A nice illustration of the ECMWF EPS's struggle with predicting the surface -AO. The model is now predicting a stable -1.3 sigma event. pic.twitter.com/WPvP6pvKjE

— Mike Ventrice (@MJVentrice) January 5, 2019