How far ahead can you predict the weather?
d ~ 210'969'600 km: day 82
Weather forecasting has made amazing progress in the past century. The days of relying on fir cones, star-gazing, folk sayings or supersized rodents are long gone. But how much better can it get? Will we ever know for sure that it'll be sunny three weeks from Tuesday?
To answer this, we first need to know a little about how weather forecasting works. What we call "weather" is just a description of what our atmosphere is up to - how hot the air is, how dense it is, what direction it's moving in and whether or not it's carrying much water with it. What weather forecasters do comes in two stages. First of all, they collect as much weather data (temperature, pressure, wind speed and so on) as they can from as many places as possible, to build up a picture of the weather right at this moment. This stage is generally automated, since there are lots of weather stations and satellites out there continually making measurements. Then, the forecasters use the laws of physics and a very large computer to work out how all that weather will have changed a minute from now, and then a minute after that. And so they step forward in time, using the answer from each step to work out the answer for the next step. So far, so good. To have perfect weather forecasts, all we need are amazingly powerful computers, and perfect information about what's happening now, right?
Sadly, no. Weather is an example of a chaotic system, and "chaos" here has a precise scientific meaning. What it means for weather is that however perfect our calculations are, there is no single answer about what will happen a month from now. This is because of the "butterfly effect" - tiny differences in the starting information can grow in importance as time goes on. Although this was only scientifically described in 1963, the concept isn't new. Do you remember this proverb?
For want of a nail the shoe was lost;
For want of a shoe the horse was lost;
For want of a horse the battle was lost;
For the failure of battle the kingdom was lost--
All for the want of a horse-shoe nail
This verse has been around for at least 500 years, and it's basically a description of a chaotic system. A single nail turns out to have a massive influence, but there was no way of knowing that beforehand.
We have all heard that a butterfly flapping its wings in Brazil could cause a tornado in Texas. But there isn't just one mischief-making butterfly out there. There are billions, and they've all got four wings each. And then you've got the ladybirds, the sparrows on the roof, my cousin's chickens and all the paper aeroplanes in the world. And you - don't touch that coffee! You might cause a tornado in Torquay.
The serious point here is that there is no way of measuring and including all of the tiny effects that everyday life has, or of knowing which ones will turn out to be the important ones. 10-14 days is about the length of time that it takes for some of those tiny effects to turn into much larger ones, so there is no way to predict the exact weather further ahead than that. We can still get the probabilities right - a 50% chance of a sunny day, for example. This is why medium-range and long-range forecasting is worth doing. But for exact prediction, 10-14 days is the limit and we're stuck with it.
That's fine by me though. I think that exact weather predictions months ahead would make life terribly dull. Hooray for chaos!