A Human Super Computer to Forecast the Weather

Imagine filling a stadium with 65,000 people, each person tasked to solve complex mathematical equations, being signaled by a director at the center of the arena, who notifies when to pass their computations on to the person sitting next to them for further calculation. This is the vision proposed by Lewis F. Richardson early in the 20th century. Later he became known as the father of numerical weather prediction .

Richardson used a set mathematical equations he derived from the equations of state and motion of the atmosphere to forecast the weather. Richardson developed a set of equations that approximated a solution to the physical equations that govern the state and motion of the atmosphere. His technique solved nonlinear partial differential equations by numerical approximation. This paved the way for modern day computers to use this very same technique in creating our everyday weather forecasts.

A devout pacifist and follower of Quaker teachings, Richardson joined the Friends Ambulance Unit in France during World War I, ferrying wounded troops from the front lines hospitals. During his free time, he set up his model using his "primitive" set of equations that predicted pressure and wind changes over a set of points in Europe. He set about calculating his result armed with a slide rule and sheets of paper to keep track of all of the terms in his equations.

Unfortunately, after furious calculating over a two year period, he came up with a wild result, showing a highly unrealistic pressure change in just a six hour period. His equations were mathematically correct, but Richardson gave as much weight to smaller scale physical processes as the larger ones. Thus, they did not have the proper balance to handle mass and motion fields. The result was a lot of noise in his solution, and an erroneous forecast.

Undaunted, Richardson published his work in a book called, "Weather Prediction by Numerical Process" in 1922. The book received early criticism by the meteorological community not only because it contained an erroneous result, but also because of its wild premise that some day machines would be able to perform the task of these computations in a timely manner.

Richardson stated that forecasting using numerical prediction would be possible 'some day in the dim future.' Even with the thousands of people producing frantic calculations in his "forecast factory", Richardson surmised that they would just barely be able to keep up with the changing weather conditions.

It was not until 1950, just a few years before Richardson's death in 1953 did his dream really began to take shape as reality, when numerical weather prediction experiments began on the first computers. Atmospheric scientists latched onto Richardson's ideas in the years following the publication of his book and refined his equations to properly balance the many physical processes the the atmosphere.

Richardson, discouraged by what here perceived as takeover by the military of forecasting services in England left the field of meteorology in 1926 to follow other pursuits. From the mid '30s to the remainder of his life he devoted his time to applying mathematical models to international relations, conflict, and peace.

Sources:

Richardson, Lewis, Weather Prediction by Numerical Process. Cambridge University Press, Second Edition 2007