Skip to content

Atmospheric Temperatures & El Ninos

October 15, 2014

By Paul Homewood 





It is generally accepted that there is a time lag between an El Nino event and its effect on global temperatures, with a consensus of between 3 and 6 months.

This does not only apply to atmospheric temperatures, but also surface ones. For the heat released by an El Nino to be distributed around the world, it first needs to find its way into the atmosphere.

The question then is have we still to see a peak in atmospheric temperatures, as measured by satellite?

First, let’s look at the Multivariate ENSO index (MEI) from the ESRL.





Anything over 0.500 is regarded as El Nino. So far this year, the MEI has peaked in April/May, although there may be signs that it is returning.

Now let’s look at when the various global temperature datasets have peaked, both this year and also in the major El Nino years of 1998 and 2010. (UAH and RSS are the satellite datasets, measuring lower troposphere temperatures, while GISS is a surface dataset).


1998 Feb April April
2010 March March March
2014 May May June


Although GISS peaked two months earlier in 1998, they all peaked together in 2010. It is reasonable, therefore, to assume that this year UAH and RSS will peak up to two months later, but no more.

We actually find that all three have peaked in May and June, i.e. similar to 2010.


We may find that the El Nino comes back to bite us over the NH winter, but at this stage it seems reasonable to say that the peak for this year has passed.

If this is so, the satellite dataset anomalies for 2014, not to mention 2015, will remain well below the levels seen in 1998 and 2010.


1998 0.42 0.55
2010 0.40 0.47
2014 Sep YTD 0.25 0.25

Temperature Anomalies C

One Comment
  1. David permalink
    October 16, 2014 7:32 am


    Going back to the start of 2013 the best correlation I could get from the raw monthly data was between MEI and RSS (UAH wasn’t as close) with a one month lag. If that pattern continues then you may well be right, because that would imply a fall in RSS for October.

    Going back to Jan 1998, the best fit I got between MEI and RSS was by smoothing the data with a centred running 13 week average (retaining the 1 month lag). That gives a slightly different picture, because it finishes with MEI still on the up and with RSS following, though rather sluggishly.

    There could well be a short term fall in RSS, but I suspect the full lower troposphere response to the recent SST increase hasn’t yet fully developed and could still come over the next few months. Having said that, it doesn’t look like it’s going to be as large as I’d thought. We’ll soon see.

Comments are closed.

%d bloggers like this: