Windswept_ The Story of Wind and Weather - Marq de Villiers [62]
Across the Atlantic, the British Meteorological Office was founded in 1854 as a small department in the Board of Trade, and by 1861 was already issuing gale warnings to shipping by telegraphing predictions to harbormasters, who would then hoist appropriately colored cones up a mast. Their forecasts were ostensibly for forty-eight hours, though they acknowledged that their day-two predictions were, to put it kindly, erratic, because most British weather came in from the Atlantic, where few observer stations were located. These forecasts persisted for a decade and then abruptly stopped, over the protests of the sailors who had been using their output.
The same thing was happening elsewhere in Europe, and by the 1870s data from weather observing stations all across the globe led to the construction of the first crude multinational weather maps. Which in turn led to the development of synoptic forecasting—the compilation and analysis of weather data from many different regions in the same period. In September 1874, the official weather map showed a hurricane for the first time.
Around the turn of the twentieth century efforts were made to develop what was called numerical forecasting—that is, forecasting the weather by solving mathematical equations that described the physical laws involved. This wildly optimistic notion—the real complexity of weather data had not yet been recognized—was first expressed by Norwegian weatherman Vilhelm Bjerknes in 1904, the year before Einstein wrote his paper expressing the special theory of relativity.
A short time later British mathematician Lewis Fry Richardson tried to put Bjerknes's ideas into practice and, working furiously for six months, produced a six-hour forecast for Munich. The futility of producing a forecast six months after the event happened was not lost on Richardson; nor was the fact that his forecast was wrong in almost all aspects. Rather bravely, Richardson reported on his failure in his 1922 book, Weather Prediction by Numerical Process. He suggested, tongue firmly in cheek, that the difficulties could easily be overcome: To predict the weather before it actually happens, you would need a roomful of people, each computing separate sections of the equations, and a system, not yet invented, for transmitting the results as needed from one part of the room to another. He guessed no more than 64,000 mathematicians would be needed.
The next real technical advance came in the 1920s, with the addition of huge amounts of high-altitude data. This was made possible by the invention of the radiosonde, a small lightweight box tricked out with weather instruments and a radio transmitter. Radiosondes were sent aloft tethered to helium balloons, climbing to almost eighteen miles before bursting. On the way down, they transmitted wind velocities, temperature, moisture, relative humidity, and pressure information to a ground station. Even now, for regular weather data collecting, radiosonde probes are the workhorses. Literally hundreds of them go up every day. Twice a day, every day, the little weather offices in places like Abidjan and Dakar and Niamey in West Africa, and on the Cape Verde Islands, and for that matter in Honduras, Cuba, a scattering of Caribbean islands, and yet again all up the eastern seaboard as far as Newfoundland, in Greenland and Iceland and the British Isles, helium- or hydrogen-filled balloons carrying a small payload of instrumentation are released into the atmosphere. At noon Greenwich time every day all this data is transmitted to regional offices, where there are any, then to national ones, and then the data flies across the oceans. Within minutes the computers in the national weather centers in Halifax and Miami and Ottawa and London and Hong Kong all have the same data to crunch.
The synoptic maps ("synoptic" here means a general overview) you see on your television weather channel or published in newspapers are based on this data.
Over the years, certain transborder conventions have been developed. Everyone measures the same things