Response to Willis Eschenbach’s WUWT Post “How The El Nino Is Changing”

UPDATE 12/14/22: I replaced the illustrations of the 7 graphs I created for this post, to reflect the correction of a glitch in my spreadsheet that impacted one month. The correction had no effect on the content of the post or the trendlines illustrated.

. . .

Does the start date of Willis Eschenbach’s comparison graph, his Figure 5 (my Figure 1 below), in his recent post at WattsUpWithThat titled “How The El Nino Is Changing”, impact the trends toward La Niña conditions? Answer: Yes. A 9-year earlier start date flattens the trends.

INTRODUCTION

Initial Note: The title of Willis’s recent post at WUWT caught my interest, because it was ENSO related. As you may recall, back when I was blogging regularly, for many years, I wrote many dozens of posts that were cross-posted at WUWT about the El Niño Southern Oscillation, and the coupled ocean-atmosphere processes that drive El Niño and La Niña events, and the aftereffects of those events. In Willis’s post, I was hoping to find analyses and documentation of ENSO coupled ocean-atmosphere processes to show how those were changing. But what did I find? A comparison graph of ENSO indices with linear trends and a curious start date.

So, with that in mind, let’s take a look at that comparison graph, Willis’s Figure 5, which I’ve included as my Figure 1, below. His caption for that comparison graph reads,“Figure 5. LOWESS smooths of four El Nino indices, along with their straight-line trends. El Niño conditions are more positive, La Nina conditions are more negative.”

Figure 1

First: When I was blogging regularly, I had been examining, preparing, and posting graphs of ENSO indices for many years so the first thing that stood out to me was the start year, just before the 1980 hashmark. That caught the attention of at least one other person commenting on the post at WUWT, because most ENSO indices include many decades of data prior to 1980. See the comment by Richard M here and my comment here. Willis’s response to those comments about the start year was that he wanted to include the newest ENSO index, the Multivariate ENSO Index (MEI), and that its data began in 1979.

Second: Notice in Willis’s graph that color-coded index for the curve and trend for the Southern Oscillation Index data is incorrectly identified as the “Southern Ocean Index”. Also note a second typo. NOAA has a NINO3.4 Index, not a NINO34 Index.

Third: Willis has inverted the off-equatorial Southern Oscillation Index (SOI) data. Normally, Southern Oscillation Index (SOI) data is presented without being inverted so that El Niño spikes are downward and La Niña spikes are upward, which is the opposite of how they appear with the other indices. Of course, he inverted the SOI data so that he could include that off-equatorial ENSO index in his comparison graph.

Fourth: Following his Figure 5, Willis states in the text of his post:

“You can see the peaks representing the big El Ninos around 1997-98 and 2015-16. Recall that according to my thermoregulatory hypothesis, the Pacific should be trending towards a more La Nina condition which is more negative.
“And all four indices, in varying amounts, show this exact outcome—in response to the slow gradual warming since 1980, we have more La Nina conditions cooling the planet.”

I had never heard of Willis’s “thermoregulatory hypothesis” prior to reading his recent post (linked above). Therefore, it will be up to you readers to comment on whether the flattening of the trend lines when the data for the ENSO indices starts in 1970, as shown in the graphs below, has any impact on his “thermoregulatory hypothesis”. I also have a question for you readers about the above quoted paragraphs from Willis’s post. Why did he identify the strong 1997/98 and 2015/16 El Niños, but not the strong El Niño of 1982/83, or the strong one of 1987/88 (portion of the 1986/87/88 El Niño), or the strong 1991/92 El Niño? Those other three El Niño events all exceeded NOAA’s +1.5 deg C threshold for a strong El Niño.

Fifth, regarding the start year of Willis’s graph, I stated in a comment on that WUWT thread that the 1980 (1979) start year was curious, considering that it was a few years prior to the strong 1982/83 El Niño. Further on that thread, I suggested to Willis that he exclude the data for the Multivariate ENSO Index (MEI) and he start a set of reference graphs in 1970 to include the La Niña dominant early-to-mid 1970s in an effort to see what happened to the trends. I advised him that I would present them, if he chose not to. He chose not to, as far as I know, so I did. They follow.

So, let’s start the presentation.

A FEW NOTES ABOUT THE FOLLOWING GRAPHS

The following three pairs of time series graphs are provided to show the easily discernable flattening of the trend lines when the start date for the ENSO index graphs is changed from 1979 (used by Willis Eschenbach in his post at WUWT) to 1970. In the graphs starting in 1970, the strong 1972/73 El Niño and the three La Niña events (one moderate, two strong) in the early-to-mid 1970s really stand out in the graphs of the ONI and NINO3.4 data…not so much in the SOI data.

Why did I use 1970 as the start year? The early to mid 1970s are dominated by La Niña events, as are the most recent years. In other words, with the 1970 start year, the graph starts and ends in periods dominated by La Niña events, making it difficult for anyone to accuse me of cherry picking the start year.

I didn’t use any smoothing on the data in the graphs. That would reduce the magnitude of the ENSO variations and make the trends appear greater by comparison. Additionally, as I presented the data for each index individually, there was no need to standardize the data. As a result, the graphs are of data as provided by the suppliers, which are linked in the following discussions.

OCEANIC NINO INDEX (ONI) GRAPHS

Figures 2a and 2b present the unsmoothed monthly Oceanic Nino Index (ONI) data (data here), with 2a starting in January 1979 and 2b with the data starting in January 1970. With the 1979 start (Figure 2a), there is a noticeable trend from positive values to negative, but with the 1970 start (Figure 2b) the trend line flattens greatly.

Figure 2a

Figure 2b

NOAA NINO3.4 INDEX GRAPHS

Figures 3a and 3b present the unsmoothed monthly NOAA NINO3.4 Index data (data here), 3a with the data starting in January 1979 and 3b with the data starting in January 1970. With the 1979 start (Figure 3a), there is a very minor trend from positive values to negative, but with the 1970 start (Figure 3b) the trend changes sign to a very minor trend from negative to positive.

Figure 3a

Figure 3b

SOUTHERN OSCILLATION INDEX (SOI) GRAPHS

A reminder: The noisy, off-equatorial, sea-level-pressure-based Southern Oscillation Index (SOI) data shows El Niño events as downward spikes and La Niña events as upward ones…the opposite of the sea surface temperature-based ENSO indices. Figures 4a and 4b present the unsmoothed monthly Southern Oscillation Index (SOI) data (data here), 4a with the data starting in January 1979 and 4b with the data starting in January 1970. With the 1979 start (Figure 4a), there is a noticeable trend from negative (El Niño) values to positive (La Niña) ones, but with the 1970 start (Figure 4b) the trend flattens very noticeably.

Figure 4a

Figure 4b

A FINAL GRAPH: UNSMOOTHED MULTIVARIATE ENSO INDEX (MEI) DATA SHOWS SOMETHING VERY CURIOUS

Figure 5 presents the unsmoothed monthly Multivariate ENSO Index (MEI) starting in January 1979 (data here). A couple of things stand out: Note how the strong El Niño of 2015/16 has a lesser peak value than strong 1982/83 and 1997/98 El Niños, while with the sea-surface-temperature-only-based ENSO indices (ONI and NINO3.4 Index) show the 2015/16 El Niño peaks at a higher value than the other two. Note also how the La Niña events toward the end of the MEI data have much greater negative values than the ONI and NINO3.4 data, while toward the beginning, the La Niñas have lesser negative values.

Figure 5

The MEI data appears very skewed toward a negative trend compared to the sea-surface-temperature-only-based ENSO indices; therefore, the next time some alarmist says that strong El Niños are growing stronger due to global warming show them a graph of the MEI data.

CLOSING

In closing, here’s a very brief introduction to El Niño events from one of my recent short stories. It was also included in a 2019 post here at WUWT:

Most news stories about El Niños call them unusual warming events in the eastern tropical Pacific Ocean, off the west coast of South America. They’re the cause of the huge upward spikes we see in the global surface temperature graphs.”

They are much more than just warming events, and, further, regardless of what the numbskull science reporters say, there’s nothing unusual about them.” … “Magnificent would be a better word. Here are the facts. El Niño events occur every two to seven years. El Niños are the most-amazing, and the most powerful, weather events ever devised by Mother Nature. How powerful? El Niños are often kick-started by series of tropical storms in the western tropical Pacific.

Further from that post:

El Niño and La Niña events act together as a chaotic, naturally occurring, sunlight-fueled, recharge-discharge oscillator, with El Niño events acting as the discharge phase and La Niña events acting as the recharge phase...

I hope you enjoyed this post. Have fun.

Regards,
Bob