Guest Post by Willis Eschenbach

In my peripatetic meandering through the CERES satellite data, I’ve been looking at the correlation between the temperatures in the NINO3.4 region and the temperatures of the rest of the planet.

The NINO3.4 region is an area in the equatorial eastern Pacific Ocean. It covers five degrees north and south of the Equator, from 170° West to 120° West. Temperatures in that area are used to measure the strength of the El Nino / La Nina phenomenon.

Now, people often discuss procedures like “removing the effects of the El Nino from the global temperature record”. What they mean is that they have noted the similarity between the temperature of the NINO3.4 region and the global temperature. Figure 1 shows that relationship as seen in the CERES data.

Global and enso34 surface temps

Figure 1. Surface temperature of the globe (blue) and of the central Pacific ENSO3.4 area (red). Note the large “El Nino” event at the end of 2015. Both datasets are normalized (set to a mean of zero and a standard deviation of one).

Seeing this relationship, people have “removed” the NINO3.4 temperature variations from the temperature record. They have done this by subtracting out, in one form or another, the variations that are “caused” by the El Nino swings. However, I have disagreed numerous times with this procedure. Let me propose a more encompassing way to understand the relationship shown in Figure 1.

This is to note that although there are areas of the surface which show a good positive correlation to global temperatures, there are also areas that show a good negative correlation to global temperatures. Figure 2 shows this relationship on a gridcell by gridcell basis. It displays how well the temperatures in each gridcell agree or disagree with the global average temperature variations shown in Figure 1.

correlation 0 month lag global and gridcell temps

Figure 2. Correlation of each gridcell with the global average temperature. Note the large areas of negative correlation (green and blue)

Looking at that, I ask you to reconsider the idea that we can simply subtract out the temperature variations in the NINO3.4 area (blue box) from the global temperature … clearly, the relationships are far from simple. Is the fact that certain areas correlate well with the global variations a sufficient reason to “remove” them from the temperature record?

And if so, why limit ourselves to the ENSO3.4 area of the Pacific? Why not use a much larger area of the Pacific and “remove” half of the Pacific from the temperature record?

Setting those questions aside, the overall pattern in the Pacific is clearly related to the heat which is moved by the El Nino / La Nina pump. These two phenomena act together to pump warm Equatorial water across the Pacific in a westward direction. Once this warm surface water hits the Asian mainland/islands it splits and moves toward the two poles.

Now, many people say that this shows that the El Nino / La Nina is causing the global temperature changes. I say that the causation is going the other way. When the earth warms and excess heat accumulates in the eastern tropical Pacific, it triggers a cycle of the El Nino / La Nina pump. This pump moves warm water to the poles, where it is lost to space. Overall this cools the planet. The results of this pumping action can be seen in Figure 2 as the green areas in the western Pacific heading towards the north and south polar regions.

In other words, the El Nino doesn’t control the temperature—the temperature controls the El Nino.

We can look at this from another perspective. Rather than comparing gridcells to the average global surface temperature as in Figure 2, we can compare gridcells to the average NINO3.4 area temperature. Figure 3 shows that result.

correlation 0 month lag enso and gridcell temps 2
Figure 3. As in Figure 2, but comparing ENSO3.4 area temperatures with gridcell temperatures. Note different color scale than that used in Figure 2.

Again, a most interesting result. It makes the El Nino pattern even clearer. Note that both the Western Pacific and the North Atlantic move in opposition to the NINO3.4 area.

The source of the pattern seen in Figure 3 is clear. It is driven by the El Nino / La Nina pump. When enough heat has accumulated in the eastern Pacific, the El Nino / La Nina pump pushes warm water first westward, then poleward. This cools the eastern Pacific and warms the…