BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

IBM Developed A High-Resolution Weather Model That Covers The Entire World

Following
This article is more than 4 years old.

Some of the most powerful hurricanes begin with thunderstorms that move off of Africa. What if a weather model could predict the development of those individual thunderstorms when they’re still over the African savannas? IBM, in collaboration with its subsidiary The Weather Company, has developed a new high-resolution weather model it claims will cover the entire world with a resolution and update frequency unseen in other weather models in use today.

The new IBM GRAF, short for Global High-Resolution Atmospheric Forecasting system, will utilize a finer grid—giving users a more detailed view of the atmosphere—and a faster update speed than other global models currently in operational use. IBM GRAF boasts a grid of just 3 km and updates on an hourly basis, which would allow forecasters to quickly predict the development of small-scale weather features like thunderstorms or bands of heavy rain and snow anywhere on Earth.

Meteorologists use two major global models to help them predict the weather. NOAA’s GFS model and the ECMWF (European) model both run several times a day—four times daily for the GFS and twice for the European—with a grid that allows the models to accurately resolve large-scale weather systems like the jet stream and nor’easters. Unfortunately, the tradeoff to having a large grid (the GFS model’s grid is 13 km) means that these models have a tough time resolving small-scale features. That’s where high-resolution models prove their worth.

We’ve had high-resolution weather models for a while. It’s common to see high-resolution models on the nightly news or on social media before a big severe thunderstorm outbreak. These models can simulate radar imagery to give you a good idea of what towns are at risk for strong thunderstorms, and about what time you can expect them to roll through.

The most common high-resolution model in the United States is the High-Resolution Rapid Refresh (HRRR) model. The HRRR’s fine grid helps meteorologists issue accurate short-term weather forecasts down to the county level during weather events where we can see big differences over short distances.

Up until now, only a handful of regions around the world had access to these high-resolution weather models. The Weather Company touts IBM GRAF as the first weather model to bring that kind of granular resolution to the entire world. IBM’s press release ahead of the model’s rollout touts the model’s benefits, ranging from farmers who live in flood-prone areas of Africa to airplanes routing around hazardous storms.

IBM GRAF will not be available for public consumption in the same way we have open access to models like the GFS or HRRR. Instead, the general public will indirectly benefit from the model through improved forecasts and products offered by IBM and The Weather Company, which includes forecasts issued by The Weather Channel and Wunderground.

Follow me on TwitterCheck out my website