Teleconnections: A Technical Discussion - Page 6 - Forecasting and Discussion - 33andrain Weather Discussion Community Jump to content
Snowy Hibbo

Teleconnections: A Technical Discussion

Snowy Hibbo

Welcome to the Teleconnections thread!

Check out some of the other driver threads and resources here:

Arctic thread

Stratosphere thread

Weather research portal

Message added by Snowy Hibbo

Recommended Posts

23 hours ago, Bring Back 1962-63 said:



The latest monthly chart for September with the August chart below for a comparison. 



Little change, which is somewhat surprising - just delayed by a month or so. In fact no models are -ve now. Actually if you focus closely, there is a slight shift downwards on the strength of the predicted upcoming El Nino. The statistical average of all the models (the thick green line) just about manages to get to an anomaly of +0.5c average over the fall (SON) whereas that was nearer +0.65c in the August prediction. It gets to its highest level of around +0.8c during the winter (DJF) before it falls back slowly during the first half of 2019. The August chart climbed to just shy of +1.0c (perhaps +0.95c) for rather longer and only fell back to around +0.7c.  So we have a slight delay in getting there and a very weak El Nino going forward. 4 models never get there at all - just hovering below 0.5c and then falling back more quickly. Fewer models get above the +1.0c anomaly. 



Rather than me trying to estimate the exact figures from the charts, here are the tables:



Discussion of Current Forecasts

Most of the models in the set of dynamical and statistical model predictions issued during mid-August 2018 indicate neutral ENSO conditions lasting through summer, shifting to weak El Niño starting with the Aug-Oct season and nearing moderate strength during winter 2018-19. In the most recent week, the SST anomaly in the Nino3.4 region was 0.2 C, in the neutral range, and 0.30 C for the month of July, also at a neutral level. All of the key atmospheric variables now reflect neutral conditions, although in the most recent week low-level westerly wind anomalies have developed. The subsurface sea temperature anomalies continue to be moderately positive. Between 80% and 85% of the dynamical and statistical models predict a warming to El Niño levels by the end of the year, and this outlook is now accepted more by forecasters since the spring predictability barrier is passed. Based on the multi-model mean prediction, and the expected skill of the models by start time and lead time, the probabilities (X100) for La Niña, neutral and El Niño conditions (using -0.5C and 0.5C thresholds) over the coming 9 seasons are:



Meanwhile, the very short term conditions are moving in the precisely the opposite direction! 


The -ve SST anomalies in the sub tropical south Pacific are back with a vengeance. They are pushing right up to the equator - not just pale blues but some darker colours now showing up.


The main Nino region used as the yardstick for measuring the ENSO base state has SST anomalies struggling around neutral.


The west CP is cooling off too.


The east CP is further below 0c.


Even the EPAC which took a very brief excursion above 0.5c for the first time in 2 years has now headed back below 0c.


Right now the only thing that I can safely say with any confidence is that things are very uncertain ?


The "Discussion of Current Forecasts" states that low level westerly wind bursts have developed which would be a sign of "Nino-like" conditions developing.  I image that our knowledgeable AAM (Atmospheric Angular Momentum) and GWO (Global Wind Oscillation) members will be updating us on the very latest position there (I'm still learning more in this vitally important area) - that is Tams, Tom, Zac and James @Tamara  @Isotherm @Snowy Hibbo and @Singularity.  We have so much more to discuss in our ENSO debate.  Will this set back prove to be temporary or will it start to confound the model consensus.  We shall try to keep up-to-date and even slightly ahead of the curve on this thread. 


@Andrew Maddis I noted your comments above and a warm welcome to our debate.  I copy your comments below:


"The summer had a lot of similarities to the 1983 summer, which was a developing NINA. My number 1 analog right now is 1983-84. Given how ENSO has been this isn't a bad analog."


I'm not really sure if using analogues will be particularly helpful in predicting the ENSO outcome this time around, given the larger than normal uncertainties and all the variables to be considered. It would very unusual to go from one Nina, to only "Nino-like" conditions and straight back into another Nina phase. In fact if you study the table below carefully,  you'll see that it has never happened since 1950.  Every Nina is eventually followed by a Nino and vica versa.  Full episodes have weakened and strengthened again at times but never going to the opposite side of neutral unless it gets to the other ENSO phase. That's not to say that it cannot happen, we could have a period of prolonged Nino-like or Nina-like conditions but without ever meeting the criteria for a full on event.  The phase process is a sort of balancing one between the two phases.  The measurements are really pretty arbitrary and the established criteria to meet the definition are far arbitrary!


 The NOAA weekly reports include the ENSO phase table but this only goes back to 2006.  Less often shown is the full table back to 1950 - here it is:


If we look at the table, 1983 saw a pretty strong El Nino episode that switched quite quickly to a long La Nina episode, albeit broken by a period of weaker -ve anomalies. That is not to say that your comments are invalid. Perhaps you can elaborate on some of the conditions that prevailed then and why you feel that this is your number one analogue - are there particular causation factors that you have identified?  I will be interested to hear your comments and please get involved with our discussions.  :) 


@mbaer1970 re: your IOD/ENSO post - I have placed several papers in the portal and have several more to add tonight and tomorrow morning. I'm sorry that it has been taking longer than expected to respond but only the first part of my review post is ready.  I should be less busy with my online business tomorrow and hope to get it finished later on Thursday. As you can see, you have already seen short posts on this from Tom @Isotherm  and Zac @Snowy Hibbo with rather differing comments. The papers that I'll be reviewing also throw up some quite divided opinions.  This is a great part of our debate.  Having different opinions or perspectives helps us to examine more of the details and the reasons for that array of views.  David :) 

Summer 2018 and Summer 1983 had these similarities...and they are noteworthy, you cannot just ignore them. I am talking about the NYC Metro area but you will find similarities elsewhere as well.


1. The +NAO for JJA for 2018 beat the previous highest +NAO summer of 1983. The AUG 2018 NAO was 1.97 while it was 1.61 in 1983.


2. Summer 2018 was the first time since the summer of 1983 were the heat and rainfall combined were well above is very rare to get a summer here that is hot and wet like we had.


3. The Heat Ridge was centered over Eastern Canada both summers.


4. Dewpoint readings of both 70 and 75 degrees were broken during the summer of 2018 by a large margin from the previous record holder which guessed it, the Summer of 1983. 


Of course there were differences as well in the QBO and with ENSO, but similar weather still prevailed this summer and I believe it continues throughout the Autumn and into The Early winter at least. I am not saying things will be exactly like the 1983-84 winter which oddly enough was similar to last winter in some ways. 


I believe that we will have excellent blocking this winter..........the chances of it NOT happening are very slim. 


Heat Ridges over Eastern Canada during the summer translate to Cold 1040HP during the winter. 

  • Like 2

Share this post

Link to post
Share on other sites
7 hours ago, mbaer1970 said:

This thread has really got my brain working and every new entry keeps the wheels rolling. 

Could the Indo-Pacific Warm Pool and the IOD be considered extensions of ENSO? I apologize if this has been spoken on already.

I'd say it's debatable either way. I think there is a entrenched connection to the SOI/ENSO, that makes the IOD fairly interconnected to it's attributes IMO. Because of this, it's impacts are not dissimilar from those of ENSO, except that it can portray them upon the regions that surround it, and factors that run through it (like the MJO, AAO, etc). That's just my observation, and this opinion doesn't rely on any research as of yet (but it is supported by various datasets, which I said above can often be conflicting). I will have a look at the papers from @Bring Back 1962-63 in the Tele portal later today.

  • Like 3

Share this post

Link to post
Share on other sites



Well since @mbaer1970 posted his queries on here on this topic, we have had responses from both Tom @Isotherm and Zac @Snowy Hibbo with rather different takes on the relative significance of the IOD (Indian Ocean Dople).  I have now placed some of my papers and presentations from my store into the Research Portal on this subject but I still have some more to add. So, I'll split this into several part posts to spread out my work load and to shorten them. 


Firstly a reminder of the points raised and the query from "mbaer1970":


If I remember correctly a positive phase of the Indian Ocean Dipole equates to El Nino conditions and a negative phase to La Nina conditions. The current Aussie BOM trend for the IOD has it moving into neutral territory during late November/early December and continuing a dive towards 0 in Feb. If the IOD connection to the ENSO state is correct, then we should start to see a trend in regions 4.0 & 3.4 moving towards neutral states as we approach winter. I am not sure if 3.0 will respond in kind, but its NDJF trend is 1.075. I am a bit confused at this Aussie BOM IOD trend considering the current strong negative state of the QBO, could this be a sign that we may be about to see the QBO begin to rapidly rise towards a positive westerly state. I can't say that I know enough to speak a neutral QBO, is that even possible. If this ENSO to IOD connection is valid, the El Nino Modoki trend could flip to a Neutral ENSO state. I also wonder how this IOD trend could impact trends and states of North American teleconnections like the PNA, EPO, and AMO all which seem to be trending towards states that would allow for negative AO and NAO setups.  


While I do not propose to provide all the answers, I do hope to demonstrate that Tom's and Zac's diverse views are reflected across many of the research papers extending over nearly 20 years. Once I have completed my reviews, I would imagine that we can pick up on some of the conclusions and views expressed and assess them on here. Like many on here, I always try to take a balanced view with an open mind. My general view is that the ENSO is a major teleconnection which interacts with others in ways that we only partly understand. There are numerous variable factors and forces which can come into play. Some of these conform to 

certain patterns in relation to particular phases of the ENSO cycle while others seem to behave much more randomly.  No two El Nino, La Nina or neutral phases are identical.  Sometimes another specific teleconnection might be quite dominant but at other times it hardly plays a part. Understanding which are the key factors at play and how they might interact is essential.  For example, I did another paper review post about a week ago on here, looking into the PMM, NPMM and the SPMM (the Pacific Meridional Mode and its northern and southern ocean components).  It's only in more recent years that the importance of the role of the sub tropical south Pacific has started to be appreciated.  How this interacts with other factors is key and it would appear to have at least some influence on the strength, timing and type of developing El Nino events.  Too much research seems to attempt to isolate a  particular factor and focus on how that influences the ENSO and draw specific conclusions but without seriously taking account of other factors.  This can sometimes undermine the legitimacy of that research which might otherwise contain some very useful pointers. 


We also have the cause and effect dilemma - which teleconnection is influencing the other and which is being influenced by the other. Sometimes there might even be some sort of feedback loop in place where one leads to a change in the other which then goes on to influence the first one. On some occasions one might be the dominant factor but at others it might be dominated.  This might well be the case with some of the oscillations as well as some of the longer term cycles which can be in very different phases to the ENSO episodes. With all that in mind, I shall proceed with caution.


I should say right away, that there are far more papers dealing with the ENSO influences on the IOD and in turn the IOD impacts on regional climates, the Indian and Asian Monsoons and on the Australian climate and only a smaller number that consider how the IOD might have some impacts on ENSO. I shall review some of those papers in part 2 of my post. 


I'm aware that some readers who might wish to follow this debate, might like to learn about or brush up on their knowledge on exactly what the IOD is.  So in this part I'll deal with the basics.  Just click on the titles for a link to our Research Portal entry for each one. From there you'll find a link to the paper or presentation.


1. Indian Ocean Dipole, Meaning, Definition and Explanation - YouTube Presentation  - a concise 5 minute audio presentation, explaining the simple terms and the IOD basics.


2. Understanding the IOD - YouTube Presentation  - an excellent 4 minute video presentation from the Australian Bureau of Meteorology.  


3. El Nino, El Nino Modoki, La Nina, Indian Ocean Dipole - YouTube Presentation  - this is a brilliant 24 minute video presentation from 2015 and one of best that I have seen. The presenter explains all the processes, circulation patterns ocean currents involved for El Nino and El Nino Modoki, La Nina, the less widely known La Nina Modoki, the SOI (southern oscillation index) and with the emphasis on the IOD. The walker circulation is beautifully demonstrated. The whole presentation contains some very clear charts and for those wishing to learn the basics and slightly beyond that, i strongly recommend this.  Apart from a link to the video, there is another link to the slides and charts with clearly written notes.  The extract below is from the part dealing with the IOD:


Indian Ocean Dipole effect (not every El Nino year is the same in India)

  • Although ENSO was statistically effective in explaining several past droughts in India, in the recent decades the ENSO-Monsoon relationship seemed to weaken in the Indian subcontinent. For e.g. the 1997, strong ENSO failed to cause drought in India.
  • However, it was later discovered that just like ENSO was an event in the Pacific Ocean, a similar seesaw ocean-atmosphere system in the Indian Ocean was also at play. It was discovered in 1999 and named the Indian Ocean Dipole (IOD).
  • The Indian Ocean Dipole (IOD) is defined by the difference in sea surface temperature between two areas (or poles, hence a dipole) – a western pole in the Arabian Sea (western Indian Ocean) and an eastern pole in the eastern Indian Ocean south of Indonesia.
  • IOD develops in the equatorial region of Indian Ocean from April to May peaking in October.
  • With a positive IOD winds over the Indian Ocean blow from east to west (from Bay of Bengal towards Arabian Sea). This results in the Arabian Sea (western Indian Ocean near African Coast) being much warmer and eastern Indian Ocean around Indonesia becoming colder and dry.
  • In the negative dipole year (negative IOD), reverse happens making Indonesia much warmer and rainier.


  • It was demonstrated that a positive IOD index often negated the effect of ENSO, resulting in increased Monsoon rains in several ENSO years like the 1983, 1994 and 1997.
  • Further, it was shown that the two poles of the IOD – the eastern pole (around Indonesia) and the western pole (off the African coast) were independently and cumulatively affecting the quantity of rains for the Monsoon in the Indian subcontinent.
  • Similar to ENSO, the atmospheric component of the IOD was later discovered and named as Equatorial Indian Ocean Oscillation [EQUINOO][Oscillation of warm water and atmospheric pressure between Bay of Bengal and Arabian Sea].

Impact on IOD on Cyclonogeneis in Northern Indian Ocean

  • Positive IOD (Arabian Sea warmer than Bay of Bengal) results in more cyclones than usual in Arabian Sea.
  • Negative IOD results in stronger than usual cyclonogenesis (Formation of Tropical Cyclones) in Bay of Bengal. Cyclonogenesis in Arabian Sea is suppressed.


The red colour shows the regions with warmer, moister air and the clouds show the areas of greatest tropical convection and rainfall.  The blue colour shows the cooler, drier regions.



Although we see +ve and -ve phases of the IOD there are rare occasions when both the west and east regions are -ve.  The chart above shows this (east and west of India) with the warmer waters further south which coincided with a full El Nino episode. This shows that there are sub-variants with different impacts - it is far from straight froward and I'll try to make more sense of it in part 2.


I'll leave it there for now.  David :)  

  • Like 2

Share this post

Link to post
Share on other sites



For my introductory comments plus learning guides, please refer to part 1. For a review of a couple of early papers, please see part 2. In this part, I move on to more recent papers and presentations. For each paper, just click on the title for a link to the Research Portal entry where there is an abstract and a link to the full paper or presentation:


6. Interaction Between the Indian Ocean Dipole and ENSO Associated with Ocean Subsurface Variability - Presentation   There are links to the paper and also the presentation slides and charts.


This fascinating workshop presentation was delivered at NOAA's 41st Climate Diagnostics and Prediction Workshop,  held in Orono, Maine, from 3rd to 6th October, 2016



The Indian Ocean dipole (IOD) is an intrinsic coupled mode of variability in the tropical Indian Ocean. It has broad impacts on regional climate. An IOD index is defined as the difference between sea surface temperature (SST) anomalies averaged over the western Indian Ocean (WIO, 50°–70°E, 10°S– 10°N) and eastern Indian Ocean (EIO, 90°– 110°E, 10°S–Eq.). An important issue in the studies of IOD is the relationship between IOD and the El Niño-Southern Oscillation (ENSO) and the potential feedbacks from each other. Previous studies have shown that the development of IOD can be independent of ENSO, but ENSO may exert significant influence. In recent years, it has also been found that IOD can affect ENSO. Clearly, there exists an intimate interaction between IOD and ENSO but the detailed phenology of their mutual evolution has not been reported thus far.

Although a positive (negative) IOD tends to co-occur with El Niño (La Niña), the spatial-temporal covariations of these two major climate modes have not been well documented. Our earlier modeling study (Wang et al. 2016) documented the time evolution of IOD and the associated ocean subsurface variability in the absence of ENSO. The current study is aimed at examining the time evolution of IOD in the presence of ENSO and characterizing the interaction between IOD and ENSO. The present work complements our previous analysis by looking at the spatial-temporal covariations between IOD and ENSO, identifying any leadlag relationships between them, and quantifying the influence of ENSO on IOD. This is done by analyzing a 500-year long fully coupled model simulation, which retains the ENSO variability (referred to as ENSO run hereafter), and comparing the results with a parallel 500-year simulation with the ENSO variability suppressed (daily SST nudged to its climatology in the tropical Pacific; referred to as no-ENSO run hereafter). The latter was analyzed in Wang et al. (2016) to characterize the spatial-temporal evolution of IOD in the absence of ENSO. The differences in the characteristics of IOD between the two simulations will indicate the impact of ENSO on IOD. Both simulations were conducted with the NCEP CFS version 1 coupled model.


The analysis considers the ENSO impacts on the IOD as well as the influence of the IOD on ENSO prediction.  I move on to the conclusions.



The interaction between IOD and ENSO is examined using coupled global climate model simulations. The covariability of IOD and ENSO is analyzed by applying the EEOF*** method to the surface and subsurface ocean temperatures in the tropical Indian Ocean and western Pacific. The first EEOF mode shows the evolution of IOD that lags ENSO, whereas the second mode exhibits the transition from a dipole mode to a basin-wide mode in the tropical Indian Ocean that leads ENSO. Both modes have high loadings in the tropical ocean subsurface. The lead-lag relationships between IOD and ENSO suggest a two-way interaction between them. A comparison between two 500-year model simulations with and without ENSO suggests that ENSO can enhance the variability of IOD at interannual time scale. The influence of ENSO on the IOD intensity is larger for the eastern pole than for the western pole, and is stronger in a negative IOD phase than in a positive phase. The influence of IOD on ENSO is demonstrated by the improvement of ENSO prediction with a linear regression forecast model when considering SST in the western pole as an ENSO precursor. The improvement of the ENSO forecast skill is found not only at a short lead time (0 month) but also at long leads Fig. 4 Anomaly correlation skills of CFSv2 dynamical forecast (black) and statistical forecasts using one predictor (blue for WWV; red for WIO SST) and two predictors (WWV + WIO SST, green) for DJF Niño 3.4 SST with lead times from 22 months to 0 month, corresponding to forecasts made from January of previous year to November of current year (Jan to Nov0 with  and 0 for previous year and current year, respectively). Solid (dash) gray line denotes the threshold of the anomaly correlation at the 99% (95%) significance level. WANG ET AL. 5 (10-15 months). The eastward propagation of surface and subsurface temperature signals from WIO that precede the development of heat content anomaly in the tropical western Pacific is the key for extending the lead time for ENSO prediction. Our results are consistent with previously reported findings but add finer points to the mechanisms of ENSO-IOD interactions and improve the predictive understanding of the monsoon-IOD-ENSO system. Forecast experiments with CFS are underway to quantify the impact in the full coupled framework so that the details of the oceanic tunnel and the atmospheric bridge of this active partnership between IOD and ENSO can be fully exploited.


***EEOF = Extended Empirical Orthogonal Function. This was part of their analysis based on: "the spatial-temporal covariance matrix of monthly mean ocean temperature from the last 480-year ENSO run averaged between 10°S and 5°N with a temporal window of 18 months. The longitude depth domain for the EEOF analysis is from 50°E to 180° over the tropical Indian Ocean and western Pacific and from the 5-m depth to the 225-m depth below the sea surface."




The authors have been carrying out research into this complex relationship for a number of years and this is no one off knee jerk reaction or an isolated claim. I am always cautious about model simulations and the conclusions drawn from the results as this does not always play out the same way in the real world. These days there are so many papers on a diverse range of meteorological and climatological topics that are based on simulations.  I like it when some research studies use "hindcasting" where the simulated model is applied to past events and the results compared to what is the known outcome (ie: still in the past). Almost all of this research was conducted effectively using these hindcasting principles and applying them to many past ENSO and IOD cycles. Just read the three main conclusions above. 1 states "strong" co-variations between IOD and ENSO with sub surface ocean variability; 2 is a definite statement or claim that ENSO enhances the intensity and lowers the frequency of the IOD;  3. the IOD "may" help ENSO prediction at longer time leads - ie: far less certainty in terms of the IOD influences on ENSO.  So more research needed and more papers and studies to be analysed.


Please note that the authors' earlier full paper: "Evolution of Indian Ocean dipole and its forcing mechanisms in the absence of ENSO" from January 2016 is still behind a paywall.  I shall keep the abstract to it in my "availability pending" file.  If anyone wants to read the paper who subscribes to the "Springer Link" library, please PM me for details.


7.    Causes and Predictability of the Negative Indian Ocean Dipole and Its Impact on La Niña During 2016


This excellent, well constructed 2017 paper looks at the relationship between a strong -ve IOD and its possible influence on a weak La Nina with strong regional climate impacts. It is full of fascinating facts and many very clear charts.  I strongly recommend that readers who are interested in ENSO, the IOD, SSTs more generally or natural variabilty (vs human influences) on climate change find the time to read the full paper and the supplements. The paper goes beyond its primary focus shown in the title.



In the latter half of 2016 Indonesia and Australia experienced extreme wet conditions and East Africa suffered devastating drought, which have largely been attributed to the occurrence of strong negative Indian Ocean Dipole (IOD) and weak La Niña. Here we examine the causes and predictability of the strong negative IOD and its impact on the development of La Niña in 2016. Analysis on atmosphere and ocean reanalyses and forecast sensitivity experiments using the Australian Bureau of Meteorology’s dynamical seasonal forecast system reveals that this strong negative IOD, which peaked in July-September, developed primarily by the Indian Ocean surface and subsurface conditions. The long-term trend over the last 55 years in sea surface and subsurface temperatures, which is characterised by warming of the tropical Indian and western Pacific and cooling in the equatorial eastern Pacific, contributed positively to the extraordinary strength of this IOD. We further show that the strong negative IOD was a key promoter of the weak La Niña of 2016. Without the remote forcing from the IOD, this weak La Niña may have been substantially weaker because of the extraordinarily long-lasting warm surface condition over the dateline from the tail end of strong El Niño of 2015–16.


I will show the first part of introduction below (as I did in part 2 with a much earlier paper) as this covers those interim years too (from 2002 to 2016).


Introduction - edited:

The Indian Ocean Dipole mode (IOD) is the leading mode of interannual variability of sea surface temperature (SST) in the tropical Indian Ocean during the boreal summer-autumn seasons. The positive phase of the IOD (+ve IOD) is characterised by negative SST anomalies to the west of Java and Sumatra and positive SST anomalies in the tropical central-western Indian Ocean. Opposite signed SST conditions are found during the negative phase of the IOD (−ve IOD). The IOD typically develops during boreal summer, peaks in autumn, and then rapidly decays in November and December when Australian summer monsoon starts. The IOD significantly affects the climate of the Indian Ocean rim countries such as eastern Africa, India and Indonesia and remotely influences the climate of southern Australia and north eastern Asia that are located under the pathways of equivalent barotropic Rossby waves emanating from the tropical Indian Ocean.


Development of the IOD is often linked with the El Niño-Southern Oscillation (ENSO) in the Pacific because of variations of the Walker Circulation, which has been explored by a number of studies. In general, the +ve IOD tends to occur with El Niño and the −ve IOD with La Niña, although exceptions exist. This relation between the IOD and ENSO peaks in boreal autumn when remotely forced wind anomalies over the Indian Ocean, as a result of the alteration of the Walker Circulation in response to ENSO in the Pacific, are reinforced by strong local positive air-sea feedbacks. The remote forcing of the IOD by ENSO is an important source of long-lead predictability of the IOD. The IOD may also act to reinforce or weaken concurrent development of ENSO and subsequent development of ENSO in the following year.

The current study is motivated by the extraordinarily rapid and strong development of the −ve IOD in 2016, which was the strongest −ve IOD event observed during June to September in the last 60 years as diagnosed using the POAMA ensemble ocean data assimilation system (PEODAS) ocean reanalyses and the merged Hadley-NOAA/OI SST analyses (Fig. 1). Although the −ve IOD of 2016 was not an extreme event as depicted in the HadISST dataset, it was a strong event with a greater than 1 standard deviation (σ) magnitude of the Indian Ocean Dipole Mode Index. This strong −ve IOD of 2016 was reported to have played a key role in promoting the extreme wet conditions over Indonesia and Australia from May onwards and was later accompanied by a weak La Niña  of 2016.

Our approach is to conduct forecast sensitivity experiments using the Australian Bureau of Meteorology’s dynamical seasonal forecast system, POAMA. This system has demonstrated skill to predict peak season ENSO and IOD at the lead times of beyond 9 months and up to 5 months, respectively and therefore, is a suitable tool to investigate impacts of the IOD on ENSO and vice versa.


The Indian Ocean Dipole amplitudes and the sea surface temperature anomalies of June to September 2016.


(a) Standardized Indian Ocean Dipole Mode Index (DMI) obtained with the June to September (JJAS) mean SST anomalies of PEODAS ocean reanalysis (dark blue bars), the merged Hadley-NOAA/OI SST analyses (orange bars), and the HadISST dataset (green bars) from1960 to 2016. The dotted horizontal line indicates negative 1 standard deviation (σ) of JJAS DMI. The SST anomalies used in the DMI calculation were relative to the climatology of 1981–2010.


(b) 2016 anomaly pattern of PEODAS JJAS mean SST relative to the JJAS mean climatology of 1981–2010. The colour shading interval is 0.2 °C. 


The authors explain their experiments into 7 differing ocean states and their methodology (explained in detail in several supplements). They point out some important facts about the state of the thermocline that I wasn't aware of and merits much deeper investigation when we analyse SSTs and sub-surface temperatures: 


"The linear trend in the tropical SST computed over the 55 years is significantly positive in the central Indian Ocean and far western Pacific Ocean, reaching up to 1.5 °C/55 years  The trend in the equatorial subsurface temperature is also strongly positive (i.e. deepening of the thermocline) in the eastern Indian Ocean, reaching up to 6 °C/55 years (Supplementary Fig. S5). On the other hand, the temperature trend is positive and negative in the western and eastern Pacific subsurface (+1 °C and −3 °C over 55 years, respectively), and so is the trend in the western and eastern Atlantic subsurface (+1.8 °C and −6 °C over 55 years, respectively). Consequently, the tilt of the thermocline has steepened across the Pacific and Atlantic Ocean subsurface over the period of 1960–2014 on 21 April."


The 21st April date is a measurement date taken at the same time each year for consistency. They have a link to a supplement with charts and further details. This data alone, needs to be examined not just in relation to ENSO and IOD phases but in terms of wider SST implificationss as deeper water upwellings are likely to have more significant impacts (with some greater temperature contrasts).  I wonder if this might also be related to the longer term oscillations like the AMO.  It may well become part of the climate change debate in terms of the "natural variability" processes.  Our fascinating ENSO debate looks like developing a new dimension!  I've been intending to prepare a post on global SST changes anyway.


Back to the paper and following their experiments they set out the results. This is just one of a few tables that are shown and described.



Observed and forecast DMI and NINO3.4 index. Forecasts of the DMI and NINO3.4 index of 2016 initialised on 21 April 2016 from POAMA control experiment (CTRL) (upper panels; a,c) and five different forecast sensitivity experiments (DTRND, IO, IOAO, IOPO, ClimIO) (lower panels; b,d). In all four panels, thick green lines denote the indices from PEODAS reanalysis (OBS), and solid thick blue line indicates the 11 member ensemble mean forecasts from CTRL. In (a) and (c), thin light blue lines indicate individual 11 forecasts from CTRL. In (b) and (d), dotted blue and solid red, pink, light blue and orange lines indicate the ensemble mean forecasts from the DTRND, IO, IOAO, IOPO and ClimIO experiments, respectively. See Supplementary Figs S6 and S7 for 11 member forecasts of each of the five forecast sensitivity experiments for DMI and NINO3.4 index, respectively. 


There is an analysis of the 2016 ENSO state and the "weak" La Nina" whose strength had been overstated previously.

"The difference in NINO3.4 SST between the IO and ClimIO experiments is statistically significant at the 99% c.l. from August onwards, and this contrasting evolution of NINO3.4 SST in the IO and ClimIO experiments highlights that Indian Ocean processes were critical to keeping the tropical eastern Pacific cooling in the 2nd half of 2016.  So, why was the −ve IOD of 2016 so strong (and predictable), how did the IOD act to promote La Niña, but why was La Niña so weak?"    


The next section entitled "Tug of war between strong −ve IOD and long-lasting El Niño" is fascinating but I'll jump to the conclusions with edited extracts:


Our analysis revealed:

  • occurrence and main features of the strong −ve IOD and the weak La Niña of 2016 were skilfully predictable from late April 2016 by the POAMA CTRL run;
  • the record strong −ve IOD event in June to September 2016 was promoted by the ocean subsurface and surface dynamics primarily within the Indian Ocean;
  • the long-term trend over the last 55 years in sea surface and subsurface temperatures in the Indian Ocean in late April, which is characterised by strong warming over the central Indian Ocean and deepening of the thermocline in the tropical eastern Indian Ocean, contributed to the extraordinary strength of this IOD event;
  • there may have been a synergistic interaction among the three ocean basins that strengthened the −ve IOD in September;
  • the strong −ve IOD was a key driver of the weak La Niña of 2016 that developed in the tropical Pacific. Without the −ve IOD, the weak La Niña may have been substantially weaker despite the encouraging subsurface conditions set in the 1st half of 2016 by the preceding 2015–16 El Niño;
  • the late demise of El Niño of 2015–16 over the tropical central Pacific disrupted the air-sea coupling that was required for the early stage of La Niña development

Overall a fascinating piece of research that should promote interest as part of our ENSO debate beyond the ENSO/IOD relationship.


I still haven't finished and my final part (in due course as I'm running out of time now with some major business commitments coming up during the next two weeks which will significantly reduce my time on this forum - on all threads) will cover several really recent papers, including one published this week! They are not in the portal yet. David :) 




  • Like 2

Share this post

Link to post
Share on other sites
1 hour ago, WxBillMoHill said:

Howdy I'm new here and interested mostly in teleconnections and seasonal forecasting.  I'll get the hang of posting here eventually. Feel free to help me as I paddle around.

Thanks for the AAM stuff from WDT. That's a gold mine.


Hi @WxBillMoHill and a big welcome to the 33andrain forum.  We look forward to your involvement and particularly on this Teleconnections thread and using the Research Portal.  Feel free to ask questions or PM any of us. Geoff @33andrain and Pat @NJwxguy78 run this superb and friendly forum and I'm copying them in on this message too.


David :)   

  • Like 4

Share this post

Link to post
Share on other sites

Have the September Qbo numbers been put out yet? I believe August was around -20.0?

  • Like 1

Share this post

Link to post
Share on other sites
2 hours ago, mbaer1970 said:

Have the September Qbo numbers been put out yet? I believe August was around -20.0?

What level? From the chart above, looks like just above -10, which is a bit of a progression from last time I checked a week or so ago. Reality is that it will be in positives by winter, although QBO forecasting is fraught with danger I might add.



  • Like 2

Share this post

Link to post
Share on other sites
47 minutes ago, Snowy Hibbo said:

What level? From the chart above, looks like just above -10, which is a bit of a progression from last time I checked a week or so ago. Reality is that it will be in positives by winter, although QBO forecasting is fraught with danger I might add.




Yes, from the ECMWF chart for Oct 3rd the Zonal Mean Zonal Winds at 30mb were running negative (easterly) at 10 m/s. But a look at the forecast for Oct 13th shows the 0 m/s line to be very, very close to 30mb. So from ECMWF forecasts it looks as if the QBO might well be switching positive (westerly) by end of October as the westerly Zonal Winds continue their descent over the Equator.


Oct 3rd: QBO Actual 03Oct2018.jpg

Oct 13th: QBO Forecast 13Oct2018.jpg



  • Like 4

Share this post

Link to post
Share on other sites
8 hours ago, Blessed Weather said:


Yes, from the ECMWF chart for Oct 3rd the Zonal Mean Zonal Winds at 30mb were running negative (easterly) at 10 m/s. But a look at the forecast for Oct 13th shows the 0 m/s line to be very, very close to 30mb. So from ECMWF forecasts it looks as if the QBO might well be switching positive (westerly) by end of October as the westerly Zonal Winds continue their descent over the Equator.


Oct 3rd: QBO Actual 03Oct2018.jpg

Oct 13th: QBO Forecast 13Oct2018.jpg




I have read that Qbo stalls, neutral trends, and around transitional periods are clues to watch for storm signals, as well as when the Qbo is between -10.0 & +10.0. Has anyone else seen any verification for this? I have always thought that a negative Qbo was an important driver for negative NAO setups, but am learning that is not the end all be all where negative NAO drivers are concerned. I am actually beginning to believe that the PDO may be a bigger influence on the NAO, than the Qbo.

  • Like 1

Share this post

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Create New...