Oh, how far we've come: Forecasting weather over the years

"At its heart, predicting the weather is one big, ugly math problem – downright frightening for those with a weak stomach for equations."

A mere 30 years ago, no one would dare speculate on how much snow would fall three days from now, suggest whether next week would be abnormally warm or cold, or think about giving specific forecasts for different neighbourhoods.

But technology is changing our ability to see accurately into the future and the way weather forecasts are made. In fact, the same revolution in technology that makes your phone more powerful than a 1980s supercomputer has transformed meteorology and how we forecast at The Weather Network.


At its heart, predicting the weather is one big, ugly math problem – downright frightening for those with a weak stomach for equations. But the idea of using math to forecast the weather goes back over 100 years.

An early pioneer of meteorology from the 1920s estimated that he would need a room with 64,000 people crunching numbers to forecast the weather this way. It wasn’t until the 1950s that there was a realistic way to solve the complex equations that describe the weather. This is when computers ushered in the area of Numerical Weather Prediction.

The advances were slow but steady for the first few decades. However, in the last 30 years the output from computer models has gone from being a useful tool for meteorologists to being the base of all weather forecasts. Our meteorologists still perform a critical role in interpreting the data, improving upon it, and communicating the forecast. But computers are now the most important link in the weather forecasting chain, bringing together promise and possibility in one neat package. In fact, some of the most powerful supercomputers in the world are used for modelling the atmosphere and making weather forecasts.

Computers and math alone cannot predict the weather. Observation data about what is happening right now with the weather is necessary to make forecasts. This data comes from high-tech and some low-tech ways of measuring what’s happening above our heads.

Content continues below


The single most important source of data is from weather satellites. Satellites are about the size of a school bus, but cost 100,000 times more to build and launch into space. These eyes in the sky that orbit the earth have the ability to see vast swaths of the earth’s atmosphere, sensing temperature and humidity. Satellite data, along with information from ground-based RADAR stations, weather stations, weather balloons, buoys, aircraft and ships are fed into supercomputers that then compute how heat and moisture will move in the atmosphere, giving us a weather forecast.

At The Weather Network, we use the computer model output from many agencies such as Environment and Climate Change Canada. Our meteorologists then select the best of this data and use their own experience to make the final forecast. We also use RADAR (Radio Detection And Ranging) data to pinpoint the timing of precipitation in the next few hours.

There is a lot under the hood that goes into driving the weather forecasts you see. The best way to fully understand the forecast is to watch our weather map videos and TV presentation to see all of these tools in action. You get to see what’s really behind the weather forecast and can often get an idea whether the forecast may change.


Some think that data from your smartphone could make weather forecasts better. We have experimented with integrating smartphone pressure and temperature data into weather computer models, but it’s unlikely this will result in a revolution in forecast accuracy. What’s more likely is that weather forecasts will continue to gradually improve as the Internet of Things, including your smartphone, provides more real-time data along with improvements in computer technology and the meteorological science that tells the computer what to do.

Despite the progress, perfection in this business will never be attainable. The ‘Butterfly Effect’ describes how tiny variations in our mathematical description of the current weather can cause wildly different weather forecasts. If there is one thing that most people don’t know about weather forecasts, it’s that they are not all created equal. Some parts of a weather forecast can be guarantees, while other parts are highly uncertain. This will always be the case.

And for all the amazing technology that we use, there is one piece of meteorological equipment that has endured for over a century. The yardstick or metre-stick is still the gold standard when it comes to measuring snow. Fancy ultrasonic snow depth sensors and billion dollar satellite measurements still cannot compete with a person sticking a ruler in the snow multiple times to get an average of how much fell. You may just have one of the most reliable pieces of weather observing technology in your garage.

Content continues below

Follow The Weather Network's Chief Meteorologist on Twitter @ChrisScottWx

Editor's Note: This article was originally published in March 2019.