AI Models Promise Neighborhood-Level Weather Forecasts

In Hong Kong, a startup promises to simulate and predict potential flooding for every street in the city in under three minutes with the help of artificial intelligence. The firm, Stellerus, is one of many around the globe looking to capitalize on the need for hyperlocal forecasting as climate change makes extreme weather events more dangerous and less predictable.

AI weather modeling offers not only higher-resolution predictions, but it can also make them faster and at a lower cost than traditional methods. Some insurance companies hope it can help them tailor policies to specific customers’ needs. But as Mary Hui writes, the growing ecosystem of commercial weather forecasters and risk modelers could also mean greater dependence on proprietary models that can’t be double-checked. Today on CityLab: AI Weather Models Promise Super Granular, Specialist Forecasts