Following the October storm that cut power to thousands of customers, researchers say it may be time to devise new models to predict storm outages. Emmanouil Anagnostou is a professor at the University of Connecticut. He says existing models do really well at building connections between historic and new storm data. But they’re not great at predicting more extreme weather events.
“That also ties with climate change,” Anagnostou said. “If we want to use the model in a way to understand the impact of future storms, you cannot do that unless you have the opportunity to extrapolate.”
Anagnostou says utilities need a lot of information to prepare for a storm, including weather forecasts, maps of where the power lines are and data on the condition of nearby trees.
“So that’s the complexity, really, of the problem,” Anagnostou said. “That we really need to have a model, or a system, that can predict that range of possible impacts from different storms.”
According to Anagnostou, the statistical models currently in use failed to predict the severity of October’s late-month storm, and as a result it took up to five days to reconnect some customers.
Anagnostou suggests tweaking the system by using new data. And pairing that data with more autonomous “machine learning” models to better predict storm outages.
This report comes from the New England News Collaborative: Eight public media companies, including Rhode Island Public Radio, coming together to tell the story of a changing region, with support from the Corporation for Public Broadcasting.