Fugaku is the fastest supercomputer on the planet. Recently built by Japanese company Fujitsu, it’s capable of 2.6 quadrillion operations per second. This staggering processing power is now at the service of the Japanese Meteorological Research Institute to help weather and climate forecasting and, above all, disaster warning.
Extreme weather, largely fuelled by climate change, is an increasing liability to the world’s economy. More than ever, countries and companies need accurate predictions in order to protect their people and their operations from disaster. This is why weather forecasting has grown into an industry that, in 2015, was estimated at US$56 billion, and is certainly bigger today.
Last year, as typhoon Hagibis was approaching its shores, Japan issued a special warning, suspended the Shinkansen trains and shut down supermarkets well in time. As extreme events become the new normal, trucking companies, commodity traders and utility providers – not to mention insurance companies – need reliable hour-by-hour forecasts and analysis, just to save money. Weather services used to be run by governments, but now a growing number of start-ups are attracted to a booming market: the market of climate insecurity.
In late August, hurricane Laura made landfall at 150 mph in almost the exact location in Louisiana predicted 3.5 days earlier. Such a result would have been unthinkable a few decades ago, yet it’s precisely what is needed as we move forward into a more uncertain world. A constant increase in processing power, coupled with artificial intelligence, machine learning and cloud-based systems are anticipating the near future. Not only do they tell an airline to reschedule flights to avoid storms, or suggest to a farmer when to irrigate crops, they also inform millions of people when to evacuate from a hurricane’s path, or simply when it’s time to grab an umbrella.
However, the very people who rely on their smartphone’s weather apps often confuse meteorology with climatology. The former tries to predict atmospheric behaviour in the very short term, the latter in the very long term (including other factors, such as the Sun’s radiation). It’s strange that countries, companies and families promptly react to the flash warnings of meteorologists, but still fail to act in the face of the dire predictions of climatologists. They both use the same processing power, the same artificial intelligence and essentially the same science.