Posted by: erl happ | June 10, 2009

El Nino. How big? How long?

This ‘plume diagram’ of ENSO model predictions for Sea Surface Temperature in the ENSO 3.4 region appears in the May 20th 2009 forecast of the International Research Institute for Climate and Society (IRI). A five month running mean greater than 0.4 that continues for six successive months constitutes an El Nino event. Ten of the 22 models predict that the ENSO 3.4 region will reach the 0.4°C threshold in the September to November period. ENP 1

A record of recent prediction appears as figure 2.

Figure 2

Figure 2

The IRI remarks: “The graph shows forecasts made by dynamical and statistical models for SST in the Nino 3.4 region for nine overlapping 3-month periods. Note that the expected skills of the models, based on historical performance, are not equal to one another.”

Well said!  In fact there is little evidence of any predictive skill at all. It appears that the models take more notice of each other than the phenomena that they purport to predict. When sea surface temperature changes direction, it leaves all the models pointing another way.

So,  ‘climate science’ has no clue as to the origin of ENSO events. ‘ Climate Science’  is ignorant of the mechanism that drives the change in global temperature that we observe on a month by month and year by year basis. How can ‘climate science’ and its political manifestation, the UNIPCC rule out ENSO as the source of the recent temperature increase? How can the conclusion that warming is due to ‘ greenhouse gas’, or any other possibility, be seen as ‘settled science’.

When you have read this presentation you may conclude, as I have, that there is no need for recourse to ‘ greenhouse gas mumbo jumbo’ to explain the warming of the surface of the Earth between 1976 and 2005.  Natural modes of variability are more than adequate.  ENSO is the mode of natural variability that is responsible for recent change.

The data that follows, in all instances, is a moving 12 month average centered on the seventh month. This procedure removes the seasonal influence. The raw data is accessible at: This data is referenced as: Kalnay, E. and Coauthors, 1996: The NCEP/NCAR Reanalysis 40-year Project. Bull. Amer. Meteor. Soc., 77, 437-471.

Figure 3

Figure 3 Pink is SST in East Pacific

Figure 3 shows that change in sea surface temperature (SST) in the East Pacific frequently precedes the ENSO 3.4 region. Instances are marked with red arrows). Where ENSO 3.4 leads I have marked the turning point with blue arrows.  Red rectangles on the axes indicate that the scale is adjusted so that the axes are directly comparable. The interval is one degree C.

Figure 4

Figure 4 Pink is SST off Chile

Figure 4 shows that change in SST off Chile leads SST near the Galapagos Islands. Rarely is it the other way round.

Figure 5

Figure 5 Green is atmospheric pressure in SE Pacific Purple is SST in SE Pacific

The differential between atmospheric pressure across the Pacific determines the strength of the the trade winds, the long observed manifestation of EL Nino warming being a slackening of the Trades.  In figure 5 the pressure axis is inverted by the simple expedient of multiplying the data by minus 1. So,on this graph, pressure increases in the downward direction. Figure 5 shows that the change to La Nina cooling is signaled by a rise in atmospheric pressure (green) in the near permanent high pressure cell located in the South East Pacific. The rise in atmospheric pressure is accompanied by a fall in sea surface temperature (purple) in the same area.

The south east Pacific ultimately delivers its waters to the Galapagos region and then to the ENSO 3.4 region. It is apparent that the this region signals the forthcoming change in the global tropics. In June 2009, SST is already falling in the waters off the coast of Chile. It is only a matter of time before these waters reach the equator.

In figure 5 the slope of the dotted lines bounding the change in pressure is a gauge of the tendency for the tropics to warm or cool over time. It is apparent that the tendency since 1983 is for La Nina cooling events to reach a lower temperature in each successive cooling event. A similar pattern marked the cooling cycle of the 1970’s. The maximum temperature reached during successive El Nino events is also in slow decline. The warming cycles of 2007 and 2009 are remarkably warm in the context of the trend from 1983 through to the present time. If this declining trend in atmospheric pressure is maintained, a deeper La Nina than that of 2008 might be expected in 2009-10.

Figure 6

Figure 6

Figure 6 confirms that a fall in sea level atmospheric pressure in the south east Pacific frequently heralds a decline in sea surface temperature in the global tropics. A cooling cycle could be imminent. But, will it go all the way?

Figure 7

Figure 7 Purple is SST off Chile

Figure 7 shows that SST off Chile (blue) is more volatile than SST in the global tropics. Change in SST near Chile frequently signals the direction of global change. In the historical context, SST off Chile has never been higher than in was in the early 1980’s. The very recent warming cycles of 2007 and 2009 constitute a  rude  interruption of the trend for gradual cooling of the south east Pacific that set in after 1983.

Figure 8

Figure 8 Red is 20hPa temperature 10N to 10S Latitude

Figure 8 shows the relationship between the QBO in stratospheric temperature and sea surface pressure in the south east Pacific. There appears to be a connection. Frequently, sea surface pressure falls about the time that 20hPa temperature peaks.

Figure 9

Figure 9

The quasi biennial oscillation in stratospheric temperature at 20hpa is due to changing ozone levels. In figure 9, 20hPa peaks align with peaks in  upper troposphere temperature at 20-30° south latitude. A projection, taking into account the lagged response, indicates that the peak in the next warming cycle should be reached about May 2011.

Note the historically high temperature at 20hpa in the stratosphere during the 1970s when the tropical ocean was cool. After 1978 increased SST  resulted in increased evaporation from tropical waters. The incursion of moisture into the stratosphere depletes ozone, lowering temperature. Notice the gradual recovery in stratospheric temperature at 20hpa since 1996. This recovery is consistent with ocean cooling and a drying stratosphere.

Figure 10

Figure 10 Red is 200hPa temperature

Figure 10 shows the dependence of SST on temperature in the upper troposphere. It is apparent that 200hPa temperature is independent of sea surface temperature.  The temperature of the upper troposphere is driven by its ozone content as well as the temperature at the surface.

A massive increase in 200hPa temperature occurred with the climate shift of 1978 and it has has been in slow decline since 1983.

200hPa temperature is an indicator for ice cloud density in the upper troposphere.  Ice cloud density determines the amount of sunlight reflected versus that which reaches the surface of the ocean. The evidence for this statement is the subject of a forthcoming post. At this point let me note that there is an inverse relationship between ISCP high cloud data and 200hPa temperature  in many areas between 10° and 60° latitude in both hemispheres with strongest correlations at about 30° latitude.

There is no evidence that ice cloud inhibits the transfer of outgoing long wave radiation to space. On the contrary, the presence of ice cloud is an indicator of strong convection that occasions a de-compressive cooling of the atmosphere. Diminished outgoing long wave radiation (OLR) is used in meteorology as an indicator of increased convection and it is the means by which the Madden Julian oscillation is charted. Outgoing long wave radiation from the tropical regions has been in decline as the tropical sea has warmed. In regions where the descending air is warmed by compression, ice cloud density will fall away. In this way more solar radiation  reaches the ocean in the enormous traveling high pressure cells of the subtropics.

Figure 11

Figure 11

Figure 11 pertains to the northern hemisphere. It shows that 200hPa temperature increases directly with 20hpa temperature. Peaks in temperature in the stratosphere impact 200hpa temperature to an extent dependent upon latitude. The effect seems to be least lagged at 20-30° north and, where it lags, it is most lagged at the equator. This may be due to a lag in the warming response as warmed waters are driven towards the equator by the trade winds. It may also be due to some quirk in the stratospheric circulation and stratosphere-troposphere exchange.

The author suggests that the ENSO 3.4 region is not the most appropriate place to monitor the the ENSO phenomenon. Change is initated in the low mid latitudes, well away from the ENSO 3.4 region. Temperature in the ENSO 3.4 region is a dependent variable. Generalizing, change in SST is associated with change in upper troposphere ice cloud phenomena consequent upon a quasi biennial increase in the ozone content of the stratosphere/upper troposphere. The best place to to monitor ENSO dynamics is the stratosphere and upper troposphere.

Figure 12

Figure 12

Figure 12 examines the relationship between 20hPa temperature in the stratosphere and 200hpa temperature in the southern hemisphere.

  • There is a strong impact of change in 20hPa temperature as far as 40°south.
  • The change in 200hpa temperature between 1976 and 1983 was much greater at 30-40° south than at any latitude closer to the equator.
  • The increase in 200hpa temperature between 1976 and 1983 was much greater in the southern than the northern hemisphere.
  • 200hpa temperature at all latitudes has been in slow decline since 1983.

The importance of the change in stratospheric and upper tropospheric temperature can not be over-emphasized.  Change over the last six decades represents a fundamental shift in influential atmospheric parameters.  There is abrupt change, as between 1976 and 1983, and also  gradual evolution .  At 30-40° south, 200hpa temperature gradually increased between 1948 and 1976 as it fell at the equator.  This change  fundamentally alters Hadley cell dynamics in the southern hemisphere. The zone that is impacted by ozone dynamics expands southward. This increases the tendency to drier winters in the sub tropics because it affects the passage of the fronts that are responsible for winter rainfall. The change in cloud cover affects the temperature regime. Where I live, in South Western Australia, rainfall has been in decline for 100 years.  This gives an idea of the length of natural cycles of climate change. ENSO dynamics operate on very long time scales. It is apparent that the warming trend in the tropics and globally, and the current cooling trend, has much to do with the change in the phenomena driving the Southern Oscillation in the southern hemisphere. There is no evidence that the sunspot cycle per se is the agent of change. However, the sun is involved in a manner that will now be described.

Figure 13

Figure 13

Figure 13 records the highest and lowest 20hPa temperature reached for each 10° latitude band over the entire period of record. This diagram therefore shows the extent of variability in the stratospheric temperature between 1948 and 2008.

  • A slightly expanded variability is seen in the region of the equator. If we bear in mind that 20hpa temperature oscillates with an amplitude of up to 4°C in the QBO, this zone is relatively stable in its temperature regime.  The larger variations occur elsewhere.
  • Between 20°N and 50° north there is enhanced variability associated with the presence of the great landmasses in the northern hemisphere. This is associated with flux in atmospheric humidity on ENSO time scales. In other words, a La Nina dries the atmosphere and tends to produce warmer conditions in the warmest months in the northern hemisphere. Global surface and atmospheric temperature peaks in July. At this time cloud cover is least.
  • The greatest variability in stratospheric temperature is associated with the southern pole. The Antarctic vortex is stronger than the Arctic vortex and unlike the Arctic it prevails all year round, albeit weakened in summer.
  • Enhanced variability at 80-70°N in the Arctic is associated with the competing draw of Siberia and Greenland on the one hand and the North Pole on the other. The Arctic vortex operates fitfully in northern winter and not at all in the summer. Hence, average temperatures are high. Warming in the Arctic winter is associated with strong warming of the sea at 20-30°S just as warming in the Antarctic winter is associated with warming at 20-30°N. See my post re ‘sudden stratospheric warming’ at:
  • Short term variability in the Arctic is a winter phenomenon. This variability is associated with change in the strength of the vortex and change in the concentration of erosive nitrogen compounds descending from the mesosphere.
  • When temperatures rise at the pole they also rise strongly throughout the winter hemisphere and into the mid latitudes of the summer hemisphere. This increase in temperature is due to the reduction in the flow of nitrogen oxides from the mesosphere. Consequently the concentration of ozone rises throughout the stratosphere. The change in stratospheric temperature is accompanied by an increase in the temperature of the tropical upper troposphere, so driving ice cloud loss and increased sea surface temperature. This dynamic operates on micro and macro time scales. The shortest time scale is associated with the 27 day rotation period of the sun.
  • The ozone hole and the relatively low concentration of ozone in the southern hemisphere is largely a product of vortex dynamics and the relatively consistent supply of erosive nitrogen compounds from the stratosphere to that hemisphere. I have seen these compounds described as “responsible for the cooling of the atmosphere”. Their presence explains the coolness of the mesosphere. It is cool because it lacks ozone.
  • The concentration of erosive nitrogen compounds in the winter vortex has been shown to depend upon both solar irradiance and geomagnetic activity with strongest correlations in the southern hemisphere.
  • Relatively high 20hPa temperature between the equator and 50° south, combined with high variability, and also the dominance of sea over land,  is the basis for the Southern Oscillation. The S.O. is the chief source of change in tropical and global temperature.
Figure 14

Figure 14

Figure 14 shows the extremes in 200hpa temperature over the period of record and therefore records variability by latitude.

  • It is apparent that variability is strongest in the southern hemisphere and the difference between the hemispheres increases from 20° to 60° of latitude. This lends weight to the observation that the Southern Oscillation is the chief source of variability in tropical and global temperature.
  • The weakness of the northern vortex is apparent in high 200hpa temperature in the Arctic. This accounts for the high concentration of ozone in the northern stratosphere. This in turn accounts for the dramatic change in Arctic, northern hemisphere and low latitude ozone concentration consequent upon a change in the northern vortex.
  • The Arctic has dramatic short term impacts on the Southern Oscillation during sudden stratospheric warmings.  However, the long term dynamic driving the southern Oscillation is the Antarctic vortex and the flux in the concentration of nitrogen oxides emanating from the mesosphere.  The strength of the vortex and the flux of nitrogen oxides can be observed to vary independently. To some extent the vortex is dependent upon tropical convection and perhaps to some extent the sun. The decline in stratospheric temperature over the equator during a sudden stratospheric warming at the pole may be due to a redistribution of the atmosphere towards the equator. Atmospheric pressure in Antarctica is remarkably low suggesting that the atmosphere is to some extent held in position by an electromagnetic field. That field is in constant flux.
Figure 15

Figure 15

Figure 15 shows that flux in 20hPa temperature at the equator follows that at 10hPa at the equator. Although planetary wave activity may play some part in modulating the oscillation in temperature and wind in the tropical stratosphere, it is solar and mesospheric influences that drive the flux in ozone concentration throughout the stratosphere. The troposphere can not do this.  It is the flux in ozone concentration that drives temperature variations in the stratosphere and cloud cover in the upper troposphere.

Figure 15  shows that instances of failure of the sea to warm when 20hPa temperature peaks (shown inside red boxes) is associated with depressed 200hpa temperature. This failure to warm when 20hPa temperature peaks is however, highly unusual. Why is it so:

  1. Sometimes the temperature increase is lagged. This may be associated with strong temperature gain in the sea remote from the equator.
  2. If this occasional slowness in response is not due to transport phenomena perhaps it is due to some quirk in the stratospheric circulation of ozone.
  3. The lagged response in 200hpa temperature near the equator is particularly evident after 2003 as the ocean cools.
  4. The lagged response may also reflect change in the way in which each polar vortex separately drives ozone concentration in the equatorial zone. At some times there is warming of the sea at 20-30° south between peaks in 20hPa temperature in the stratosphere (shown with red arrows). This can be shown to be associated with sudden stratospheric warming at the poles. Such warmings can cut in at any time during the equatorial QBO time scale. It is proof positive of the influence of polar vortex influence on upper tropospheric temperature per the medium of changing ozone content.


This presentation explains the dynamics behind the warming and cooling of the tropical ocean that is commonly known as ENSO.

It is apparent that the temperature of the tropical ocean is tied to the temperature of the upper troposphere, changing ice cloud content and therefore the flux of solar radiation into the sea. It is enhanced solar radiation that warms the sea and is ultimately responsible for increased winter temperature at high latitudes, the major manifestation of so called ‘global warming’.

The temperature of the upper troposphere is tied to the temperature of the stratosphere by change in ozone content.

The temperature of the stratosphere depends upon the flux of erosive nitrogen compounds from the mesosphere.

The flux of erosive nitrogen compounds from the mesosphere depends upon solar activity.

The next warming event should reach its peak early in 2011. A cycle of cooling is already underway. However it is at this stage still remote from the equator. On the basis of recent experience it is expected to continue for a year. It should plumb new depths for tropical sea surface temperature and have a strong impact on water temperature at high latitudes.



  1. Very few of the predictions in the “plume diagram” begin with a downward movement and very few end with an upward movement, no matter what is happening with actual SSTs. That suggests that they all have an upward trend prediction built in. I am guessing that this is based on the IPCC prediction of a rapidly warming planet.

    Given that the IPCC is almost certainly wrong (they are certainly wrong in their REASONING, having omitted solar magnetic effects entirely from their predictions, despite the .6 to .8 correlation between solar activity and temperature observed on all time scales), that means the ENSO predictions from the present should be lowered by at least the amount of the IPCC’s predicted rate of global warming.

    Am I correct in guessing that most of the ENSO predictions incorporate IPCC predictions?

    • Dear Alec,
      I am sure you are correct. While these models can not reproduce ENSO variations they are of little use to anyone. Just finished reading a great article on Australian drought at The greatest Australian droughts preceded the era of human settlement. Google ENSO and one finds references to climate change wrought by ENSO in sediments laid down thousands of years ago.

      But, I don’t know what the models have in them. I am sure its all mind bogglingly complex, mathematically challenging and relies on the assumption that the basic parameters driving natural climate change are completely static. Like you, I have a background in economics.

      The litmus test for ENSO should not be ENSO 3.4 anyway. That’s a complete red herring. The ocean off Chile is a good place to start. I can think of some rude parallels but I won’t because this is a serious discussion.

      You note the correlation between solar activity and temperature. Yes,its the influence of magnetic affects that first attracted my attention. But what has kept me going is the opposition of one particular observer who knows a lot about solar activity and magnetic matters in particular, but little about the Earths atmosphere and doesn’t seem to be interested in learning.

      It seems we face both a poverty of the imagination and a wealth of religious-ideological zeal.

  2. I’m bothered by the use of “SST off Chile” as a variable, apparently associated with quantity of heat available to be transported equatorwards in the surface ocean circulation of the SE Pacific. SST off Chile must surely be dominated by variability of coastal upwelling, unless the sampled rectangle is very large indeed. But I dont see where you specify the rectangle?

    Then, surely the SST of Nino 3.4 is dominated by the strength of local equatorial upwelling, and upwelling further east, upwelling being determined by the strength of local trade winds, not by any consequence of transport in the gyral circulations?

    Alan Longhurst

    • The sampled rectangle is indeed very large and well away from the zone of coastal upwelling. Longitude is 250 to 280 East. If you investigate the temperature change in a specific latitude band (say 15 to 30 South latitude) sampling every 20 degrees of longitude you will discover that, in the southern hemisphere in recent years SST anomalies are uniformly positive in the December to March period because the Arctic vortex is currently weak at that time leading to a global increase in stratospheric ozone content and a fall in ice cloud density. Anomalies are calculated with respect to 1948-2009 data. The other thing that would ease your mind is finding that anomalies move in unison on both sides of South America.

      It is my observation that surface temperature change is much greater away from the equator than it is at the equator. In both hemispheres most of the action is in the 30 to 50 degree latitude band, and this is especially noticeable in the north.

      SST in the ENSO 3.4 zone is atypical, exaggerated, unrepresentative and shows a strong lag factor. SST at 5N to 5S is a product of change in cloud cover elsewhere. This is a zone dominated by heavy convectional cloud cover that builds on a daily basis. Its the last place I would look if I wanted to discover the dynamics of climate change.

      As you suggest, ENSO 3.4 temperature is strongly affected by upwelling phenomena and the strength of the Trades. But the strength of the Trades is also heavily influenced by upper air phenomena driving changes in surface pressure that are closely related to climate change phenomena. So, there is a connection, albeit at a degree of separation.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: