Long before this year’s hurricane season officially began on June 1, researchers at the Eversource Energy Center were analyzing storm and utility infrastructure characteristics to predict outages from hurricanes and other storm types. Severe weather is among the leading causes of outages on the overhead electric distribution grid, and historic events like Storm Irene (2011), the October nor’easter (2011), and Hurricane Sandy (2012) have shown a need for utilities to improve their emergency preparedness, response and mitigation strategies. Adequate planning before these disasters can relieve emergency preparedness issues by giving better predictions of storm damages and the expected length of time before power is restored. Such preparation can help utilities allocate equipment and personnel more efficiently, and the public can better manage their expectations about when the power will return.
The UConn Outage Prediction Model (OPM) predicts an upcoming storm’s impact, including the number and location of outages, so that a utility can proactively dispatch crews before storms arrive, and provide intelligence as to whether outside crews should be put on standby or called in. The OPM is trained by state-of-the-art high-resolution weather simulations for more than 160 storm events, which span over a decade (2005 – 2016), occurred during different seasons and represent varying severities (from isolated thunderstorms to hurricanes). To stay relevant, the model is updated regularly such that each storm event that impacts the region is included as a data point in the model.
At this point, most forecasting groups expect the 2016 Atlantic Hurricane season to be a near average to above average season, due to a combination of factors including an expected transition to La Niña and warmer than normal sea surface temperatures in the Gulf of Mexico, Caribbean, and Western Atlantic.
To access UConn’s daily weather forecasts, please go to www.cee-wrf.engr.uconn.edu.
Published: July 25, 2016
Categories: Recent News
Available Archives