If you want to know what the weather is doing, go outside and raise your hand.

If it gets wet, it’s raining. If it’s dry, it’s not raining. If it casts a shadow then it’s sunny and so on.

It’s 100% foolproof.

It’s an age-old joke but the forecast is not as simple if you try to predict the future.

‘We have no idea what’s going to happen [in the weather] beyond three days out,’ a TV meteorologist reportedly told a study in 2008.

When TV weather presenters predicted 100% chance of rain in that same study, it only rained 66.7% of the time.

The BBC even had to apologise for predicting dry, hot weather as a thunderstorm rolled into south-east England.

But it’s not meant to be this way.

The Met Office has supercomputers capable of more than 14,000 trillion calculations-per-second, there’s a so-called ‘arms race’ between the US, Europe and private companies battling for the most accurate weather forecast.

‘Long-range (3-10 days) are as accurate now as one to two-day forecasts were 20 years ago,’ meteorologist Paul Knightley, group lead for UK and Canada at MeteoGroup, tells Metro.co.uk.

‘We’re getting much better at understanding the physics of the atmosphere and clouds.’

New technology is hoped to increase the amount of data used, with sensors attached to planes and even drones to collect more information.

‘Sensors are attached to an aircraft and they take readings every time the plane takes off and lands,’ Tom Schmutz, CEO of aircraft data company Flyht, tells Metro.co.uk.

‘The technology we own can augment and potentially in the future replace weather balloons.

‘We have a fairly advanced prototype for UAVs (drones) for weather data. As drones become more and more prevalent, they’ll have a bigger role.’

But it’s not perfect yet.

A seven-day forecast is accurate around 80% of the time, the US government’s scientific agency NOAA (similar to the Met Office in the UK) has said and a five-day forecast is correct approximately 90% of the time.

That accuracy falls to 50% for a 10-day or longer forecast.

So why isn’t it 100% accurate?

‘It all boils down to the butterfly effect, small changes in the atmosphere,’ Knightley says.

READ  Inside Musk's Tunnel to Fix 'Soul-Destroying' Traffic

‘For example, a small gust of wind might pass a mountain unusually quickly in a way you can’t simulate and this could, over the course of a few weeks, change weather patterns.

‘Computer power isn’t enough and you still have to make assumptions. We don’t know what every single part of the atmosphere is doing now and we never will.

‘Small-scale tiny errors become much bigger over time.’

It’s an example of chaos theory, where small changes have huge impacts over time.

Paul Knightley compares the weather to a horse race, where there are often clear favourites but a lot can happen over a race.

Even if a horse is odds-on to win, it is still not a certainty and that’s how bookmakers make their money.

This ‘chaos’ is combined with audiences not fully understanding what the forecast means.

If your weather app says there is a 50% chance of rain, it doesn’t actually always mean there’s a 50% chance of rain.

And yes, it is confusing.

The 50% accounts for the chance of rain somewhere in the forecast area.

So if forecasters give Birmingham a 50% chance of rain to affect half the city, there would be a 25% chance that you would get rained on (chance of rain multiplied by the percentage of area affected).

Even more confusingly, it depends what app you’re using.

A US-based app is more likely to use the complicated method above. A British app is more likely to mean 50% when it says 50%.

A lot of this comes down to the scale of measurement.

On a global map, the UK is split up into 10km by 10km squares.

Each of these squares is displayed as an average of the measurements taken. This is down from 13km in 2016 and from 338km in the early 1900s.

The idea is that the smaller the scale, the more accurate the picture will be.

‘We’re working on tighter grids,’ Knightley says.

‘It will be 5km for example and for a UK map, we’ll get the resolution higher and the grid down to hundreds of metres.’

A strange quirk of this technological race is that it could all be ruined by the 5G mobile signal rolling out at the moment.

READ  IT top 5 hire 57% more in Q1

Water vapour, obviously an important measure for the weather, is detected by satellites emitting a faint signal into the atmosphere at around a 23.8 GHz frequency, extremely close to the 24Ghz signal sold for 5G.

Because the frequencies are so close, it risks satellites picking up regular interference.

The US agency NOAA’s acting chief executive Neil Jacobs said that the interference could end up reducing forecasts’ accuracy by 30%, equivalent to the accuracy achieved in around 1980.

It’s seen as a big enough problem for US senators to have written a letter trying to limit the use of 5G until this is resolved.

In the UK, official bodies are less concerned.

‘The Met Office is in no way opposed to the introduction of 5G,’ a Met Office spokesperson told Metro.co.uk.

‘[It might be possible for] meteorologically relevant information to be extracted from 5G technologies.

‘However, if appropriate transmission limits are not applied there will be a significant and detrimental impact on satellite sensing of atmospheric water vapour, also impacting our weather forecasting ability.’

Even if the 5G issue is resolved, the big question still remains: will the weather forecast be 100% correct?



The best of The Future Of Everything

‘The simple fact is that everything from oceans to mountains has an impact on the weather and it’s impossible to predict with 100% accuracy,’ Knightley says.

‘An ocean might slightly change temperature over a week, influenced by the very weather you’re trying to predict, which then again influences the weather and the cycle goes on.

‘On a microscopic level, we don’t fully understand how clouds form. Of course, we know so much about them but on a microphysical level there is still a lot we don’t know.

‘Computer power has come a very long way but we are never going to get a perfect weather forecast because physics doesn’t allow it. You can’t know exactly where an atom is and its speed.’

Then it’s settled: it’s physically impossible.

Or is it?

Scientists at the University Of Maryland, US have modelled a potential solution to the chaos.

Focusing on the chaotic system of the Kuramoto-Sivashinsky equation (no? Us neither), they created an algorithm to learn how the system had evolved in the past to be able to predict the future.

READ  Facebook has had numerous scandals. So why does user engagement keep growing?

The researchers argued that once you have enough data, things seemingly entirely random can be predicted accurately.

Accurate predictions were made eight times longer than had previously been possible.

‘This is really very good,’ Holger Kantz, a chaos theorist at the Max Planck Institute for the Physics of Complex Systems in Dresden, Germany, has said.

‘The machine-learning technique is almost as good as knowing the truth, so to say.

‘This paper suggests that one day we might be able perhaps to predict weather by machine-learning algorithms and not by sophisticated models of the atmosphere.’

Given that bad weather is said to cost the economy $1trillion-a-year and extreme weather is predicted to kill 150,000 people a year in Europe alone, getting the weather forecast right is about so much more than just whether you need to take a coat.

Hurricane Dorian in the Bahamas, for example, kept defying predictions and caused huge damage and has cost 51 lives.

Meteorologists understand the importance of their work and Paul Knightley is very proud of his company’s record of ‘getting things right’ with improved technology and seasonal forecasting.

But until machine-learning algorithms find the answer, the best way to understand your personal weather forecast is still to go outside and look up.

You’ll be right every single day.



The Future Of Everything

Future Of Everything

This piece is part of Metro.co.uk’s series The Future Of Everything.

From OBEs to CEOs, professors to futurologists, economists to social theorists, politicians to multi-award winning academics, we think we’ve got the future covered, away from the doom-mongering or easy Minority Report references.

Every week – new pieces every Wednesday morning – we’re explaining what’s likely (or not likely) to happen.

Talk to us using the hashtag #futureofeverything  If you think you can predict the future better than we can or you think there’s something we should cover we might have missed, get in touch: hey@metro.co.uk or Alex.Hudson@metro.co.uk

Read every Future Of Everything story so far





READ SOURCE

WHAT YOUR THOUGHTS

Please enter your comment!
Please enter your name here