“Nothing is independent, the entire world is connected which every event influencing each other. The world is a giant chaotic non-linear system.”

I wrote this a few years ago the core of it is correct. However, due to advancements in machine learning chaotic systems may not be as unpredictable as they once were.

The weather possesses scant regard for forecasts and never acts as expected. I have become so despondent with meteorologists that I’ve replaced them with my cat. If she puts on her coat before going outside I ease off the mescaline… But why are weather forecasts awful, and will they get better? The answer to this lies in Chaos.

Chaos is a scientific term, stolen by the layman to describe his shopping trip, “Oh, it was just chaos!” this usage should be met with disapproving glares. Chaos and Chaos theory, are precise terms and describe systems where minute changes in initial conditions transform outcomes. A fancy way of saying: what you start with will change what you end up with. Dave down the pub could tell you that.

Dave will describe chaos: “When I start my night with a pint of beer, it’s a casual night, easy. But, and listen to me very carefully, if I drink a pint of gin I get well wrecked, it’s chaos.” No Dave, drinking gin instead of beer is not chaos. In our world the more alcohol you drink, the drunker you get. But what if that didn’t happen, let’s imagine it didn’t, in a magical place called Chaos Bar.

wavy lines

At Chaos bar, you can order a beer in any measurement, 500ml, 20ml, 501.05ml, etc. It’s very tedious work for the bar-staff, but nobody cares. In Chaos Bar, only your drink size changes. The bar is always the same, the people the same, everyone sits in the same places, and have the same conversations. It’s deathly boring, but not in any discernible way different from village pubs. Except in Chaos bar, imperceptible changes in the size of your beer radically alter your night.

If you drink 500ml of beer you’ll arrive home and drift off to bed, on a bed of roses (sans thorns) – wonderful. However, if you drink 500.01ml you will end the night tied up in a bin, with your dignity plastered on the inside of your pants.

That is chaos theory. A tiny change in starting conditions creates a completely different outcome. Described scientifically, the size of your beer is the ‘initial condition’ and the end of the night is the ‘outcome’. You might say, “that’s fine John, but I’ve drunk loads before, and have never found myself in a bin or shat myself” And granted, alcohol isn’t really chaotic, its effect on us is linear – the more we drink, the more attractive we become. But, Chaos bar gives us a good grounding to understand the non-linear nature of our world. And why weather forecasters despise butterflies.

The butterfly effect
Lorenz – a leading figure of chaos theory and enemy to butterflies – gave the lecture in 1972: “Does the flap of a butterfly’s wings in Brazil set off a tornado in Texas?” The lecture dragged Chaos into mainstream science. Previously, chaotic systems were oversimplified, with the chaotic part ignored. This led many to believe our world was inherently predictable. With more data and better models, we would be able to predict everything. Chaos theory changed that, and with it extinguished the dream of long term weather-forecast. A boon for the British.

British People love talking about ‘the weather’. If one visits England you are likely to overhear, “This weather we’re having, right?”
“Yes, the weather!”
The Brits don’t talk about the weather because they are deathly boring, although we are. But, because it is so damn unpredictable. Even the weather forecasters have no idea what it will do next. Below is an actual transcript of a forecast that I’ve made up.

‘Today it will be generally sunny, with the possibility of a few light showers, which might turn into thunderstorms as the day goes on. Be on the lookout for snow and ice, as blizzard Bertie blows around the British Isles.’

A few sentences cover every possible weather. It’s like filling in all the boxes on a multiple-choice test and then smugly announcing you got 100%.

The weather forecasts aren’t inaccurate because all meteorologists are all stupid. Although they maybe – I’ve never actually met one. No, weather is difficult to forecast because it is a chaotic system, and dependent on initial conditions, like you in Chaos Bar.
In weather forecasting, the initial conditions are the position of every particle in the atmosphere. And obviously, you can’t know that. Instead, forecasters use satellite images, and weather stations, to look at what the weather is doing. They feed that data into a supercomputer which runs thousands of simulations and out pops the weather forecast. It’s a little more complicated than that. But what you need to know is: forecasters run thousands of simulations and the weather forecast is the most probable simulation. They have to run thousands of simulations because their starting data is inaccurate. So each simulation is run with slightly modified initial conditions,

This is where our butterfly reenters the scene, a butterfly has scant regard for our weather forecasting, they are sinister beasts who ruin our initial conditions. When the butterfly flaps its wings, the molecules of air it pushes will push on other molecules, these push on other molecules, until eventually, the paths of many molecules have changed due to the dastardly butterfly.

Because the world’s weather is the consequence of interactions between every atom in the atmosphere, slight variations anywhere will eventually alter the weather globally. In fact, anything that moves in the atmosphere influences the weather. Why the butterfly got the blame nobody knows. The butterfly effect could be called the ‘dropped sausage effect,’ with no loss of meaning.

To provide an idea of how difficult it is to predict the weather. If poles, which reached from the ground to the top of the atmosphere, were placed on every square meter of the planet, and these took measurements of pressure, temperature, humidity, wind speed, from the base to the top. These measurements still wouldn’t enable an accurate long term forecast. You see, it is not the poles that are important, but what’s in space between them that matters.