A sample text widget

Etiam pulvinar consectetur dolor sed malesuada. Ut convallis euismod dolor nec pretium. Nunc ut tristique massa.

Nam sodales mi vitae dolor ullamcorper et vulputate enim accumsan. Morbi orci magna, tincidunt vitae molestie nec, molestie at mi. Nulla nulla lorem, suscipit in posuere in, interdum non magna.

What’s in the future for weather forecasting?

My name is Harry Enten, and I’m addicted to weather forecasting.

In my teens, I went to Penn State Weather Camp, where I visited the National Weather Service and AccuWeather, and spent a week diving into the art and science of weather forecasting. Around that same time, I got featured in the New York Daily News about snow measurements.
I didn’t end up going into meteorology because I hate calculus. My addiction to weather forecasting and particularly snow remains, however. (There’s a reason I’m not moving to Washington, D.C., as it averages less than 14 inches of snow per season.)
However, I’m not a person who pays a lot of attention day to day to climate change.
And as I learned in the latest episode of my podcast, “Margins of Error,” having this disconnect between weather and climate change is surprisingly common.
A 2018 Pew Research Center survey asked people what was the most important topic covered on their local news broadcast. Weather, far and away, was No. 1 at 70%.
By comparison, a 2019 Washington Post poll found 10% say they often talk with their friends about global warming.
This split does make some sense. But as we discuss on the podcast, we can’t really separate them as much as we used to do.
Weather forecasts are an immediate concern to pretty much everyone. If a forecast says it’s going to rain later today, you’re probably going to take an umbrella with you. Climate change refers to long-term changes.
An aerial view of devastation after tornadoes struck the Bowling Green, Kentucky, area in December.
Still, for something that interests so many, there are a lot of misconceptions about weather forecasts.
I cover a lot of what goes into the forecasts in the podcast (so tune in), but one big misconception is how accurate they are. While people love to rag on meteorologists who get it wrong, it turns out, forecasts are better than they’ve ever been.

They’ve gotten a lot more accurate.
As New York Metro Weather’s John Homenuk told me, there are “amazing minds working on producing these weather models. They’ve gotten so detailed. I keep mentioning (that) we can predict individual thunderstorms, things like that. (An) incredible technology boom has helped us a ton.”
One quick way to know forecasts have gotten better is to examine hurricanes and see how the error rate in terms of nautical miles for these storms has declined.
The forecast error rate has dropped by anywhere from about 70% (for a 24-hour forecast) to about 90% (for a 72-hour forecast) since 1970. To put that in perspective, the average error for a 72-hour forecast was about 450 miles off in 1970. Today, it’s about 50 miles off.
That improved accuracy has saved countless lives.
Put another way, your local meteorologist has gotten really good — thanks to weather models and a better understanding of weather patterns.
That said, we’ll probably never achieve perfect accuracy.
A big problem is the butterfly effect. As Homenuk pointed out to me, “The big storms are still very complex and very difficult to figure out. … The joke sometimes in the thunderstorm community is a farmer can sneeze in Oklahoma and change the whole setup.”
Another potential problem is climate change. While climate and weather are different, the former may be having an impact on near-term forecasting accuracy.
I spoke with Aditi Sheshadri, lead researcher of a 2021 study at Stanford University that explored how warmer Earth temperatures will affect weather prediction. Researchers did some modeling revolving around how different warming patterns could affect weather forecast accuracy in this and other parts of the world.
What they found was a “pretty systematic relationship” between temperature changes and how far out you can accurately predict the weather. Sheshadri noted that we could make more accurate long-term forecasts “if the Earth was very much cooler.” As the Earth warms, “it’s the other way around, and this window of accurate weather prediction narrows.”
According to the Stanford study, you lose a day of accurate precipitation forecasts for every rise of 3 degrees in Celsius (5.4 degrees in Fahrenheit).
It may not seem like much, but according to data from the National Atmospheric and Oceanic Administration, Earth has warmed by .08 degrees Celsius (0.14 degrees Fahrenheit) every decade since 1880. By now, Earth is about 1 degree Celsius (1.8 degrees Fahrenheit) warmer than it was then. The World Meteorological Association warns that the Earth’s temperature will continue to rise over the next few years.
Here’s the bottom line: Those who love weather forecasting but don’t discuss climate change should understand the latter’s impact. Weather forecasting may not be as accurate as it could be because of climate change.
Once a meteorologist has thoroughly reviewed the current weather and ascertained what processes are producing it, the forecaster can begin to look in the future. Forecasters will typically use the “Forecast Funnel” technique. The forecast funnel is a concept that focuses the forecasters’ attention first on large scale processes, and then on increasingly smaller scales.

Numerical weather prediction models are computer simulations of the atmosphere. These models provide the foundation of the weather forecast. The models use an analysis of the current weather as a starting point and then project the state of the atmosphere in the future. The models use complicated physics and fluid dynamics equations that require supercomputers to solve them.

In addition to the numerical weather prediction models, forecasters will draw upon conceptual models, experience, and research to produce the forecast.

The forecasters will then conduct a thorough review the output of these models. At times, the models yield different results, and in these circumstances, forecasters will try to determine which models perform best for the given situation or they will seek a consensus solution. Computer models and other weather data are viewable on the AWIPS workstations. The main graphical interface has the ability to show several “windows” with different data. These can be switched back and forth between the large window and the small windows.