Weather forecasts: The tech giants use AI but is it any good?
A wave of machine-learning weather models have been unleashed by some of the very biggest businesses on the planet.These challenge the orthodoxy of traditional physics-based computer forecasts that have been incrementally developed and improved over many decades. But are the machine learning models any good? The weather is a national obsession for us Brits, and it is no wonder given the huge changes that are seen and felt from one day to the next.
Accurate weather forecasts are not just vital for planning our daily lives but knowing about upcoming severe weather can help us to change our behaviour, save lives and mitigate damage to property.It is impossible to assess the full economic value of weather forecasts globally, but the numbers are huge. According to NOAA (National Oceanic and Atmospheric Administration), in the US alone - and just taking into account the biggest weather disasters that caused over $1bn (£740m) in damage - the fallout from severe weather in 2024 amounted to $182bn, with 568 fatalities.Since 1980, this damage figure stands at nearly $3tn! Meanwhile, in the UK, there were 1,311 excess deaths caused by heatwaves in 2024.A study from consultants, external, London Economics, concluded that the Met Office would bring £56bn of benefits to the UK economy over the course of a decade through providing meteorological services. Multiply these kinds of numbers across the whole world, with a growing population exposed to increasingly extreme weather fuelled by climate change, and weather is big business.
The biggest computers on the planet
Traditional weather forecasts are produced on some of the biggest supercomputers on the planet; the Met Office super computing contract is worth £1.2bn. That huge sum of money buys you a machine that can perform 60 quadrillion (60,000,000,000,000,000) calculations per second, running a model containing the understood physics, with over a million lines of code and using 215 billion weather observations. Global weather models work by crunching the numbers in a grid of boxes right around the planet. The size or resolution of these boxes varies across different meteorological models, but range between about 10sq km to 28sq km (3.86sq miles to 10.81sq miles).At this kind of resolution they cannot accurately predict showers, while mountain ranges are lower and smoother than in the real world. The highest resolution model from the Met Office, the UKV (the model that runs BBC TV graphics for the first 48 hours) can predict showers with its incredible 1.5km (0.9 mile) resolution - but it takes so much computing time and power that this model is not able to forecast for the entire world; instead it concentrates on the UK and Europe.
Machine-learning weather models: are they any good?
Machine-learning weather models have only been around for a few years, showing promise as they develop rapidly. Traditional models take hours to run on hugely expensive supercomputers, however this new breed of models can take less than a minute to run on a standard laptop. They don't need to know all the 'burdensome' laws of physics, but are instead trained on 40 years of past data to make their predictions.How do they perform? Well let's look at forecast verification data from ECMWF (European Centre for Medium Range weather Forecasting) for atmospheric pressure patterns in winter 2024/2025, external.GraphCast (Google), AIFS (ECMWF) and Aurora (Microsoft) were more accurate than the traditional IFS (ECMWF) benchmark forecast, whereas FourCastNet (Nvidia) and Pangu-Weather (Huawei) trailed behind.So some of the machine-learning models performed better, and some worse, but it depends on which variable you look at, and all of this could change quickly as the rate of progress accelerates. Just like traditional models, AI models are less accurate the further ahead in the future they're trying to predict - a consequence of the chaotic nature of the atmosphere. Looking 10 days ahead, none of the AI models (or traditional models) were able to offer forecasts considered to be of much use in terms of accuracy.
So is it time to walk away from physics-based weather models? Not yet! Machine-learning weather models are not only trained using data produced by traditional weather models, but they also use the start position of the atmosphere from traditional models as their input point. In other words, without those traditional models running, the machine-learning models wouldn't work as well.All machine-learning models can forecast large scale features like high and low pressure six days ahead very well, but they can underperform compared to traditional models at smaller-scales of 1000km or less.This means that important features like troughs and ridges could be missed, which would make the difference between a dry day, or a day with heavy rain.
The majority of machine learning models have a resolution of 28sq km, which is the same scale as the data that they've trained on. This means small features like showers would most likely be missed, so they wouldn't be able to forecast a Boscastle flood event many days ahead.News headlines have claimed that these new models are better than traditional models at predicting hurricanes. It may be true that some have been a little better at predicting the landfall of hurricanes ahead of traditional models, but at the same time they have been very poor at predicting the wind strength and therefore the likely damage the storm would bring. This may be the result of the smoothing, or averaging effect of looking at lots of hurricanes in the 40 years of training data.
AI models may struggle to forecast effects from rare events that have not been seen often in the 40 years of training data. An example of this is the 1991 eruption of Mount Pinatubo that cooled the planet down by up to 0.5C for two years.There are also questions about how well AI models will forecast in a warmer world as our planet continues to heat up as a result of climate change. The past climate that they've trained on will look quite different to our future climate, as greenhouse gases continue to accumulate in the atmosphere.So, where will we be in five years time?"I think we'll have traditional models running alongside AI models so that we are drawing on their combined strengths to enable hyper-localised accurate forecasts, delivered fast, when you need them," says Professor Kirstine Dale, chief AI officer at the Met Office. Machine-learning models haven't been around for very long, but with their speed, computer efficiency and rapid rate of development, they show great potential.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Telegraph
30 minutes ago
- Telegraph
UK chipmaker considers breakup as it is rocked by Trump's tariff assault
One of Britain's leading microchip businesses is racing to complete a $1bn (£740m) sale as it grapples with the impact of Donald Trump's trade war. Imagination Technologies, which designs graphics processor technology for customers such as Apple, has held sale talks with two US rivals. Bosses of the Chinese-owned company are also considering a potential break-up to get a deal over the line, which could include selling its valuable patent portfolio. It is understood that executives have put a $1bn price tag on Imagination, with bankers from Lazard overseeing the process. In a memo to staff, Ray Bingham, Imagination's chairman, said 'the Galway Project', which refers to the sale talks, 'continues to progress well'. The memo added: 'The second-to-none creativity and expertise of our teams is an incredible asset to capture the unique opportunity created by the current AI revolution.' It comes after the president's trade war with China has rocked the tech industry, including at UK-based Imagination. Pressure has arisen after the US introduced increasingly stringent sanctions on China's chip industry, as Mr Trump seeks to limit China's access to Western technology. This stems from fears that it could be used by Beijing's military. Imagination, which is based in Hertfordshire, was taken private by US-based investor Canyon Bridge in a £550m deal in 2017. However, the business was later engulfed in a political storm after Canyon Bridge's Chinese backers attempted an unsuccessful boardroom coup. This led to the departure of the then-chief executive, Ron Black, who blew the whistle on the Chinese effort to take control of the business. Last year, an employment tribunal ruled Imagination had unfairly sacked Mr Black, who had sued the company for £200m. Damages are yet to be determined. Canyon Bridge has been considering its options for Imagination for several years, at one stage exploring a Chinese float and then a US listing. According to its latest accounts for the year ending December 2023, Imagination's revenues increased by 3.6pc to £124.6m, although its profits plunged by 78pc to £2.6m. At the time, the company warned that 'US and UK export controls' on its Chinese customers had 'stymied top-line revenue growth'. As for its management, staff were told earlier this month that Simon Beresford-Wylie, the current chief executive, is preparing to leave the business. Didier Lamouche, a board member, will lead the business in the interim while it appoints a successor.
BBC News
38 minutes ago
- BBC News
UK start-up to launch metal-fuelled thruster into space
A UK-built spacecraft engine that runs on metal is preparing for its first test in space, in a mission that could pave the way for longer-lasting, more flexible plasma thruster, developed by Oxfordshire start-up Magdrive, heats tiny bits of solid metal into a hot gas to produce powerful bursts of thrust. In the future it could allow satellites to recycle parts of themselves, or space junk for fuel, but for now the metal is carried launch, which is due from Kennedy Space Center in Florida, US, at 22:00 BST, is backed by the UK and European space agencies. Mark Stokes, chief executive and co-founder of Magdrive, said the mission showed what British innovation could achieve with the right support. "We've spent four years building something that breaks the mould," he said."This launch isn't just about proving our tech – it's about proving the UK can lead in space." Dr Gianluigi Baldesi, from the European Space Agency, said the quick progress from project start to launch demonstrated the "bold" innovation it aimed to encourage. "In less than a year, we have gone from kick-off to launch," he from the test flight is expected in August and September. You can follow BBC Oxfordshire on Facebook, X (Twitter), or Instagram.
Auto Car
42 minutes ago
- Auto Car
Tyre makers developing EV-specific rubber to boost ranges, minimise wear
A year since it began marketing EV-specific tyres for electric vans and taxis, British company Enso has launched its first tyres for passenger EVs – specifically, the Tesla Model 3 and Y and Jaguar I-Pace. There are currently more than 1.5 million electric cars on UK roads, and this year car makers expect to sell a further 450,000. Meanwhile, in 2022, 40 million car tyres were sold in the UK in a market worth around £2.3 billion. By 2030, the market is expected to be worth around £3.9bn. Compared with giants such as Michelin and Continental, Enso is a relative minnow, but it punches above its weight thanks to initiatives such as direct sales and an emphasis on technical innovation - including reducing particulate pollution through improved tyre wear rates. 'The sheer weight of EVs impacts the tyres,' said Gunnlaugur Erlendsson, founder and CEO of Enso. 'The fact that you then have high torque and the cars being driven in urban areas also has an impact. The tyres wear down, but we reduce the rate at which they do by using better raw materials, construction and 'recipes', and by collecting data continuously from the field to understand how EVs are reacting to the road.' Balancing tyre wear with optimal rolling efficiency, the desired wet and dry handling performance and acceptable noise levels are among the challenges confronting Enso and other manufacturers of EV tyres. 'Today, tyres are coping with far heavier vehicle loads than in the past,' said Jaap Leendertse, general manager of tyre development at Falken Tyres. 'Previously, you had 50% of a tyre's load capacity utilised; now we are at 70-80%, and that means we are more focused on the tyre lasting longer.' Michelin has stated that EVs wear tyres 20% faster than ICE vehicles. According to Kwik Fit, Tesla models are the heaviest on tyres, while Citroën, Mercedes and BMW models are among those needing fresh tyres at an above-average rate.
