Many PhD researchers use climate models with little thought of how they were developed. Often there just isn’t time in modern schedules to investigate their history. How these models evolved through the ages is pretty interesting and involves battling mathematicians carving out early theory and determined meteorologists coding the way forward. The first real attempt at weather prediction was courageously undertaken on a hand-held calculator! It is a shame that the early development of numerical modelling is often swept under the carpet. What people don’t realize is that the story is filled with arguments, intrigue and great intellect.
The story really begins with the development of calculus. This branch of mathematics deals with physical rates of change and lays the foundations for the equations used in climate models. Before the 1600’s the mathematical community lacked cohesion, often notation was a personal preference and no general standard existed. Fortunately two giants appeared on the scene who managed to rigorously define and develop calculus – Isaac Newton (1642-1727) and Gottfried Leibniz (1646-1716). These magnificent men both managed to pull the ideas of derivatives, integrals and infinites out from the foggy cloud that was the current state of maths. Unfortunately a unified community was not instantly defined. Both Newton and Leibniz claimed the credit for creating calculus. Miraculously they both presented ground-breaking theories at the same time and, for some, this seemed too much of a coincident. The Continental Europeans sided with Leibniz and the British with Newton, both accused the other of plagiarism and refused to collaborate with one another. The advancement, as quickly as it came, crawled to an embarrassing halt whilst both groups squabbled unashamedly. This rift in the community persisted for over a century until efforts from both sides brought the petty arguments to a halt. After this the theories of both men were accepted and field truly prospered. Calculus as we know it finally broke free. As I see it (maybe from my jaded British Newtonian perspective!) modern calculus adheres more closely to the reasoning behind Newtons work (based on rates of change) but draws heavily upon Leibniz’s insights (based on infinitesimally small points) and his notation is often seen as the standard.
Parallel to these discoveries Charles Babbage (1791-1871) set to work on the creation of a machine which would aid mathematical computation. His vision was that the machine should be programmable, compute mathematical quantities and store these for later use. With the help of Ada Lovelace the first notions of a computer was born and he sketched out ideas for the difference engine and the analytical engine. His life work evolved into building these theoretical machines. But the journey was troublesome and wrought with pay disputes, funding and personal problems. Tragically Babbage never completed any of his machines before his death but he is attributed as the inventor of the first real computer. Despite this you can see a completed Analytical Engine on display at the London Science Museum.
As computers have evolved since the time of Babbage it has become evident that they are useful tools for solving complicated non-linear equations. In 1904 Vilhelm Bjerknes (1862- 1951) realized it would be possible to use Newton’s laws of motion to understand, and possibly forecast, atmospheric and oceanic events. He outlined the importance of having accurate initial conditions and sufficient knowledge of the laws governing dynamical transitions.
During World War One the British mathematician Lewis Richardson (1882-1953) attempted to create a numerical scheme to predict weather using nothing more than a hand-held calculator. His general interest in the subject led him to develop his scheme. Impressively he managed to complete his work alongside driving a Quaker ambulance in Northern France. He set to work in predicting one day of weather (20th May 1910) using his scheme. However, due to the calculator and his misunderstanding of error terms his results were a tad inaccurate. In 1922 he wrote ‘Weather Predictions by Numerical Processes’ which described this process. Richardson’s inspirational story led to many future prediction attempts. However, it wasn’t until the development of more sophisticated computers that more accurate results could be achieved.
A few years later, in the 1940’s, John von Neumann (1903-1957) supervised the construction of a computer powerful enough to create weather forecasts. A team, led by Jule Charney (1917-1981), was formed to use this resource and forecast weather patterns. This work was facilitated by the simplification of the governing equations; using only terms that were dynamically important. The removal of unnecessarily complicated terms reduced computational time; an important discovery attributed to Carl Gustav Rossby (1898-1957). In the 1950’s Norman A. Phillips incorporated some simple forcing terms to Rossby’s equations and his numerical solutions displayed many of the observed meteorological features. He managed to compute a month long prediction using the system. Phillips was awarded the Napier Shaw Memorial Prize in 1956 for his ground-breaking work and received the £100 prize fund – which is equivalent to £2,118.47 in modern currency!
In this day and age we invest considerable amounts of money into weather prediction. Improved computing power and scientific understanding leads to ever more efficient and accurate numerical models. The incorporation of various forcing terms, chemical processes and numerical techniques – which are often impressive and beautiful like the recently developed adaptive meshes – all remain active areas of research. Modelling of the ocean system developed during the 1960’s and comes with more complications still; such as including land masses, sea ice and mixing. It is evident that numerical models are powerful tools which are constantly improving. Every day a new page is added to the climate modelling history book with new methods being invented and refined. These models enable us to understand the climate and weather systems that are so important to our daily lives.
Pingback: ClimateSnack Norwich blossoms | ClimateSnack()