In 1854 an experimental department of the Board of Trade was set up under the illustrious command of Captain Robert Fitzroy, RN, who had sailed the globe on HMS Beagle with passenger Charles Darwin. Fitzroy was investigating the possibility of forecasting the weather, as hundreds of ships were lost to storms every year. He established the first weather charts and storm warning system, thus laying the foundations for the science of meteorology. In the 20th century, the Met Office enjoyed its own 'finest hour' when forecaster John Stagg gave General Eisenhower the weather go-ahead for the D-Day landings on 6 June 1944.
For most of its life, the Met Office was an offshoot of the Ministry of Defence, providing forecasts for the public and shipping, and also the forces. In 2003 it moved from Bracknell to a new £80m HQ in Exeter, and since 2011 it has been part of the Department for Business, Innovation and Skills. It operates commercially as a trading fund, selling data and services to both the private and public sectors. It produces 3,000 tailored weather forecasts daily.
Who's the boss?
Chief executive John Hirst has been the UK's weatherman-in-chief since 2007. He trained as an accountant and was previously boss of electronic components business Premier Farnell and a senior executive at ICI. In 2010 he was caught sans umbrella in his very own media storm, when it was revealed that he was earning over £195,000 a year - more than the prime minister.
The secret formula?
Weather forecasting these days calls for a lot more than the traditional pine cone and piece of seaweed. The science of modern forecasting was first mooted by mathematician Lewis Fry Richardson in 1922, who realised that complex equations could be used to model the behaviour of the atmosphere and thus predict the weather. In that pre-computer age, however, Richardson reckoned it would require 64,000 people to produce a forecast sufficiently rapidly for it to be useful. It wasn't until 1959 when the Met Office got its first real computer - a Ferranti Mercury - that his methods became practical. Today's short-range forecasts are produced on an IBM supercomputer capable of 100 trillion calculations per second and are perhaps four times more accurate than they were 30 years ago. But prediction remains an extremely uncertain business.
Where to start? The 'barbecue summer' of 2009 failed to materialise and this year's drought in the south-east was washed away by the unforeseen arrival of the wettest summer in a century.
*All figures for 2011-12