Short Wave

How scientists predict big winter storms

January 28, 2026

Key Takeaways Copied to clipboard!

  • The ability to predict major winter storms days in advance, as seen with Winter Storm Fern, is due to significant advances in computer weather models built upon decades of coordinated Earth system observations. 
  • The accuracy of weather models hinges on the quality of input data, which must be plentiful, granular (covering all dimensions of the atmosphere), and continuous over long periods to accurately forecast rare extreme weather events. 
  • Potential budget and staff cuts to U.S. government agencies like NASA and NOAA, which maintain crucial, long-term observational data sets, pose a threat to the continued accuracy of advanced weather forecasting in the future. 

Segments

Winter Storm Fern Impact
Copied to clipboard!
(00:00:20)
  • Key Takeaway: Winter Storm Fern affected at least 29 states, causing widespread snow, sleet, and ice, and knocking out electricity for hundreds of thousands of people.
  • Summary: The recent winter storm stretched over 2,300 miles, impacting about half the U.S. population with heavy snow, sleet, and ice. The storm affected at least 29 states across the country. This event resulted in power outages affecting hundreds of thousands of residents.
Advances in Storm Prediction
Copied to clipboard!
(00:01:21)
  • Key Takeaway: The extended lead time for predicting major storms like Winter Storm Fern is attributed to amazing advances in computer weather models, which rely on coordinated observations of the Earth system.
  • Summary: Scientists can now predict events days in advance, a capability unavailable 50 years ago. This improvement centers on better computer weather models. These models are built upon large, coordinated efforts in observing the Earth system.
Understanding Weather Models
Copied to clipboard!
(00:04:01)
  • Key Takeaway: Weather forecasts are typically based on a weighted average of multiple computer models, such as the European model, which vary in their strengths across different weather types and spatial scales.
  • Summary: Good computer models are the key to accurate forecasting, and many different models exist, often identified by acronyms or nicknames. Forecasts seen by the public are usually derived from an average of these models. Different models excel at predicting specific weather phenomena or handling regional versus city-level scales.
Data Requirements for Models
Copied to clipboard!
(00:05:58)
  • Key Takeaway: The most crucial requirement for effective weather models is good data, specifically data that is plentiful, granular (covering all atmospheric dimensions), and continuous over long time scales.
  • Summary: Computing power is one element, but good data is the most crucial requirement, following the principle of ‘garbage in, garbage out.’ Plentiful data is needed due to the atmosphere’s complexity, including clouds, moisture, and temperature interactions. Granular data requires measurements from the land, air column (balloons, planes), ocean (ships), and space (satellites).
Threats to Data Collection
Copied to clipboard!
(00:08:38)
  • Key Takeaway: Government budget and staff cuts targeting agencies like NASA, NOAA, and the National Center for Atmospheric Research threaten the continuous, publicly funded data collection essential for feeding accurate weather models.
  • Summary: Much of the vital observational data is maintained by governments because satellites, buoys, and balloons are publicly funded. Cuts pursued by the Trump administration have already led to staffing shortages impacting weather balloon launches. If these cuts proceed, maintaining the current level of accurate, early storm forecasting will become difficult as weather becomes more extreme.