Andrew Blum explores how technology ensures weather forecasts don’t come out of the blue.
Yes, he famously put a rocket under Nasa to overhaul the Soviets in the space race. But it was at JFK’s urging, buried deep in his 1962 “Special Message to Congress on Urgent National Needs” – which mostly dealt with an increase in Cold War US military spending – that he promised US$75 million to “help give us at the earliest possible time a satellite system for worldwide weather observation”.
Kennedy may have fired the starter’s gun for the race to the moon, but he also sparked what we now take for granted, the ability to see the path of storms, rain and winds across the face of the planet, courtesy of a garland of weather satellites. It is one of the rare examples of the nations of the world working together on a public good.
The dream of a world weather map is not a 20th-century aspiration, but dates back, says Andrew Blum, to the essayist John Ruskin, who in 1839 wrote of a “vast machine” to run “perfect systems of methodical and simultaneous observations”. These would link the solitary prairie dweller as “part of one mighty mind, a ray of light entering into one vast Eye”. The closest the 19th century got to Ruskin’s dream was at the Smithsonian Institution in 1848, when the museum organised a network of observers from across the US to telegraph each day their weather observations. A giant map of the country in the museum lobby was refreshed each day with paper disks pinned to locations, white for sunshine and black for rain.
Today’s forecasts are built from hundreds of millions of observations that are in turn fed into simulations of the world’s weather. Supercomputers do their algorithmic magic and out come the forecasts that save lives and crops and provide material for untold millions of small-talk exchanges.
There are three main players in the arena of global weather simulations: the UK’s Met Office, the US National Weather Service and the European Centre for Medium-Range Weather Forecasts, commonly known as the Euro. When it comes to predicting the future of weather, the Euro is the champ. The simulations are not machines in the industrial sense, with input, processing and output. Instead, they seek to be a perfect representation of the actual global weather system, a virtual simulacrum with a clock the scientists set. Speed up time and it is possible to glimpse the future, hours and days ahead of time in the real world.
Building and improving the simulations takes massive creativity, effort and brainpower. The Euro’s seven-day forecast is now as accurate as its five-day forecast, but it has taken 25 years to make that 48-hour improvement.
Blum is an affable guide to the behemoths of weather prediction, as he meets scientists at the three main centres of forecasting. Best of all, he distils complexity into a largely engaging read that explains how, with a few taps of a smartphone, we can check the week’s weather, whether in Auckland or Addis Ababa, thanks to some largely unsung scientists.
THE WEATHER MACHINE: How We See into the Future, by Andrew Blum (Bodley Head $35)
This article was first published in the September 28, 2019 issue of the New Zealand Listener.