How do weather forecasters know how hot it will be? Weather forecasters predict how hot it will be by inputting as much meteorological data into a supercomputer as they can. They also look at patterns. They look at what the weather has done in the past. And they go on the basis that a weather system will continue.
Weather forecasts can only be as good as the available data and the predictions are more accurate the closer they are. It is far easier to predict what the weather will do in the next hour than it is in a month’s time. The reason for that is that there are just too many variables. It is said that if you could map out every single molecule in the entire Earth and its atmosphere, and then track them, you could predict the weather. However, that is obviously impossible.
The first step to making a weather forecast is getting as much data as possible. Meteorologists have access to weather stations, both on ground and weather buoys at sea. These send observations and data about the atmosphere. They often have access to weather satellites, which can send even more data about the atmosphere. They have data from commercial airlines and from ships at sea. Then all of this data is fed into a supercomputer. It has to be a supercomputer because there is too much data and there are too many variables. Our atmosphere is basically a fluid and movement in one part of the fluid can affect another part of the fluid. All of these connections have to be mapped out and only a supercomputer can do that. The American National Weather Service uses a supercomputer that has a processing power of 8.4 petaflops. One petaflop is 10,000,000,000,000,000 operations a second. And their computer has 8.4 of those. That is about 10,000 times more powerful than the computer I am typing this on now. And even a computer like that isn’t powerful enough to keep track of all the data it has. What it can do, though, is analyze all of that data and then analyze all of the weather from previous years that it has in its database. Then it can compare them and give a series of predictions.
Weather forecasters have to look at all of the possible predictions the supercomputer kicks out and try to work out which is the most likely. They can look at previous weather and they know that when the weather is doing X now, then, in the past, it did Y. That doesn’t mean it will do the same thing again. There are far too many variables for that, but it is likely. And that is why predictions get more accurate the closer we get to the event. As we get closer, the weather forecasters can get rid of predictions that don’t fit with what the weather is doing and they can refine their predictions with what is happening. Weather forecasters also now that if a weather system is doing one thing, it is likely to continue doing that until something changes it. Remember that the atmosphere is a fluid. If you mix cold water into a fluid at one point and stir it, you know that it will reduce the temperature at another place until all of the cold water has warmed up. That is predictable, but exactly where it will happen and how long it will take is very difficult to predict.
So, how do forecasters know the temperature. Again, this is a lot of guess work based on patterns. The temperature of a place is decided by a huge number of factors. The amount of sunlight is obviously one of them. The amount of humidity in the air is another. The cloud cover and the strength of the wind are two factors. The wind can either blow cold air in and lower the temperature, or blow hot air in and raise the temperature. The air pressure can have a big effect as well. When there is a high pressure system, it can trap warm air underneath it and this can lead to heatwaves as the area gradually warms up. So, how do the forecasters know how warm it will be? They look at all of these factors and they run all of the data through their supercomputer. Then they look at past data and they decide if that data applies or if other things will affect it. And then they make their prediction. They tend to get trends right, but they might be off on the numbers. Weather forecasting gets better year by year and AI might have a big impact on it. And this is what I learned today.
Sources
https://en.wikipedia.org/wiki/Weather_forecasting
https://www.weather.gov/about/forecast-process
https://www.weather.gov/about/supercomputers
Photo by Brett Sayles: https://www.pexels.com/photo/photo-of-white-and-gray-clouds-1101104/