along the northern coast of the Gulf
of Guinea. Please clarify.
The two panels were mistakenly switched. The discussion in the text is meant
for the left not the right panel. This is now fixed.
17. page 29: The discussion of the Giannini et al (2003) papers disrupts the
flow of the section. Please clarify the connection ('Incidentally' betrays a lack of
connection). The discussion of anomalies off the western coast of African south
of the equator does not fit well. What is the purpose of this brief section?
We have reworded the sentence. It now reads: “It is also important to point out
the distinction between the rainfall pattern associated with this coupled mode
and rainfall variability over …”. The purpose of the brief discussion on
Benguela Ninos/Ninas is to point out the fact that this phenomenon is not
necessarily associated with the Atlantic zonal mode. Given some recent
interests in this phenomenon, we feel the need to include this brief discussion
to make the review more complete.
18. Figure 8: Please provide a more thorough figure caption. What is the Atl-3
index, what are the sources and characteristics of the data?
The caption has been rewritten, which now includes the definition of Atl-3 and
19. Figure 9: Please expand the caption: what is the region of the zonal
average? What is the x-axis, what does ENSO(+) mean, what are the SST
The caption has been rewritten, which now includes the definition of ENSO(+)
and NAO (-) and other information about the plot.
20. Page 34: what are the 'important consequences on seasonal predictability'?
During the years when the existing Atlantic conditions reinforce the remote
influence of ENSO (constructive interference), the boreal spring climate
anomaly can be forecasted skillfully up to two seasons in advance.
21. Page 34: what do the cited investigations of the role the Southern
Hemisphere find ('begun to investigate the issue')?
The initial finding suggests that the southern summer atmospheric variability
(and to a less extent the winter variability) can play a pre-conditioning role in
the onset of the inter-hemispheric anomalies in the deep tropics during the
following austral fall.
22. page 40, editorial: replace 'proceeded' with 'precedes'.
23. page 47/48: Amplification of the extratropical signal in the thermocline:
What is the evidence that local air-sea feedback amplify the extra-tropical
signal? Why does this feedback only act and amplify signals that originated
from the extratropics, and not on perturbation generated in the tropics?
There is no such evidence at the moment.
Indian Ocean Dipole (IOD)
24. page 51, second paragraph: Please clarify the following incomplete
sentence: '.. strong positive IOD events of 1961, 1967 and 1994, independent
nature of the IOD is also seen in pure composites ...'
The sentence is rewritten.
25. Page 51: define 'physical mode'. How can the IOD be distinguished from
intrinsic atmospheric variability that causes changes in the ocean through
cooling by the turbulent heat flux, changes of upwelling off Sumatra, and
adjustments of the equatorial Indian Ocean thermocline?
The IOD signal is distinguished in both oceanic and atmospheric variables for
more than a couple of seasons. It is unlikely that the intrinsic variability in the
atmosphere can be sustained for such a long period of time in coherence with
the oceanic conditions. The only possibility is the case in which the wind
anomalies in the Indian Ocean are remotely generated by the ENSO. Based on
recent coupled modeling studies, we have discussed the unlikelihood of such
remote forcing; the IOD events occur in the tropical Indian Ocean even in
absence of ENSO variability in the Pacific in those coupled model experiments.
Moreover, the heat content anomalies and SST anomalies have shown
significant correlation at one month lead not only to the short rains over East
Africa but also with the equatorial zonal winds crucial for maintaining the
thermocline tilt. Furthermore, the ocean conditions can not be sustained with
out a coupling to the atmosphere in the quick response regime of tropical
26. Page 55: please describe the model results of Baquero-Bernal et al. (2002)
27. Pagae 58: reference to Figure 14: ' Fig 14c .. cold SST anomaly near the
Java coast propagate along the west coast of Australia'. Figure 14 does not
show the west coast of Australia. Please clarify figure or statement.
The statement is modified.
28. page 59: Editorial: correct 'telelconnections'
29. page 60: What determines the one season lag between occurrence of the
IOD and extreme African short rains?
It is determined by the seasonality of the IOD evolution and African short rains
in relation to the fall transition of the monsoon. The short rains season
coincides the peak season of the IOD: The western pole of the IOD locates
close to East Africa during boreal fall.
30. page 64, bottom: editorial: fades >a<way.
31. page 67: bottom-5 lines: correct 'mixing'
32. page 71: explain 'northern escape gateway'
The northern escape gateway refers to the North Brazil Current/North
Equatorial Counter-Current System. During the boreal spring, the NECC is
absent and the NBC flows northward. This causes the warm tropical water to