Arguments
The oceans may be lulling us into a false sense of climate security
Posted on 5 March 2015 by John Abraham
A paper published last week in Science casts more light on oceans and how they may have contributed to a false sense of security about what we face in the future. The paper, coauthored by Byron Steinman, Michael Mann, and Sonya Miller, approached the problem in a new way that connected real-world observations with state-of-the-art climate models. What the authors find casts severe doubt on other work which had oversold the role of natural climate’s ability to halt global warming for the next 15 years. Instead, by correcting others’ errors, the new paper shows that things may be worse than we thought.
First, some definitions. The authors of this paper, in particular Michael Mann, are well known in the scientific community for researching various natural climate processes that recur periodically. These processes are often called oscillations and they are key components to short-term climatic fluctuations. The two oscillations focused on in this paper are the Atlantic Multidecadal Oscillation (AMO) and the Pacific Decadal Oscillation (PDO).
The AMO is a cyclical variation in North Atlantic temperatures that lasts for 50–70 years. On the other hand, the PDO can actually be thought of as a short (16–20 year process) and a longer (50–70 year) process. Currently, the oceans are characterized by a slightly positive AMO and a more negative PDO. A recent publication discusses the role of the PDO and the continued warming of the plant, readers can go there for a basic description.
The authors focus attention on the longer of the two PDO processes which is multidecadal and they ask whether the current status of the oceans can explain what we are observing at the Earth’s surface. What they find is interesting. These oscillations are “found to explain a large proportion of internal variability in Northern Hemisphere mean temperature.” The authors also show that other researchers who have incorrectly defined natural variability using simple linear detrending have been mistaken.
In fact, recently published work on the climatic “stadium wave,” which is a highly speculative interconnection of climatic waves that reportedly travels around the globe, is likely not correct. According to this new paper, their analysis “reveals any putative correlation between the AMO and the PMO (the longer part of the PDO) and arguments of a stadium wave climate signal to be an artifact of the linear detrending approach.”
Some recent work, including publications of the “stadium wave” advocates, have argued that large parts of the changes we’ve observed in temperatures are attributed to these oscillations. The new research shows that “the methods used in these studies tend to inflate and distort the estimated internal variability owing to an incorrect partition of internal and forced variability”.
Here is what Dr. Michael Mann had to say.
Our work reinforces the notion that there is no pause in human-caused global warming. If anything, we’ve been lulled into a false complacency by the fact that internal oscillations in the climate system temporarily masked some of that warming. That may come back to bite us as these oscillations swings back in the other direction and add to global warming in the decades ahead.
The study’s lead author, Byron Steinman told me,
Thanks for covering this important story. Abraham (from St. Thomas in St. Paul) interviewing Steinman (from UMN Duluth) just drives home how many of us Minnesotans are important players in GW issues!
Meanwhile, breaking news:
NOAA officially calls it: We are in El Niño conditions.
www.cpc.ncep.noaa.gov/products/analysis_monitoring/enso_advisory/ensodisc.html
[Rob P] - A sizeable Kelvin wave is heading east across the Pacific Ocean. Combined with observed slowdown of the South Pacific subtropical gyre, there is the potential for a strong El Nino developing this year too. I suspect that the sustained trade winds in the southern hemisphere last year prevented proper ocean-atmosphere coupling. 2015 is shaping up as another interesting year weather wise.
Thanks for the reply and the cool graphs.
Sooo, is it time for another El Nino update?
Typo alert for the main article above:
"continued warming of the plant"
> potential for a strong El Nino developing this year too.
Did you mean to quantify that as "low potential for a strong ..."?
I haven't found an agency suggesting a strong El Nino is at all likely.
E.g. http://www.realclimate.org/index.php/archives/2015/02/climate-oscillations-and-the-global-warming-faux-pause/comment-page-3/#comment-626577
[Rob P] - Don't have much time to explain at the moment, but I'm writing a series on the Interdecadal Pacific Oscillation (IPO) which ties into this.
You are, however, misinterpreting what I have written. The potential for a strong El Nino exists because the southern hemisphere subtropical cell and subtropical ocean gyre circulation have spun down, unlike last year where the trade winds were persistent south of the equator. There is a greater potential for the current Kelvin wave to cross the Pacific and not be weakened by the subtropical cell and South Equatorial Current pumping heat out of the tropics - thus encouraging atmospheric reinforcement.
This doesn't mean we are due for a strong El Nino, merely that underlying oceanic conditions are more conducive to such. The situation could quickly change. As ever, I keep a beady eye on the situation.
Hi Wili,
I think you will be proven correct, but you're slightly premature at the moment.
The key phrase to look out for is "el Nino conditions".
For an el Nino to be declared, the 3-month rolling average has got to be 0.5oC (or more) up on climatology for at least 5 consecutive periods. The last 5 such periods were as follows...
ASO(+0.2) SON(+0.5) OND(+0.7) NDJ(+0.7) DJF(+0.6)
We therefore need to wait for the March numbers, but - assuming I've got my sums right - unless the monthly anomaly drops to below about +0.2oC, then the 3-month JFM figure should be at least +0.5oC (with rounding).
The rolling-3 figures are given here.
cheers bill f
bill f: I do believet hat wili is correct...
Just when everyone had pretty much written it off, the El Niño event that has been nearly a year in the offing finally emerged in February and could last through the spring and summer, the National Oceanic and Atmospheric Administration announced Thursday.
This isn’t the blockbuster, 1998 repeat El Niño many anticipated when the first hints of an impending event emerged about a year ago. This El Niño has just crept across the official threshold, so it won’t be a strong event.
“We’re basically declaring El Niño,” NOAA forecaster Michelle L’Heureux said. “It’s unfortunate we can’t declare a weak El Niño.”
After Much Ado, El Niño Officially Declared by Andrea Thompson, Climate Central, Mar 5, 2015
The NOAA report indicates that current Nino 3.4 conditions are in the El Nino range. It also indicates the expectation is for low level El Nino conditions through 2015.
The NOAA report includes Figure 7 showing the ONI predictions of models (The NOAA ONI is the 3 month average of the Nino 3.4 region). Only a few models indicate the potential for a reasonably strong El Nino in 2014/15. Even the strongest model predicted event is significantly less than the magnitude of the 1997/98 event which had ONI peak values of 2.3 and 2.4 C as can be seen in the NOAA ONI value history here). Most models indicate a low level El Nino and a few indicate no El Nino.
The latest Australian summary of ENSO conditions here also indicates the expectation is only weak El Nino conditions to develop.
Modelling of long term generalized multi-year average climate conditions has been proven to be quite reliable (with the primary uncertainties being the actual impacts of things like how much excess CO2 will be generated), however, the accurate prediction by models of near term things like the ENSO is less 'reliable'. In spite of the variable success of near term prediction, in the big picture the average of the model predictions of many such potential climate related events is likely quite reliable. Said another way, if you looked at many years of predictions of the ONI the average predictions of all the models through all of the years would probably be seen to be quite reliable, even though specific years could be found where the outliers among the predictions were the ones that got 'that year right'.
So, for a strong El Nino event to develop in 2015 an outlier model prediction would need to be the correct one, which has a low likelihood of occurring, but such unlikely events can still be what happens.
What exactly will happen in the near term can be quite uncertain and is best forecast by "Wait and see", even though what will happen long term, in the averages of many years, is able to be quite reliably predicted.
@ John Hartz #6
Hi John,
On the Oceanic Nino Index page to which I linked in #5 (and unwittingly repeated in #7), it states that...
"For historical purposes cold and warm episodes (blue and red colored numbers) are defined when the threshold is met for a minimum of 5 consecutive over-lapping seasons"
I've always taken that to indicate the difference between el Nino conditions being met for some undefined transient period, and a formal el Nino declaration. (That's why the red/blue periods are always in groups of at least 5.
However, I was sure I had also read something to that effect elsewhere. The page I was trying to remember is here, and the relevant bit says...
"Niño 3.4 SST anomalies are averaged over the three months ending with the current month, and that value is called the Oceanic Niño Index (ONI). If the ONI exhibits warm or cool phase conditions for at least five consecutive values, it officially becomes an El Niño or La Niña event."
Obviously, if the basis of your understanding supersedes the above, then I stand corrected and humbly defer to you and Wili.
Either way, this is almost at the semantics level: I'm sure we all agree that it's on the cards - if not now, then next month.
cheers bill f (and apologies again for the inadvertent repetition)
[PS] Fixes done as requested
billthefrog@8,
I share your understanding of the terminology. El Nino/La Nina conditions are the average surface temperature anomaly of the of the Nina 3.4 region. As indicated on the NOAA ONI history page we both linked to in our earlier comments, the anomaly is being measured against a 30 year average. And that 30 year average is updated every 5 years because the ocean has been warming. The NOAA ONI page includes a link to a page that presents how the 30 year average has increased. It also shows that the 30 year average is not just 'a temperature', the 30 year average that the anomaly is calculated from has a different value for each month.
So an El Nino condition occurs any time the Nino 3.4 region anomaly is 0.5 C or warmer (the term is probably even applicable when weekly average values are 0.5 C and above). And NOAA declares an El Nino event to have occured when a set of 5 consecutive 3 month averages of the Nino 3.4 anomaly, what they call the ONI, are 0.5 C or warmer.
The Australian Bureau of Meteorolgy definition of what constitutes an El Nino event may not be exactly the same. Their latest ENSO update here states there is a 50% chance of El Nino forming in 2015 even though by NOAA methods only one more month of warm enough waters would be needed for NOAA to declare that an El Nino event has occurred.
There is also a range of strengths of El Nino events and even the linking of the ocean surface temperatures to trade wind patterns that could spread the warming effect of air passing over the warmer ocean surface to other areas of the planet. And there is other Pacific Ocean surface anomalies like the current large warmer area off the west coast of N. America shown on the following Australian link that can occur without an El Nino event being declared.
It is very complex, but regardless of the potential variations of terminology there is no doubt that the warming due to excess CO2 from human activity continues to occur.
bill f:
Wili and I both documented that NOAA had declared an El Nino as also detailed in the following:
National Oceanic and Atmospheric Administration climate scientists reported the switch to official El Niño status in their latest technical bulletin on Thursday, and outlined their decision process in a blog post .
El Niño Has Arrived, and It Could Produce the Warmest Year on Record by Eric Holthaus, Slate, Mar 5, 2015
Here is the running 30 day SOI index:
Negative values indicate El Nino like conditions, while positive values indicate La Nina like conditions. Clearly, since July 2014 El Nino like conditions have prevailed, and by comparison with the SOI figures of 2010, you could argue that we have been experiencing an El Nino if you used SOI figures alone. However, the sudden rise to positive values call into question the continuation of those conditions. Such sudden changes often immediately reverse themselves, as in May 2013, or April 2014, but they can also be followed by a sustained reversal.
The SOI is a proxy for the strength of the trade winds, and we can now look at direct measurements of that strength. The BoM writes:
The weakening, and even reversal, of the trade winds in the western tropical Pacific would weaken the Tropical Warm Pool as surface water flows east to the central Pacific. The continued strength of the trade winds in the eastern tropica Pacific, however, will prevent the warm water flowing further east, and the formation of a true El Nino.
This can be seen in the SST data:
There has been a marked warming in the Nino3 to Nino 4 region, but the rest of the tropical Pacific is neutral with respect to 1961-1990 average. Taking into account the effect of global warming, that means they are slightly cooler than we would expect. Crucially, that means while we currently have drier conditions for eastern Australia, as would be expected from an El Nino, we do not have a pool of warm water of the central American coast bringing wetter conditions there.
It is possible that the eastern Pacific trade winds could weaken, turning this into a true El Nino, but it is by no means certain. I think the BoM is right to be cautious. I also think looking at more data than just the SST in a single constrained region shows my distrust of such indices (ie, Nino 3.0, 3.4 and 4) is justified.
Unfortunately, the pages for the Multivariate Enso Index are currently down, so I cannot add them into the analysis.
@OPOF #9
The NOAA page showing the rolling 5 yearly baseline increments to which you linked does indeed contain a wealth of information .
Up until the last 4 such periods, the rise was pretty monotonic, but has subsequently gone up in punctuated lurches. If you calculate an overall average for the ONI in each of the 4 most recent 30-year windows, you see that it also lurches in an analogous fashion. Each time the baseline is recalculated, it is somewhat affected by the relative frequencies and intensities of ENSO events occurring within that 30 year window.
In December last year, I thought I had noticed a typo in the page describing the changes, but I was just being thick. Michelle L'Heureux of NOAA had to patiently explain to me that the first of the 5-year shuffles was anomalous because it wasn't 5 years long - it was 6 years.
cheers bill f
@ John H & Wili
Hi again guys,
As Tom Curtis sagely states in #11, a single number is a pretty restrictive way of thinking about ENSO. The people at NOAA obviously think so as well, as, back in January, they put out a blog piece describing the need for, and relationships between, a multitude of ENSO-related indicies.
In the body text of NOAA's diagnostic discussion and the NOAA blog piece that were both linked to in #10, the phrase "el Nino conditions" occurs more often than does "el Nino" without the trailing modifier. When restricting the context to references to the current state of affairs in the Pacific, the phrase "el Nino conditions" is vastly more prevalent.
By way of an analogy, in the whacky world of economics, there exists the dreaded word - recession. A country can often (always?) be described as being "in recession" when it has experienced two consecutive quarters of negative growth. (I just love that expression!)
If we were 70%-80% of the way through the second quarter of negative growth, and things were still in the shit, some analysts would say the country was in recession. Others (especially those representing the party in power) would cavil at this, saying a necessary criterion had not been reached - yet.
OK, call me Mr Cavil. I am aware that I am raising a petty objection, but, in the NOAA blog relating to multiple indicies, it clearly states in the opening paragraph...
"At NOAA, the official ENSO indicator is the Oceanic Niño Index (ONI), which is based on sea surface temperature (SST) in the east-central tropical Pacific Ocean."
It is also perhaps worth remembering that, in addition to not yet having 5 consecutive rolling 3-month averages >= + 0.5oC, we don't even have 5 individual months, as the October figure was +0.46oC.
To avoid going round in circles further, I have sent an email to Michelle l'Heureux at NOAA to ask if she, or a colleague, might care to drop in and tell it like it is.
cheers bill f (pedant to the nobility)
bill f: To complicate matters further, see the discussion contained in:
El Nino declared as climate scientists watch on with 'amazement' by Peter Haman, Sydney Morning Herald, Mar 6, 2015
It appears that meterologists in both Japan and Australia were surprised by the NOAA declaration.
The bottom-line seems to be that the chararcteristics of the current El Nino do not completely align with those of the typical El Nino.
At what level of Global CO2 emission would the Atmospheric PPM of CO2 come down?
billthefrog,
The NOAA presentation of the history of 30 year averages of the Nino 3.4 region that they re-baseline every 5 years also shows that the most recent 30 year baseline values (1981-2010) are about 0.4 C warmer than the first of the 30 year baselines (1936-1965). This means that an el Nino event today has suface temperatures that are 0.4 C warmer.
And as you noted there are definite leaps of the baseline that were due to significant El Nino events during the 1991-1995 and the 2000-2005 period (which can be seen in the NOAA ONI history). Note that the 1997/98 event did not produce a step up because it was in a set of years that included significant la Nina events. So I would suspect that the next update of the ONI baseline for 1986-2015 will not be another leap up.
Cool Hand Luke @15, there has been some discussion of this at And Then There's Physics which you may find useful, starting with Jac's question.
In short, CO2 will be drawn down essentially by four processes. They are:
1) Draw down and equilibrization into the surface ocean and biosphere;
2) Draw down and equilibrization with the deep ocean;
3) Reaction of oceanic CO2 with calcium carbonate (CaCO3); and
4) Chemical weathering of rocks leading to a long term draw down of CO2.
All of these occur at different rates. The first occurs over the first year after emissions, and accounts for most of the reduction of the airborne fraction (ie, the fact that only 45% of total emissions actually remain in the atmosphere). The rates of the other three are illustrated by this diagram:
The diagram assumes an instantaneous pulse of CO2, so it does not distinguish between the surface and deep levels of the ocean. Further, it shows all CO2 initially entering the atmosphere. In the actual situation, with the gradual ramp up of CO2 emissions (in human terms), the start point for draw down would be at about 700 ppmv for the emissions shown in the diagram. Therefore ocean equillibriation (step 2) would draw down from 700 to 650 ppmv over 300 years, or at a rate of 0.167 ppmv per annum. As there are 2.12 (IPCC AR5) to 2.13 (CDIAC) Gigatonnes of Carbon per ppmv of atmospheric CO2, that amounts to 0.355 GtC per annum to allow a stable atmospheric concentration over the first approx 300 years of after peak emissions*.
After that, calcium carbonate buffering (step 3) draws down CO2 from about 650 to 450 ppmv over 5000 years. That represents a reduction of 0.043 ppmv per annum, or about 0.09 GtC per annum to maintain constant CO2 over that period.
Finally, the remainder of the CO2 is drawn down by chemical weathering over the course of about a million years. That represents 0.00017 ppmv per annum, or 0.0004 GtC per annum, for what is practically for ever (as David Archer points out).
To summarize, that is approximately 0.355 GtC per annum for the first 300 years, 0.043 GtC per annum for the next 5000 years, and 0.0004 GtC per annum thereafter. In terms of current emissions, that is 3.55% for the first 300, 0.43% for the next 5000 years, and diddlysquat thereafter.
You should not get too hung up on these figures. Different models of carbon uptake will vary the results by up to 30%. Further, the exact figures change significantly with increased cumulative CO2 emissions. Further, increased climate sensitivity reduces the permissible short term emissions (due to reduced capacity for the ocean to store CO2), but increases permissible long term emissions (due to increased chemical weathering). Finally, and obviously, I have made linear estimates of obviously non-linear functions so that there will be substantial variance over some years, particularly in the short term. In short, these are ballpark figures, not exact values. The important points to realize are that:
1) Currently only 45% of emissions remain in the atmosphere because we are using the bulk of short term storage in other earth systems (ocean, biosphere). Therefore we cannot simply reduce CO2 emissions to 45% of current values and expect CO2 concentrations to stabilize. We need a much larger reduction than that.
2) Within a very short time in historical terms, we will need to reduce CO2 emissions to essentially zero. Given that, we might as well make that our current target.
3) Stabilizing CO2 is a bad call in any event, as it ensures we experience the full equilibrium climate response, whereas by reducing CO2 levels (by natural draw down if by no other method) we can ensure an earlier and lower peak climate response, significantly reducing the damaging impacts of anthropogenic global warming.
* I have not read a decription of the model run used for this graph, and may be misinterpreting the net emissions and time period over which they occur. Calculating the short term stabilization target by a rule of thumb by David Archer, ie, that we will 25% of emissions will remain in the atmosphere in the short term, with the rest lasting effectively forever, we can calculate that there will be a 20% further reduction over the three years. For current emissions, that amounts to 55 ppmv over 300 years, or 0.18 ppmv per annum (0.4 GtC per annum). Ergo, the figure is in the right ballpark. More importantly, it is nowhere near the 45% (or even 55%) of current emissions some people mistakenly estimate.
@ John H #14
Of course what's far more significant than semantics is the effect that this el Nino-ish thing is going to have on global temps. (And more importantly still on people's lives.)
As you know, Gistemp, HadCRUT, NCDC, BEST and JMA all had last year as the warmest on record. Focussing just on Gistemp (as an example - hopefully not cherry-picked) Jan 2015 was 0.07oC up on the equivalent month in 2014. This will already have pushed the rolling 12 month average up by another ~0.006oC.
Feb 2014 was the real party-pooper, as (on Gistemp) it was almost a quarter of a degree down on the J-D annual average. However, the Nino 3.4 monthlies are frighteningly different. This year, the Feb anomaly is about 1.3 or 1.4oC up on last year. Similarly, the deltas on the Nino 3.4 monthlies for Dec and Jan were both a fraction under 1.2oC up on their equivalents from a year earlier.
If there is any predictive skill there, one might reasonably expect Feb 2015 global surface temps to be > ~0.1oC up on Feb 2014, hence pushing the rolling 12 up by about another ~0.01oC.
Of course, as the Pacific has an area of about 165 million sq kms, as compared to the paltry 6.2* million sq kms of the Nino 3.4 region, some caution is called for. (* Assuming I can still remember how to work out areas on a sphere.)
@OPOF #16
"I would suspect that the next update of the ONI baseline for 1986-2015 will not be another leap up"
Yep, that's exactly how I interpreted it as well. Let's wait and see if we're correct.
cheers bill f
This discussion about 'declaring an El Nino' has brought up some new details that help me better understand what is going on. It appears there are actually two different El Nino related matters being discussed:
Reviewing the NOAA ONI history it is clear that many El Nino conditions begin in May, June, or July of one year and continue until the early part of the following year. However, the ONI record includes several instances where the El Nino events get established later in a year and continue through the next year only waning after a nearly 2 year period. The following are the cases where this occurred:
What is certain is that the CO2 concentration in the atmosphere continues to rise rapidly. And that increased amount of CO2 will produce significant changes, including changes ine the oceans bue to a large percentage of human produced excess CO2 going into the oceans. And those changes can be seen in the history of many measurements, including in the changes in the 30 year average SST values in the Nino 3.4 region that NOAA updates every 5 years seen here.
Regarding my previous point about how the El Nino season of 2014 started not beaing an indication of how it will develop, I was referring to the strong SOI values in early 2014 whish are often precursers of a developing El Nino. THe SOI history is here. In the SOI history the March value was strongly negative at the time of the season an El Nino condition has historically been seen to begin. That SOI was very short lived and the ONI never reached +0.5 C. However, the Australia Bureau of Meteorology ENSO forecasts for 2014 were for a higher probability that an El Nino would develop in 2014 than they are currently indicating for 2015 (current report is here).
[Rob P] - Current status of the Kelvin wave making its way east across the equatorial Pacific Ocean...
Rob P, The Kelvin wave appears to be significant and has grown since the Mar 4 image you shared in response to the first comment. It will be interesting to see what ultimately develops. It will also be interesting to see if the SOI values grow more suportive of El Nino. The Prelimiary SOI values here indicate that the SOI may be returning to steady negative values. Time will tell.
Also, a further clarification of my comment about the Australia Bureau of Meteorology ENSO forecasts of last year compared to this year. The history of their ENSO Tracker levels is here. The February 2014 level was El Nino Watch. The February 2015 level was El Nino Neutral. The current ENSO Tracker level on the most recent ENSO Wrap-up here has reterned to El Nino Watch
Presumably, if the authors have a handle on the periodicity of these two waves that influence climate, they should be able to predict when will be the next time that they reinforce each other in the warming direction for the climate.
I think the ocean cycles are quasi-periodic and thus do not have the predictability which would allow prediction of when they would next reinforce.