# Supercomputers

Without computers weather forecasting as we know it today would not be possible. From simple desk calculators to complex supercomputers which perform trillions of calculations a second, information technology has been at the forefront of understanding the weather for decades.

## Mathematics and weather

The idea of creating a weather forecast using dynamic equations was first put forward by English mathematician, Lewis Fry Richardson, in 1922. He realised the dynamics of the atmosphere could be modelled by doing thousands of equations, thus being able to predict the weather.

In a pre-computer age, however, the only way to apply his numerical method was by hand. He estimated it would take 64,000 people to perform the calculations needed to make a forecast in time for it to be useful. While this wasn’t practical, Richardson’s theory formed the basis for weather forecasting as technology improved.

## First steps for technology

In the 1950s the Met Office acquired an electrical desk calculator — which was cutting edge technology at the time. A mathematician specially trained in computational methods was hired to work it and thus began the first efforts to fulfil Richardson’s dream.

This was taken further when Met Office staff had access to a ‘real’ computer owned by catering company, J Lyons. Called Leo, the machine was bought to calculate the value of Lyons’ bakery sales, but its power was harnessed to help speed up calculations needed to make a weather forecast.

## First computer

The Met Office embraced the computer age in 1959, when a Ferranti Mercury, nicknamed Meteor, was purchased. Capable of doing 30,000 calculations a second, it was a major step forward in the evolution of making weather forecasts. For the first time, scientists were able to regularly use numerical methods to make their forecasts.

By building an understanding of the way the atmosphere works, equations are created which seek to mirror these processes. The equations, built out of lines of computer code, combine together to make ‘models’, which are effectively attempts to recreate the dynamics of the atmosphere through maths. They work by taking all the current weather observations available to understand the current situation and applying the model to see what might happen next.

As our understanding of the atmosphere improved and the number of weather observation inputs increased, the need for more computer power also grew. Thus the Met Office bought a new supercomputer in 1965, an English Electric KDF9, which could do 50,000 calculations a second. This leap in speed, of more than 60%, allowed for faster, more complex forecasts to be made.

This pattern of advancing technology and increasingly complex models continued, with the Met Office buying successively quicker computers every five to ten years. By 1982, our CDC Cyber 205 could do 200 million calculations a second, but by 1997 a Cray T3E was doing more than a trillion (1,000,000,000,000) a second.

## Today

We are now using an IBM supercomputer which can do more than 100 trillion calculations a second. Its power allows it to take in hundreds of thousands of weather observations from all over the world which it then takes as a starting point for running an atmospheric model containing more than a million lines of code.

Necessarily, the supercomputer requires a large amount of energy to run and maintain — about 1.2 MW of electricity a year. While everything possible is being done to minimise this, the power consumption remains small in comparison to the socio-economic benefits delivered, including CO2 emissions reductions.

Take our global aviation forecasts, for example, which allow airlines to save fuel by using the wind to help them to their destination. We have estimated that this alone helps save around 20 million tonnes of CO2 a year through increased efficiency.

Crucially, Met Office forecasts save lives. Through our Severe Weather Warnings, mountain area forecasts and marine forecasts, with a host of other services, an independent report estimated we help save as many as 74 lives a year. Financially the same report concluded we save the UK economy more than £260 million a year.

How valuable is the Met Office?

Such is the power of the modern generation of computers, they can run ensemble forecasts, meaning they can run a model several times, each one from a slightly different starting point. This helps forecasters understand possible variations and make informed decisions on the most likely outcome.

## Results

Because of improved science and increased computing power, today’s four-day forecasts are as accurate as one-day forecasts were 20 years ago. Forecast accuracy should continue to increase as technology advances.

Beyond the present day, the power of the computers has also been harnessed to help understand the long-term future of the Earth’s climate, effectively making forecasts that stretch for decades into the future. This has allowed scientists to understand more about the dynamics of climate change and its impacts.

Our research into climate change is helping governments, businesses and individuals plan ahead for the challenges we face in the future. The predictions we make, and the detailed advice we give, has the potential to protect life on a huge scale worldwide, and potentially make even greater CO2 savings.