IBM GRAF: Vastly improving accurate weather forecasts

By The Weather Company

Contributing author: Cameron Clayton

Weather influences what people do—where we travel, what we eat, even how we feel. Weather is arguably the most important external swing factor in business performance, potentially responsible for nearly half a trillion dollars in economic impact each year in the U.S. alone.

At the same time, reporting of extreme weather events—droughts, wildfires, heavy rainfall, hail, floods—is becoming more common. As such, the world needs better, more accurate, more finely-tuned weather forecasts.

The challenges of accurate weather forecasting

Accurate weather forecasting and prediction present a number of challenges. It requires collecting and analyzing vast amounts of data from weather radar and other sources, which is often expensive and time-consuming. In addition, weather systems can range from a small thunderstorm that quickly passes to a large-scale snowstorm that is thousands of miles in diameter. Lastly, there needs to be a clear and effective way to communicate these forecasts to the community who are expecting accuracy and hour-by-hour updates.

Introducing the IBM Global High-Resolution Atmospheric Forecasting System

For these reasons, IBM is introducing the IBM Global High-Resolution Atmospheric Forecasting System. The system, called IBM GRAF for short, is the first hourly-updating weather system that is able to predict something as small as a thunderstorm virtually anywhere on the planet. 

IBM GRAF represents a huge step forward. Until now, much of the world had to settle for predictions from models that covered 10- to 12-kilometer swaths of land—too wide to capture many weather phenomena. In addition, leading weather models update less frequently, every 6 to 12 hours in much of the world. In contrast, IBM GRAF provides a 3-kilometer resolution that updates hourly, delivering reliable predictions for the day ahead.

Not only will this much-improved weather system help people and communities better plan for upcoming weather conditions, it will help:

  • Government disaster management departments better prepare & optimize resources
  • Utility companies optimize operations & keep workers safe
  • Airlines more effectively manage turbulence
  • Farmers make more informed decisions as it relates to high value tasks like when to fertilize crops
  • Broadcasters present more accurate graphics due to more rapidly updating weather information
  • Insurers or retailers better adjust messaging in weather impacted regions

Using terabytes of weather data to your business’ advantage

All of this is possible because the system is able to pull previously untapped sources from millions of sensors worldwide and in-flight data, and it also runs on a hardware platform capable of processing the massive volumes of data required. IBM GRAF is the first commercial weather system with the ability to include wind speeds and temperature readings taken by sensor readings on airplanes all over the world. 

This is made possible through a relationship with FlightAware, which works with airlines and air traffic control stations globally to collect 10 million high-altitude data points from aircraft every day. Each plane will transmit readings every 5 seconds, while flying through the atmosphere at 500-600 miles per hour.

IBM GRAF is also able to tap into additional weather data by harnessing pressure sensor readings sent from barometers found within smartphones if people opt in to sharing that information. The Weather Company will assure it conforms to the relevant operating system terms of use. Additionally hundreds of thousands of weather stations, many run by amateur weather enthusiasts, can also contribute data to the model. 

Supercomputers: Providing high precision & high consistency

Supercomputers are used to run nearly all weather forecast models. In the U.S., Japan and a handful of other countries primarily in Western Europe, supercomputers are available to run storm-scale weather forecast models many times a day. In the rest of the world weather forecasting can be less precise and less consistent, mainly due to a need for computing horsepower.

IBM GRAF is the first large scale global weather model to run on graphics processing units (GPUs), enabling it to run faster and more efficiently than older models. More specifically, GRAF is composed of 84 nodes of the IBM Power Systems AC922 server and uses 3.5 petabytes of IBM Elastic Storage capacity to keep the model fed. This is the same technology used by the U.S Department of Energy in the Summit and Sierra supercomputers, the two most powerful computers in the world.

Incorporating Global Weather Model MPAS for accurate weather prediction

This newest weather prediction system is made possible by The Weather Company’s open-source collaboration with the National Center for Atmospheric Research (NCAR). IBM GRAF incorporates the latest-generation global weather model – the Model for Predictions Across Scales, or MPAS – which was developed by NCAR with the Los Alamos National Laboratory. 

"This is a great example of how long-term basic research funded by the federal government has created an industry opportunity that is both good for the bottom line and protects lives and property,” said Antonio Busalacchi, president of the University Corporation for Atmospheric Research, which manages NCAR on behalf of the National Science Foundation.

Final thoughts

IBM GRAF is the most sophisticated and farthest-reaching forecasting system ever made. The best part? Anybody with The Weather Channel app on their smartphone will benefit from it not only with access to improved forecasts, but can also—with permission—share pressure sensor readings via the app to improve accuracy. They are contributing to democratizing weather data.

Learn more about IBM GRAF and how it can vastly improve your weather forecasting capabilities. And if you would like to learn more about keeping up with the broadcast media industry, contact us today.

Learn more about our weather solutions for broadcast media companies