IBM believes its new weather forecasting system can mitigate shocks from climate change

• IBM’s new weather forecasting model — IBM GRAF — will run on a supercomputer while using AI, cloud and analytics to predict weather events.
• The company claims its new product is three times more accurate than existing global forecasting models.
• GRAF will be able to predict the weather up to 12 hours in advance and its forecasts will be updated on an hourly basis.

Climate change isn’t only crippling the planet but also cost the world nearly $225 billion in 2018 by causing severe distortions in weather patterns making business decisions uncertain. IBM has a plan to mitigate those shocks with its new weather forecasting system that runs on a supercomputer.

"As weather conditions become increasingly more severe across the globe, it’s crucial that businesses across India have access to timely and accurate weather data to help stay prepared," said Cameron Clayton, General Manager of IBM and Weather.


Nearly 72% of Indians believe that the local economy has been affected by a severe weather event in 2019, according to Climate Change Perception in India. The new weather forecasting system — IBM GRAF — will be able to limit that damage.

According to the company, it will be able to predict weather conditions up to 12 hours in advance — something that wasn’t available on a global scale until now.

And, it won’t be doing it alone. Its calculations will be supported by artificial intelligence (AI), cloud, and analytics to help with decision making — for just for businesses, but users as well.

How is weather forecasting going to get better?

IBM promises that its predictions will be more accurate and timely. Current global forecasting models — like the European ECMWF and US’ GFS model — cover 9 to 12 kilometres at a time and are updated every six hours.

In comparison, IBM GRAF is able to shrink that distance to 3 kilometre — which means its high-resolution images will be 3-4 times greater. It will provide a resolution so fine that approximations will no longer be needed to figure out where thunderstorms are headed and how they might behave.

And, its forecasts will be updated six to 12 times more frequently — that’s nearly once every hour.

The primary difference between global models and GRAF is how they predict the weather. Where the ECMWF and GFS solve mathematical equations to figure out how the atmosphere will flow, the GRAF model uses a variable-resolution grid — which kind of looks like a net being thrown over the planet.


“With IBM GRAF, the granularity of insights offered will be an important factor whether it’s helping maximize crop production on agricultural lands or using satellite imagery to guide rescue efforts during a natural disaster,” said Prateep Basu, Founder and CEO of SatSure at the launch.

In India, 89% of people have expressed concern that climate change could have a negative impact on the economy, according to the Climate Change Perspectives report. 66% believe severe weather events are more common now than they were a year ago. In 2019, the country was rocked by natural disasters like Cyclone Fani and the initial withdrawal of the Indian monsoon followed by its delayed onslaught.

See also:
You can experience the end result of climate change today with virtual reality

TikTok now has three new stickers to remind users about climate change

India's Indus water tower is the most important in the world — but it's also the most vulnerable