MET-Office wants your money for a new supercomputer, but has given up decadal forecasting

Posted: February 17, 2020 by tallbloke in censorship, Dataset, Forecasting, History, humour, MET office, modelling, Surfacestation
Well this is a disappointment.

After the fiasco in 2018 when I revealed the data-shifting technique the MET-Office were using to never be wrong about their ‘decadal’ forecast, and the late update in 2019 , the MET-O have now disappeared the ‘decadal’ forecast altogether. This after they promised to update it in January 2020.

EDIT: The forecast has been found! See comments below.

I tried to search the MET-O website for it, to no avail. Nice picture of a windmill though.

Here’s the last iteration before it finally vanished. Retrieved from the wayback machine at

Gone but not forgotten. The MET-O ‘decadal’ forecast.

We’ll keep an eye on how well this final ‘decadal’ prediction pans out. If Ian Wilson is right, we will see an El Nino over the next couple of years. But will it soar to the heights of the MET-O’s blue blob; or fall well below? Time will tell.

  1. tallbloke says:

  2. […] über MET-Office wants your money for a new supercomputer, but has given up decadal forecasting — Tallbl… […]

  3. oldbrew says:

    They moved it to here…

    Outlook for global climate in the coming years

    Forecast issued January 2020. The forecast will next be updated in January 2021.
    . . .

    Averaged over the five-year period 2020-2024, forecast patterns suggest enhanced warming over land, and at high northern latitudes. There is some indication of continued cool conditions in the Southern Ocean. Current relatively cool conditions in the north Atlantic sub-polar gyre are predicted to warm, with potentially important climate impacts over Europe, America and Africa.

    During the five-year period 2020-2024, global average temperature is expected to remain high and is very likely to be between 1.06°C and 1.62°C above the pre-industrial period from 1850–1900.This compares with an anomaly of +1.16±0.1°C observed in 2016, currently the warmest year on record. In the absence of a major volcanic eruption a new record is likely in the coming five years, and there is a small (~10%) chance of one year temporarily exceeding 1.5°C. These high global temperatures are consistent with continued high levels of greenhouse gases.
    – – –
    Further on they say:

    ‘The forecast is for continued global warming largely driven by continued high levels of greenhouse gases. However, other changes in the climate system, including longer term shifts in both the Pacific Decadal Oscillation (PDO) and Atlantic Multidecadal Oscillation (AMO), are also contributing. Near record temperatures are predicted during the coming five years, consistent with the 2020 Met Office annual global temperature forecast. The forecast remains towards the mid to upper end of the range simulated by CMIP5 models that have not been initialised with observations (green shading in Figure 3). Barring a large volcanic eruption or a very sudden return to negative PDO or AMO conditions (which could temporarily cool climate), ten year global average warming rates are very likely to be similar to late 20th century levels over the next few years.’
    = = =
    That last sentence sounds a bit tame. Maybe the ‘climate emergency’ is over 🤨

  4. oldbrew says:

    At the foot of the MetO web page above there’s a ‘short survey’ link. First question:

    ‘Were you able to find the information you required on our new website?’

    I found it via a search engine so ‘yes and no’ to that one.

  5. tallbloke says:

    Lol. Good spot OB.

  6. oldbrew says:

    Funny how (according to the Met O) ‘continued global warming largely driven by continued high levels of greenhouse gases’ could become a ‘return to negative PDO or AMO conditions (which could temporarily cool climate)’ 🤔

    The data shows any recent warming follows El Niños, not CO2.

  7. tallbloke says:

    Yeah. The positive phases of these natural oscillations can never be responsible for warming; but their negative phases can be the cause of ‘temporary’ cooling.

  8. Chaswarnertoo says:

    Just pay Piers Corbyn, Jezza’s smart, older brother for an accurate prediction. No need for supercomputers, GIGO.

  9. Phoenix44 says:

    I am at a complete loss as to how they can come up with a 10% chance of anything in this way? This is mechanistic isn’t it?

    If they are giving probability bands because they can’t model the mechanism accurately, on what basis are they are assigning the probabilities? That their models give a 1.5 degree temperature 1 in 10 times?

  10. ivan says:

    Unvalidated computer models give garbage results especially when they leave out things like the sun and cloud cover – whoda thunk it.

    A bigger computer just gets the garbage out faster with larger error bars… woops, they don’t use those any more if they can help it.

  11. tom0mason says:

    The last time they begged for money (2009 or before?) for a new computer they said it would improve their long term forecasting, and improve the resolution of local forecasting and seasonal forecasting. Well how well did that go? IMO Seasonal and local forecasts are still as hit and miss as they’ve ever been.
    Who does the audit on the accuracy of the Met Office’s output? Who does the cost/benefit analysis of buying this technology.

    I find I can get better results from for all weather forecasting from watching/reading an amateur at . He relies on historical records, a wide knowledge of weather, gut instinct, and many models (including the M.O.) . Thankfully he steers well clear of ‘climate’ issues.

    So Met Office when will the hottest, coldest, and wettest days of this year going to occur. No, you don’t know because weather is chaotic!

  12. Saighdear says:

    Just the same OldStory, innit? Throw Money at something to make it better. I’m regularly told I’d got Loadsa money – Funny how I don’t “feel” better. Old saying, -so old I’ve forgotten where it came from – Money is the Root of all Evil