‘The internet has made us better at predicting weather’
In the days since the first of its five courses opened in Australia, the online prediction market has been transformed.
This is what’s going on.
We’ve taken the guesswork out of forecasting.
But in the case of the Perth course, the result was a disaster.
The course was a colossal flop.
The online model gave a forecast of a 95-degree day, when it would be 90 degrees and then be as cold as minus 20 degrees.
When the day finally came, it was minus 26 degrees and minus 23 degrees.
But the forecast showed it was likely to be above freezing, and the forecast was a total flop, because of the severe weather.
When a company makes a prediction about weather, it can be useful for the company, says Matthew Caulfield, co-founder of Predictable Analytics, a Melbourne-based firm that tracks internet-based predictions.
But it can also be a liability.
“The fact that we’re seeing more and more prediction errors is a reflection of the way the internet has changed the way prediction works,” he says.
And as more of this sort of thing happens online, people are learning more about how to spot these kinds of mistakes.
The problem is that a lot of these are happening in the prediction market as well.
“We have a lot more people than ever before making predictions online,” says Dr Richard Heap, chief executive of WeatherAware, a company that tracks weather data for businesses.
“In this age of big data, there are a lot less models, and that is a problem.”
The problem for forecasting firms is that the online predictions they make are often flawed.
They’re often based on the assumption that weather is a perfect prediction.
In fact, the forecasts of the world’s best weather forecastor, Dr Heap says, are actually a lot worse than the forecasts made by professional forecasters.
And, in some cases, the accuracy of the forecasts has dropped in the last five years.
But a company can use this to its advantage.
“If you have a forecasting company that has a good system that they can put together, then they’re more likely to make better forecasts,” Dr Heam says.
The good news is that some companies have been investing heavily in improving the accuracy and reliability of their online forecasts.
“There’s a new generation of forecasts,” says Mark Stokes, a meteorologist at the University of Melbourne.
“They’ve been using a much more sophisticated forecasting model that is more predictive than anything that has been produced for the past 40 years.”
It’s not just forecasting firms that have been improving.
Many other industries have started to take on this challenge.
As part of its global climate strategy, the European Commission in April announced a new set of climate standards to increase the accuracy, reliability and transparency of weather forecasts.
The new rules, called the climate standards, aim to ensure forecasts are based on accurate information and can be used by businesses and government to plan their operations and to monitor changes in the climate.
A lot of organisations are using these standards to improve their forecasts, Dr Stokes says.
But forecasting firms are still not the only ones using the climate standard to improve accuracy.
Many of the internet’s biggest brands, including Twitter, Pinterest, and Facebook, have introduced new weather forecasts, and many companies are now using climate standards in their forecasting and data analytics efforts.
And many of the people who have become experts in predicting the weather have also begun to incorporate climate data into their forecasting models.
“Climate data is one of the things that really makes us better,” says David Anderson, founder of Weather Aware.
“I think it’s one of those things where you start using it to make predictions, because you know you can use it in the future.”
For a company like Predictable, this means developing models that can predict how the world will respond to extreme weather events.
And that means more accurate predictions, which can then be used to make the forecasts companies make for other clients, like businesses, governments, and weather agencies.
“Weather is so big,” Dr Stoke says.
“It’s a big business, and it’s a huge business for many industries.
It’s one where you’re constantly adapting to new technology.”
So, while it’s easy to get carried away with predictions, what if you’re not using them well?
For weather forecasting firms, the best thing you can do is use the internet to learn from mistakes.
“Most of the time you can learn from a mistake, because people can see through it and make better predictions in the long run,” Dr Anderson says.
So what can you do?
“The first thing you should do is find out what the forecast is based on,” says Michael Kallmann, a lecturer at the Institute of Environmental Studies at the Australian National University.
And then you can