This week marks a new era for weather forecasts thanks to an upgrade to the National Oceanic and Atmospheric Administration’s flagship forecasting model.
The Global Forecast System, used to help forecasters predict severe weather, winter storms and tropical cyclones, will incorporate a new component called the Finite-Volume Cubed-Sphere. FV3 is a dynamical core, a program that turns wind and air pressure readings into a numerical weather prediction.
In a news release Wednesday, Secretary of Commerce Wilbur Ross said the upgrade is the latest in a series of forecasting improvements for NOAA and its subsidiary agency the National Weather Service.
“In the past few years, NOAA has made several significant technological leaps into the future — from new satellites in orbit to this latest weather model upgrade,” Ross said. “Through the use of this advanced model, the dedicated scientists, forecasters, and staff at NOAA will remain ever-alert for any threat to American lives and property.”
The FV3 upgrade, initially designed to predict weather patterns stretching out several decades, combines the long-range, global climate data modeling with day-to-day speed and reliability of the GFS model. Steve Piltz, meteorologist in charge at the National Weather Service office in Tulsa, said the upgraded model will help forecasters better predict all modes of weather, particularly smaller, isolated systems.
“It’s really designed to handle smaller-scale features better in the atmosphere,” Piltz said. “The global models have been doing a pretty good job of handling the larger scale systems, the ones that cover several states. The smaller scale systems are the ones we have difficulty tracking and timing over time.
“The new model will better resolve those smaller-scale features and help maintain their continuity in the forecast as we go out from day zero.”
Larger systems like long squall lines racing across the plains, a vast winter storm blowing out of Kansas, or a swath of fast-moving supercells like Oklahoma experienced several times in May, are more ideal for the global model’s tracking, Piltz said. A large system is easier to see coming, and Piltz said it’s also easier to estimate when it will arrive.
Those smaller, more isolated systems of only a few strong cells or a brief, localized swath of freezing drizzle could have been missed in the previous long-range models, Piltz said. With FV3, the weather service will have a more consistent idea of when the smaller systems will impact the area, Piltz said.
“We anticipate maybe a better result throughout the entire seven-day forecast, rather than just catching those systems at the very beginning of our forecast,” Piltz said.
Scientists tested the GFS-FV3 model alongside the legacy GFS program for a year, using real-time evaluations to compare the strengths and weaknesses of each model. After the testing, the FV3 upgrade was reportedly better “across a wide range of weather phenomena,” according to a news release.
Piltz said FV3 is also better equipped for updates than the previous GFS model. With multiple laboratories and researchers contributing to the project, there’s a greater opportunity to improve and develop the model as time goes on.
Louis Uccellini, director of the National Weather Service, said in a news release FV3 will allow greater accuracy for weather forecasts not only the day of, but for the rest of any given week.
“Switching out the dynamical core will have significant impact on our ability to make more accurate 1-2 day forecasts and increase the level of accuracy for our 3-7 day forecasts,” Uccellini said. “However, our job doesn’t end there — we also have to improve the physics as well as the data assimilation system used to ingest data and initialize the model.”
The legacy GFS model will reportedly run through September 2019 so users can compare performance, but it won’t be used in regular forecast operations.
Take a peek inside the National Weather Center in Norman.