Our ENSO-temperature paper of 2009 and the aftermath

02 Feb 2011 (updated 27 May 2020, 18 Aug 2023 to correct a link)  Category: GENERAL

In July 2009 the Journal of Geophysical Research published our paper titled "The Influence of the Southern Oscillation on Tropospheric Temperature". We showed that the Southern Oscillation Index (SOI, calculated according to the Troup method, was a good indicator of global average lower tropospheric temperature 7 months later except when volcanic eruptions around the Pacific Ocean caused cooling. The paper in PDF format is available here.

Our brief errata was published shortly afterwards to correct two captions and a label on a figure:

The caption to Figure 3 should be: "Correlation coefficients for derivatives of SOI and GTTA with different time lags". The caption to Figure 4 should be: "Derivatives of SOI (dark line) and RATPAC-A GTTA (light line) for the period 1960-2008 with principal volcanic activity indicated". In Figure 7, the dark line in (b) and (c) should be labelled "SOI (7-month lag)", as described in the caption.

The key to our paper was Figure 7, in which we graphed the monthly average lower tropospheric temperature with the Southern Oscillation Index of seven months earlier.

The image below is of Figures 7(b) and (c), along with the caption. Figure 7(a) is omitted here but shows a similar sustained relationship between SOI and RATPAC-A data obtained by weather balloons. This data was important because it spanned 1977, the year in which recent warming began and clearly that warming was related to the SOI of mid-1976. (Even though it was clearly labelled and very different due to the data being at 3-month intervals, some precious people falsely claimed that we were splicing different graphs together.)


The key image of our paper

The consistently close relationship in these graphs, especially the failure of temperature line to steadily rise above the SOI line, suggest that any warming caused by carbon dioxide is negligible.

Not surprisingly our finding of a strong and sustained close relationship between time-lagged SOI and temperature upset a few people. One such person was Grant Foster (who hides behind the nom-de-web "Tamino"). Within days of the publishing of our paper he had written a criticism and posted it to his web page, and shortly thereafter it was posted with very few alterations but formatted in the style of the journal on the web pages of Kevin Trenberth. Foster was the principal author but now J.D. Annan, P.D. Jones, M.E. Mann, B. Mullan, J. Renwick, J. Salinger, G.A. Schmidt and K.E. Trenberth were also claiming to be authors.

About 3 months later we were advised that the journal would publish this criticism/comment and our response was invited so that it might be simultaneously published. We pointed out that the comment had already appeared on the Internet in the journal's format and therefore did not comply with the journal's regulations. We said that we would respond to it but asked that the comment's authors be publicly rebuked for this breach of regulations. The journal editor ducked and weaved, declaring at one point that posting on the Internet was not publishing (then what is it?), then admitting that the Internet posting had used the journal's boilerplate (which is a format template) and then later quoting the journal regulations as saying that unformatted papers [my emphasis] could be posted to the Internet prior to publication. Ultimately our request was ignored, and so to were some of journal's own regulations.

The journal's regulations also say that a response to a comment will be reviewed by just one person and the aim is to confirm that the substance of the criticism has been properly addressed. Our response was sent to three reviewers and they aimed their criticisms at our original paper, to which the journal responded by refusing to publish our response.

What I've mentioned here is just the tip of the iceberg. For more details see Censorship at AGU: Scientists denied Right of Reply (PDF).

And was our paper correct? Observational data seems to confirm predictions that Bryan Leyland of New Zealand made based on our paper.


Comparison of delayed SOI and global average temperature anomaly since yr2000

As with the above graphs, the match isn't perfect, but that's because a variety of forces act over various time spans on the ENSO and temperature. Carbon dioxide can hardly be the cause of "additional warming" during and since 2010 when it caused no warming earlier.