Super computers can be great machines, but their use by NOAA in the U.S. and by the British Met Office demonstrate the old saying “garbage in – garbage out.”
Weather and climate are complex. Even with supercomputers, if the wrong assumptions are input, then the results are often wrong. Let’s see how NOAA and the Met Office did with winter forecasts this year.
The graphic below was made last fall by NOAA showing their predictions for winter temperatures in the U.S. The orange with an “A” shows areas that NOAA predicted would be above average, the blue with a “B” is for below normal temperatures, and the white area with “EC” are equal changes of being above or below normal.
So much for predictions. The conditions that really happened are shown on the graphic below. (Source: NoTricksZone). The blue area is colder than normal; green is much colder than normal.
Remember, these are the people who claim they can predict climate change 10 to 100 years into the future.
And, (from NoTricksZone): “This year western Europe has experienced a mild winter as a parade of low pressure systems coming in from off the Atlantic has fed the continent with a steady supply of mild southerly winds. For Germany this winter will be the first mild one in 6 years after a record 5 consecutive winters of colder than normal winters.”
This year was not the first time that the Met Office got the forecast very wrong. Back in 2012, when they touted their new supercomputer which is capable of 1,000 billion calculations every second, and uses 1.2 megawatts of energy to run – enough to power a small town, the head of the Met Office claimed that this new computer “will enable the Met Office to deliver more accurate forecasts, from hours to a century ahead.” (See my post “British supercomputer botches weather forecasts”) As it turned out, spring in 2012 was one of the wettest on record in the UK.
I pick on 2012 because my wife and I happened to be traveling in the UK that June. We rather enjoyed the cool and wet weather as a contrast to an Arizona June which is hot (35-45̊C) and dry.
Results show that even with super computers, predicting the weather is a tricky business. Government agencies seem to make assumptions based on political science rather than real science. Perhaps they should consult the Old Farmer’s Almanac more often. That publication claims to get the long-range forecasts right about 80% of the time.