Latest News (hidethedecline)

The Siberian-Pacific climate pendulum

Posted by Frank Lansner (frank) on 2nd January, 2012
Latest News (hidethedecline) >>


The SST of the Nino3,4 area (5S-5N / 120-170W) in the Pacific Ocean might hold information on how global temperatures evolve, at least since 1920 where the first somewhat reliable SST data from the Nino3,4 area begins. For each month a constant fraction of the Nina3,4 index is added to global temperatures from the month before, and this approach seems to re-produce global temperatures rather well. The 1000 $ question is if this relationship will remain tomorrow ?

 The relation on fig 1 -  if true - has the following consequence:

La Nina: One year of average -1 K in the nino3,4 area changes global temperatures approx - 0,094 K
El Nino:  One year of average +1 K in the nino3,4 area changes global temperatures approx + 0,105 K

Due to the above finding I carried out just a rough analysis of the climatic patterns in the Pacific Ocean. There are different writings on the subject, but i like hands on myself so i can investigate exactly what I find relevant.



In this analysis, temperature sets from the areas shown on fig2 are used.  CRUTEM3 land temperature index 50-90N / 90-210E is used for the ”SIBIR_ALASK” (red area) . For all Pacific areas on fig 2, data from HADISST1 is used. The area 50-60N/140-200E is called “NPAC_N”  (for “North Pacific, Northern part”).  More southern Pacific Ocean areas “NPAC_S” and "SPAC" will be used later in this writing.

NPAC_N and NPAC_S often show cold SST during cold PDO, and thus these areas are in the following used as simple SST-based indicators for the typical PDO patterns.

Fig3. Comparison PDO vs. NPAC_N: Trends have some similarities.


Fig 4.

The area SIBIR_ALASK on fig 2 was chosen since the typical W and NW winds is likely to affect Northern Pacific SST. On fig 4, we see variations in the area “SIBIR_ALASK” land temperatures from CRUTEM3 versus the Nino3,4 Pacific area SST. Often after Nino3,4 heat peaks, often rather similar heat peak appears in the SIBIR_ALASK area and vice versa. Definitely a “Chicken and egg” –problem: What is the cause and what is effect. In fact, the most satisfying explanation is proberly that Siberian air temperatures influences Nino3,4 temperatures with a 15-18 months lag and vice versa. Often this rhythm seem to be interrupted by some disturbance like volcanoes and more. However, the system of heat going forth and back between Siberia and Nino3,4 appears to re-establish quickly after disturbances. Volcanoes appear to take heat out of both Siberian and Pacific areas simultaneously.


The "Siberia– pacific climate pendulum":  Normally, temperature anomaly in Siberia appears 15-18 later in the Pacific Equatorial region and vice versa.




SIBIR_ALASK temperature variations are normally followed by Northern pacific SST variations (NPAC_N) which again are normally followed by variations in the Eqautorial Pacific regions, Nino3,4.


Variations in SST for NPAC_S appears more random than NPAC_N. Depending on factors like the Arctic Oscillation – the NPAC_S heat variations can occur faster or slower after heat variations occurs in Siberia. A positive AO directs the wind further south directly from Siberia to the Californian coast, and this can switch the sequence of heating between the NPAC_N and NPAC_S so that the NPAC_N occurs later than NPAC_S. (This happened around "the great climate shift" in 1978).




Besides AO and volcanic impacts, snow cover changes are likely to change the quantum of heat transferred from Siberia to the Northern Pacific possibly due to albedo change. For now, I do not have snow cover data from the specific SIBIR_ALASK area, just the full Eurasian snow cover trend. The red and blue dotted lines shows examples of how changing snow cover appear to affect changes the level of SIBIR_ALASK and Nino3.4 temperature anomalies.

Arctic Oscillation.
Let’s take a closer look at the possible impact from the Arctic Oscillation on the ability for Siberian air to flow over the Pacific.

Fig9. For positive values of AO, winds from Siberia can reach the North Pacific easily, even as for south as the Californian Coast. Since air masses can easier leave the SIBIR_ALASK areas, temperature peaks and dives tends to get smaller.

Fig10 For negative values of AO, it takes longer time for the Siberian cold air to affect northern Pacific and this can increase the SIBIR_ALASK temperature peaks and dives. Above right we see a strong negative AO temperature pattern where Northern Pacific SST is not affected by Siberian air. This is, the cold Siberian air is delayed. (Sometimes AO-delay of Siberian warm air seems to eventually generate large El Ninos.)

Thus, the NPAC_N SST is not only affected by the temperature of the SIBIR_ALASK air masses, but also by the volume of SIBIR_ALASK air masses moving towards the NPAC_N area to some degree regulated by AO.



We can thus for example illustrate the impact of AO index and SIBIR_ALASK SST on NPAC_N SST using an equation like the above. (From this, AO can be roughly estimated from SIBIR_ALASK temperatures and NPAC_N SST with similar "accuracy".)

Fig 12.

The graph on fig 12 was made using 9 sequences of data 1979-2008 where zero on the X-axis equals a minimum of the  SIBIR_ALASK area. Doing so, shows more clearly the more normal sequence of heat variation: Siberia -> North pacific -> Equatorial Pacific. Further more, SST data from South East Pacific was included “SPAC”. These show that variations in the SPAC data appear to occur mostly  just a few months before variations in Nina34 variaitons.



The complexity of temperature movement and dynamics on Earth is overwhelming. This small analysis is just a scratch in the surface to roughly show some climatic patterns to put  the stunning match between Nino3,4 SST and global temperatures fig1. in some perspective.

From the black Nino3,4 peaks to red SISBIR_ALASK peaks fig 4, heat seems to shift location like the stone of a pendulum. Heat from the Nino3,4 can normally be found mostly all over the world.  However, it seems that the heat "coming back" from just a smaller area (Siberia + SPAC etc.)  often seem to generate a heat peak similar to the previous heat peak... That is, even though heat is distributed to the whole globe, the following peak appears to have the same quantum of heat.

If this holds some truth, it suggests that a smaller amount of heat flowing to Siberia - possibly due to snow cover changes and thus albedo effects - is amplified before eventually reaching the Nino3,4 area again.

So perhaps heat collected in the Nino3,4 area from the pacific is not only reflecting some passive oscillation of heat, no, it may also reflect the newly produced heat from Siberia (and possibly other areas), which would then make Siberia be an important climate driver.

 This brings us back to fig 1:


Fig 1.

The mechanism of the “Siberian-Pacific climate pendulum” indicates as described, that heat to some degree may be produced in the process of heat transfer between the Pacific and Siberia. If so, the Nina3,4 heat peaks represents heat added to the Earths climate budget and the relation on fig 1. might make some sence. This is ment as an input for discussion - there are possibly other explanations.

If we once more consider the relations:

La Nina: One year of average -1 K in the nino3,4 area changes global temperatures approx - 0,094 K
El Nino:  One year of average +1 K in the nino3,4 area changes global temperatures approx + 0,105 K

- If true, then for example "just" a shorter period medium strong La Nina mode would cool the planet notably.

Numerous La Nina occur during cold PDO, and numerous El Ninos occur during warm PDO, and thus fig1 - if true - confirms that cold PDO cools the planet and warm PDO warms the planet.

If Siberian snow cover changes has impact on North pacific SST, then what drives Siberian Snow cover ? So even though some digging is done here, far from all questions have been answered. However, if cold PDO is typical for a period with temperature decline, we have to accept that presently we have cold PDO and thus should be experiencing global temperature decline.

Last changed: 15th January, 2012 at 21:23:27



To Jorge By Frank Lansner on 15th January, 2012 at 21:33:36
I agree, the Nino3,4 "corrected" temperature data would not change much in the overall picture because such changed (mistakedly) normally are year to year changes.

Since Nino3,4 cannot represent all oscillations in world temperatures (Some are better represented in the Atlantic etc.) Then R koefficients etc are not an easy approach: What Koefficient would be "correct" when we know that Nino3,4 cant possibly give a 1,0 koefficient as explained?
For now the match on fig. is the best argument i can think of to confirm that Nino3,4 variations are not year to year changes.
But im open for suggetions!

K.R. Frank
To the "FAIL" / unknown commenter By Frank Lansner on 15th January, 2012 at 21:27:25
Your personal opinion was answered at JoNovas:

You are welcome to answer.

K.R. Frank
FAIL By Unknown on 15th January, 2012 at 15:19:45
Figure 1 By Unknown on 14th January, 2012 at 07:57:10
Frank: I'm wondering how Fig. 1 would look without the Nina3.4 correction. I suspect it won't look hugely different, since global temperature anomalies tend to be strongly self-correlated, month to month. Do you have correlation coefficients for all Fig. 1 dependent variables on time?

I wouldn't expect strong correlation between land surface temperatures and SST's anywhere. It's just not that easy to transfer heat from gas to liquid, or vice versa.

There may be something here. I'm just not sure if the "wiggle-matching" has fully revealed it yet. Keep looking. Thanks.


Add Comment