In this post I use the Multivariate ENSO Index (MEI) supplied by NOAA at http://www.esrl.noaa.gov/psd/enso/mei/index.html and lower tropospheric temperature data supplied by UAH to show how much of temperature variation over the past 20 years is due to ENSO and how little is due to CO2. I will keep words brief and let graphics do the talking.
Firstly, here is the MEI data from 1950:
Fig. 1: Monthly MEI from 1950
As an aside, this is how it compares with SOI data. The SOI is inverted and both are scaled for comparison.
Fig. 2: MEI compared with SOI inverted
Now compare scaled MEI with Global UAH:
Fig. 3: MEI (scaled) and UAH
Notice tropospheric temperatures appear to lag the MEI by some 5 months:
Fig. 4: MEI advanced 5 months and UAH
Notice both datasets are noisy, and there is a clear discrepancy in the early 1990s. 12 month running means show this more clearly:
Fig. 5: 12 month means of UAH and MEI advanced 5 months:
The slump in UAH data is shown by the arrow. Mt Pinatubo’s main eruption was in June 1991. (Without El Chichon in 1982, there may well have been a much higher spike in the mid-1980s).
Now let’s look at the correlation between monthly MEI and UAH. Firstly, the whole period from December 1978:
Fig. 6: UAH vs MEI advanced 5 months 1978 – 2016
About 13% of temperature variation is associated with MEI variation. Doesn’t tell us much does it. What if we exclude the UAH data for two years from April 1982, and from July 1991 to December 1995?
Fig. 7: UAH vs MEI advanced 5 months 1978 – 2016 with periods after volcanic eruptions excluded
Considering the fluctuations in both datasets, that shows a fairly strong correlation.
Next, we examine the periods, before, during, and after the Pinatubo influence.
Fig. 8: : UAH vs MEI advanced 5 months December 1978 – June 1991, excluding April 1982 to March 1984
Again we see a similar correlation.
Fig. 9: UAH vs MEI advanced 5 months July 1991 – December 1995
The strong positive correlation of the previous plots has broken down.
Fig. 10: UAH vs MEI advanced 5 months January 1996 – June 2016
The correlation is even higher. Over half of temperature variation is associated with ENSO variation five months previously. Here is the same 1996-2016 plot but with 12 month running means:
Fig. 11 UAH vs MEI advanced 5 months January 1996 – June 2016, with 12 month running means
74% of temperature variation for the past 20 years and 6 months can be explained by previous ENSO variation alone. In the same period, carbon dioxide concentration at Mauna Loa has increased by 44.77 ppm, which is more than 49% of the entire increase from 1958, and Global temperature as measured by UAH has increased by a little over 0.1 degree C.
No wonder Global Warming Enthusiasts were pinning their hopes on the 2015-16 El Nino to put an end to the Pause, but they must also hope for the ENSO- temperature correlation to break down shortly, as a deep La Nina will mean cooler temperatures and further embarrassment for them. However, the correlation breaks down when volcanoes cause lower temperatures in El Nino conditions as we have seen, but what mechanism could there be for higher temperatures in La Nina conditions? Perhaps that magical greenhouse gas CO2? That would indeed be spectacular- there are no outliers at the low end of any of the above plots. The most UAH has been higher than expected with low MEI is about +0.2C to +0.3C, and those values cannot be described as outliers. Besides, UAH for June is already down to +0.34C, and we are only four months past the peak- the cooling has barely begun.
Finally, this is a plot of the centred 37 month mean MEI (because La Ninas can last for three years).
Fig. 12: 37 month centred mean MEI
Notice that before 1975 the 37 month average never exceeded +0.5, the majority of the time was in negative territory, and in the 1950s and 1970s reached below -1.0. Since 1975 the MEI has dropped below -0.5 only once in 2000 and approached -0.5 in 2012, but has been in positive territory for the vast majority of the time, exceeded +0.5 in six events, and was above +1.0 in the early 1990s. It would be surprising if global temperatures had not seen a large increase.
How low will the monthly MEI go with the coming La Nina, and how low will the following global temperatures go? All depends on La Nina’s length and strength, but the monthly MEI data are falling fast. Stand by.