Posted by: erl happ | June 29, 2009

The Southern Oscillation. The young persons guide to climate change.

Sea surface temperature between the equator and 30° latitude

Figure 1 Sea surface temperature between the equator and 30° latitude. Data: 12 month mov. av. centered on seventh month.

Figure 1 indicates that the ocean between the equator and 30°north is about 1°C warmer than the equivalent zone south of the equator. In northern summer the atmosphere is warmed by radiation from the continental land masses. As the atmosphere warms cloud disappears  allowing more light to reach the sea. Despite the strong annual cycle of warming in the north it is the southern waters that show the larger variation on the inter-annual, decadal and longer time scales. It is the south that has warmed the most over the period since 1948. Much of the warming in southern waters occurred between 1976 and 1980. Since 1998 southern waters have cooled faster than northern waters.

Figure 1 shows the Southern Oscillation in action. What causes it? Climate science is has no answer to this question. If it had an answer it might not be so morbidly obsessed with ‘greenhouse gases’. This article explains the working of the Southern Oscillation and shows how it is responsible for climate change.

Greenhouse theory posits that upper troposphere water vapour and ice cloud amplifies surface warming. The counter argument is that the presence of ice cloud cools the surface because it reflects solar radiation. With less cloud the surface warms. The two viewpoints are diametrically opposed. As we shall see, this element of  greenhouse theory is just plain wrong.

If :

  1. Surface temperature rises when high altitude ice cloud disappears.
  2. Ice cloud diminishes when the air containing the cloud warms.
  3. 200hpa temperature is indicative of the temperature of the high altitude where ice cloud forms.
  4. 200hPa temperature varies with ozone content.
  5. Ozone content at 200hPa depends upon solar activity.

Then, an entirely different scenario of climate change evolves, where ozone variability is responsible for the Southern Oscillation.

Until recently it has not been possible to demonstrate the response of 200hpa temperature, sea surface temperature and ice cloud density to a change in ozone content. But gridded monthly ozone data for the period since 1979 has recently been made available at:

Location under examination is 20-40°S , 80-10°W

Location under examination is 20-40°South , 80-100°West

I want to look particularly at the zone of highest atmospheric pressure in the south East Pacific. This is at 20-40°south, 80-100°west. The pressure difference between this zone and the maritime continent (Indonesia) drives the Trade winds across the Pacific. The increase in sea surface temperature in this zone frequently leads the ENSO 3.4 region. The decline in atmospheric pressure in this area leads the increase in tropical temperature as measured in the ENSO 3.4 region.  The high pressure cell that occupies this zone is a near permanent fixture determining atmospheric dynamics across the Pacific. Pressure in the west is relatively invariable but in this zone it is very variable.

The Southern Oscillation Index is computed as the ratio between the atmospheric pressure at Tahiti (circled in the map) and Darwin. But the atmospheric pressure off Chile is normally higher than at Tahiti and 200hPa temperature near Chile is colder than it is over Tahiti, in part because the surface waters are colder but in part also due to the downdraft from a colder tropopause.

Sea surface temperature and total column ozone

With respect to figure 2 the tendency of the 12 month moving average of sea surface temperature to peak when total column ozone peaks in mid winter is of interest.  This is when the ocean is coolest and cloud cover is most extensive.

Mid winter is the time when the pressure difference between the south east Pacific and Indonesia is greatest. It is therefore the time when a fall in pressure in the East alters atmospheric parameters in the most influential fashion. If a window is opened through the winter cloud to let in more solar radiation the sea warms. A warming of the atmosphere at 200hPa and also at the surface, weakens surface pressure. This loss of pressure in the eastern Pacific and a failure of the trades marks  all tropical warming events.

Run your eye across figure 2. Does the 12 month moving average of sea surface temperature tend to peak at the same time as ozone peaks in mid winter and is there not a rough two year frequency between peaks that relates to the QBO in equatorial ozone, temperature and stratospheric wind? I will now dissect this relationship further.

Figure 2, relationship between total colmn ozone and sea surface temperature 20-40°S 80-100°W

Figure 2. Relationship between total column ozone and sea surface temperature 20-40°S 80-100°W Monthly data for ozone. 12 month mov. Av. for SST.

Look now at figure 3.

Figure 3. Relationship between sea surface and 200hPa temperature

Figure 3. Relationship between sea surface and 200hPa temperature. Monthly data.

Peaks in 200hpa temperature occasionally occur in the middle of winter when the sea surface is coolest. This is quite anachronistic. Patently, something other than surface temperature is driving 200hPa temperature.

Look now at figure 4. There is a strong response of 200hPa temperature to the winter ozone peak. The most obvious instances are marked with an arrow. The very regular peak in 200hPa temperature in February-March is associated with highest sea surface temperature at that time. But, in 1983 and 1997 (two big El Nino years) the relativity between these two peaks in 200hPa temperature was reversed. The larger peak occurred in mid winter. In the case of the 1983 El Nino the 200hpa peak in the summer of 1982 was also outstanding, as was the winter ozone peak that preceded it.

Figure 4. Total collumn ozone and 200 hPa temperature

Figure 4. Total column ozone and 200 hPa temperature. Data is monthly.

20hPa temperature at the equator is a good proxy for local ozone content. Look now at figure 5. There is a clear association between 20hPa temperature at the equator and sea surface temperature in the East Pacific ( longitude 60-80°west). In three  instances sea surface temperature increased prior to the peak in 20hPa temperature over the Equator. In fifteen instances the peaks are conjunctional and in 3 instances there is a short lag between the 20hPa peak and the sea surface peak.In making these judgments I am looking primarily at SST between 20-30° south.

Latitude 30-40° south only  came into play after 1978 and it has been a strong driver of sea surface temperature change since that time. There are only three instances where a peak in 20hPa temperature is not associated with a peak in sea surface temperature. These occasions followed strong El Niño’s and it is probable that the resulting precipitation event from a strongly heated ocean generated so much cloud as to negate the response. In each of these cases a peak in sea surface temperature can be seen at 30-40°S even though it does not appear at other latitudes. So, the smothering effect of a long lasting La Nina precipitation event is greatest close to the equator.  The most outstanding illustration of this effect is the year 2000 where a strong sea surface warming at 30-40°S occurred as  latitudes closer to the equator actually cooled. Finally there is a single instance where two 20hPa peaks morphed into one large warming event. That occurred between 1971 and 1973.

Figure 5. Sea surface temperature at 20-40°S and 20hPa temperature over the equator.

Figure 5. Sea surface temperature at 20-40°S and 20hPa temperature over the equator. Data: 12 month mov. av.

This diagram tells us that mid stratosphere temperature over the equator is a better guide to warming events than total column ozone at higher latitudes. It appears that ozone flows out from the equatorial zone into the downdraft areas of the upper troposphere between the equator and 40°S. This is the region described as the Hadley Cell. It is a region where high cloud streams away from the equator and there is little low cloud, except at the inter-tropical convergence zone.

Figure 6 shows the relationship between 20hpa temperature at 10°N to 10°S and 200hpa temperature at 20-40°S , 80-100°W. The relationship is loose but nevertheless coherent.

Figure 6. Relationship between 20hPa temperature over the equator and 200hPa temperature at 20-40°S , 80-100°W

Figure 6. Relationship between 20hPa temperature over the equator and 200hPa temperature at 20-40°S , 80-100°W. Data: 12 month mov. av.

There can be no doubt that ozone content of the upper troposphere is a strong driver of temperature change at 200hpa Since the water vapour content of upper troposphere air is relatively invariable, ice cloud density will fall when 200hPa temperature rises. So, the rise in sea surface temperature in mid winter is no mystery. But of course, it confounds greenhouse theory and this phenomenon will be seen as an anomaly by those who espouse the notion that high cloud must warm the Earth. This is plainly a head in the sand response. High cloud cools the earth. Its disappearance is marked by an increase in surface temperature. That comment applies to the latitude 40°S to 40°North. It is in this latitude band that the temperature of the Earth is determined.

What have we learned thus far?

Change in total column ozone drives the Southern Oscillation by increasing 200hPa temperature and evaporating high altitude ice cloud. The greatest potential for this to occur is in winter. There is a conjunctional relationship between 20hPa temperature over the equator and sea surface warming events in tropical latitudes. As a result of the loss of upper troposphere ice cloud:

  1. The temperature of the waters that are driven towards the equator by the trade winds rises.
  2. The temperature of the cold waters of the northwards trending currents on the western side of the southern continents rises.
  3. The area where this occurs is clearly between the equator and 40°south.
  4. The warming frequently begins remote from the equator and in some instances it occurs only there.

The ocean off Chile is a proxy for what happens  in the Indian and Atlantic oceans and across the entire latitude band 10-40°S. Change may also be initiated in the Indian or the Atlantic Ocean.  In truth there are strong high pressure cells across the 20-30° latitude band right round the globe in both hemispheres but they are far more extensive in the south. There is another factor of great importance in explaining the power of the Southern Oscillation. There is more sea surface to warm in the south than the north.

The Accelerator

Once the equatorial water warms significantly,  uplift starts a train of precipitation, convection and de-compressive cooling (OLR falls) above the equator. What goes up must come down. Compressive heating then occurs between 10 and 40° north and south, further eroding cloud cover (OLR rises) stimulating the rise in sea surface temperature. This means that the big high pressure cells get larger and they are less cloudy.

It is confidently asserted that tropical sea surface temperature and thereby global climate depends heavily upon stratospheric ozone, high cloud cover and light flux into the ocean, particularly in winter.

Response of sea surface temperature to stratospheric warming at the poles

The response to the sudden stratospheric warming in February-March 2009 in the temperature of the sea off Chile is seen in figure 7. A most interesting aspect of a sudden stratospheric warming is the immediate increase in ozone concentration in the stratosphere/upper troposphere. This dynamic, and a charting of its extent, seems to have escaped the attention of ‘climate science’.  It’s not the sort of story that ‘warmers’ want to hear. How important is this phenomenon?

Figure 7. Sea Surface temperature change in response to the February-March 2009 sudden stratospheric warming in the Arctic

Figure 7. Sea Surface temperature change in response to the February-March 2009 sudden stratospheric warming in the Arctic. Monthly data.

Figure 8 answers that question in part. The stronger variability of 1hPa temperature at 60-90°S between June and February shows when stratospheric warmings have impacted the southern vortex in the past. Notice that temperature at 1hpa has bumped along at the lower limits of the record (record is only for the period since 1979) in both 2008 and 2009. We are now (in June 2009) emerging from a brief warming event. There was another in October-November last year and the major event in the Arctic in February-March. All these events have been associated with brief increases in stratospheric ozone and warming of the subtropical and tropical ocean.

Figure 8 Variability in 1hPa temperature 60-90°south latitude

Figure 8 Variability in 1hPa temperature 60-90°south latitude

But, what has happened to ozone levels over the years?

Long term change in ozone and stratospheric temperature

Figure 9 shows that, after rising gently from 1948 to 1976, then quite abruptly between 1976 and 1980, temperature in the southern stratosphere has fallen gently but continuously until the present time. Surface atmospheric pressure in the sea off Chile, while fluctuating as ever, has gradually risen (since 1983, see my last post). This strengthens the trades and the La Nina tendency. By 2003, it appears to have tipped the balance between net warming and net cooling of the southern ocean. The result is apparent in figure 1 in a cooling of the southern tropical ocean. The north will follow as surely as night follows day.  It will cool more swiftly if atmospheric specific humidity can recover the losses of recent decades. Perhaps that will happen if the tropics cool and there is less convection cannoning water vapor into the stratosphere.

Figure 10. Change in10hPa temperature in the southern hemsiphere

Figure 9. Change in10hPa temperature in the southern hemsiphere. Data: 12 month mov. av.

The cooling of the stratosphere after 1980 has nothing to do with increasing greenhouse gases, just as its warming prior to that date had nothing to do with declining greenhouse gases. It has a lot to do with the chemistry of the mesosphere and the thermosphere.

The concentration of nitrogen oxides (these substances degrade ozone) in the thermosphere and the mesosphere is governed by solar activity, both in respect to irradiance and geomagnetic activity. So ozone can be diminished in the mesosphere, at and below the Stratopause, and via the vertical winds at the poles (stronger during the polar night) and via the stronger and more prevailing Antarctic vortex.

The ozone content of the stratosphere is a primary determinant of its temperature. The decline in the temperature of the stratosphere since 1978 is consistent with the decline in ozone content, and stratospheric temperature, seen  in many figures in this presentation

How is ozone concentration nearer the equator modulated?

There is a long standing theory that ‘Planetary Waves’ are responsible for the biennial reversal of the stratospheric winds over the equator and some maintain that these waves are also responsible for sudden stratospheric warmings. Associated yes. Causative, no. Figure 10 shows us that the atmosphere itself modulates 20hPa temperature above the equator. The propagation is from the pole towards the equator. Atmospheric damping simplifies and consolidates the signal so that it approximates  a simple sine wave, and some of the pulses that propagate from the pole are lost in the process. The average period from peak to peak is 27.1 months. We have already seen in figure 5 that 20hPa temperature over the equator is the best predictor of sea surface temperature . We can also see that the globe is currently headed into a cooling cycle. The next warming cycle will not peak for about two years. If it is a large one, as is usually the case in the upswing of the solar cycle, it will be followed by a long and deep La Nina that will smother the next warming event so the one after that may be six years away, peaking in 2015.

Figure 10. Propagation of ozone cnd temperature change between 20-40°S latitude and the equator at 20hPa

Figure 10. Propagation of ozone and temperature change between 20-40°S latitude and the equator at 20hPa. Data: 12 month mov. av.


Per medium of ozone, the sun drives the Southern Oscillation. It drove the rapid warming between 1976 and 1980. It has allowed the gradual cooling of the stratosphere since that time. There has been a slow reduction in the intensity of tropical warming cycles and a progression of slightly deeper minima, the direct consequence of diminishing ozone in the southern stratosphere. This process  produced the cooling of the 1970’s and the abrupt warming of 1976 to 1980. It is responsible for the long series of strong El Ninos between 1976 and 1998 that is associated with the well documented expansion of the Hadley cell in the southern hemisphere. It is also responsible for the cooling since the turn of the century.

The future

Whereas the chemistry of ozone control by nitrogen oxides seems to be fairly well understood the dynamics behind the strengthening and weakening of the polar vortexes is not. As geomagnetic activity weakens with the solar wind, the concentration of nitrogen oxides in the mesosphere must fall. However, it appears that the polar vortex has strengthened. If we want to properly understand the suns role in climate change the waxing and waning of the vortex must be explained. A clue to the dynamics that drive the vortex may be in the observation that a sudden stratospheric warming at the pole is preceded by a marked cooling at 10hpa in the tropical stratosphere. This suggests a temporary thickening of the atmosphere over the equator and thinning over the pole. Is the solar wind responsible for a shift in the ionised atmosphere via electromagnetic acceleration? The solar wind has been weakening for decades. If the vortex continues to strengthen the ozone hole will expand and the stratosphere will cool. However, the influx of moisture  into the tropical stratosphere will diminish as the ocean cools and ozone concentration  will increase. This may well lead to more exaggerated swings in the Southern Oscillation. A cooling of the tropical ocean would tend to promote a situation of greater sensitivity over a narrower band of latitudes. The Hadley would contract, perhaps to its extent prior to 1976.

Figure 5 shows that we have just experienced a peak in sea surface temperature in the Ocean off Chile. The next is due, if the law of averages is to be our guide, in about two years. As I write the Southern Oscillation index, having headed strongly into El Nino territory in recent weeks, (perhaps under the influence of a stratospheric warming in Antarctica in May), is now telling us that the next cycle of tropical and global cooling is under way. I am well aware that BOM and NOAO are predicting El Nino. They are wrong. Improvement is possible. They must look to ozone, the stratosphere and the sun.

These are exciting times. In a few days we can chart the descent of temperature with data for the month of June at:

All temperature data for this presentation was drawn from NCEP/NCAR reanalysis at that source.



  1. I find the ozone connection intriguing, and your statement: “Per medium of ozone, the sun drives the Southern Oscillation”. I have a question regarding an observation I made:

    I explored ozone here:

    Half way down the page I have this: The following figure shows the Arctic ozone in March for 1979 – 2007 (left) [] and (right) comparing that to the multivariate El Nino / Southern Oscillation Index (MEI) with ozone lagged by one year (i.e. ENSO leads by one year). The MEI plot is from [].

    (The two figures from the page are here and

    Question: Why does the figure referenced above indicate that ENSO leads ozone if the sun drives the SO?

    (I would be happy to continue the discussion via email if you prefer.)

    • Alan, I am happy to exchange emails on this but equally happy to discuss it here. I am having trouble raising the documents you refer to but here are some comments that I think are relevant.
      What I maintain is that changes in ozone drive sea surface temperature across the globe. This is per medium of changing cloud cover (high level ice cloud from 200hpa upwards). An increase in ozone results in an increase in temperature, falling humidity and less cloud. Specific humidity above 200ha does not vary a great deal. But temperature varies a lot, much more than at the surface. The relationship between stratospheric temperature and sea surface temperature is not lagged.
      The change in ozone is driven by the changing strength of the polar vortex in both hemispheres. ‘Arctic ozone in March’ does not cut the mustard.
      The MEI includes variables other than sea surface temperature and it relates only to the ENSO 3.4 region in the mid Pacific. Sea surface temperature in this region is unrepresentative of global tropical temperature and equally unrepresentative of global temperature change, which depends upon SST change at all latitudes.
      So, the comparison you refer to is unrepresentative of SST and of ozone content.
      When I refer to the sun driving the Southern Oscillation I am simplifying the argument. The SO is important because of the area of ocean that is affected. The Southern Ocean is a big contributor to global tropical temperature. The SO very much depends upon the Antarctic vortex in driving long term change in stratospheric ozone. But the Arctic vortex is also important because it operates in northern winter, southern summer when the Earth is closest to the sun and the sun is more directly overhead. But, the big changes are driven in southern winter and come via change in the Antarctic vortex.
      I have a post coming up that looks at sea surface temperature change at all latitudes. Currently positive anomalies are experienced in the December to March period and are therefore much driven by the Arctic vortex which is much weaker than in the past. In another forthcoming post I look at the relationship between stratospheric temperature and surface pressure at high latitudes and at the equator. Change in stratospheric temperature is sudden, continuously evolving and distributed throughout in a brief interval, a month or so. Change is initiated and has the greatest amplitude at the poles. Stratospheric temperature is driven by changing ozone content.


    • Then, feel free to enlighten me.

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: