Skip to content

The Met Office’s New Decadal Forecast

Paul Homewood, Not A Lot Of People Know That

The Met Office have today published their latest decadal forecast of global temperatures, something they do every January.  The range is ridiculously large and is really saying ‘we have not got a clue.’



The forecast covers 2014-18 (no I don’t know either!), and the graph looks like this.


Tallbloke has the spaghetti graph, which is the basis for this, and as he rightly comments, it looks like kiddie’s scribble!

Nevertheless, what can we glean from it?

1) The 5-year average range comes in at 0.17C and 0.43C. Remembering that this is using a baseline of 1981-2010, last year averaged 0.20C. (Question – will HADCRUT4 now be realigned against 1981-2010, as UAH is, and as I have been arguing for a while?)

2) As Tallbloke’s version of the graph shows, the red shading is NOT “previous predictions”, as the label misleadingly claims, and which of course implies that predictions made in previous years have been pretty accurate.

They are instead “hindcasts”, meaning that the model they are running now estimates that’s what temperatures would have been under the model assumptions.

They did exactly the same last year, and after much pressure changed the label to “retrospective predictions”.

3) It is interesting that they use 90% shading, rather than 95%. If they had used the latter, there would have been times when observations dropped outside the confidence range.

4) The range is ridiculously large, and is really saying they have not got a clue.

Maybe temperatures will stay pretty much as they are, or maybe they will half a degree! In a way, this parallels the annual forecasts. For this year, for instance, they are forecasting anything between 0.14C and 0.42C (again, based on 1981-2010). (Remember, last year finished at 0.20C).

It therefore seems that they are consistently predicting far too high.

5) Nevertheless, taking the mid point of 0.30C, this would imply an increase of 0.10C from now to the 5-year average 2014-18. Assuming this builds up gradually, it implies that by 2018 the number will be about 0.20C higher.

This is really a nonsensical number. Bear in mind that even 2010 only reached 0.24C, so even another big El Nino is unlikely to see us up to their figures.

(Note – the Met Office say 0.26C, but I make it 0.24C. I am checking this with them)

6) The answer, I believe, lies in this statement they make.

 ScreenHunter_257 Jan. 30 19.21

However, the forecast initially remains towards the lower end of the range simulated by CMIP5 models that have not been initialised with observations (green shading), consistent with the recent pause in surface temperature warming.

Put simply, their models still have no explanation for the pause, so they have merrily carried on forecasting forward, as if the pause had never happened.

Full story