This is something I've seen before. Prior to #SSW, model fails to see any impacts. Once disrupted #polarvortex is initialized into forecast, suddenly the forecast is for basically the composite (historical mean) SSW pattern. (These are all forecasts for Feb, but diff init dates). pic.twitter.com/zZOMkI27HE

— Dr. Amy H Butler (@DrAHButler) January 16, 2019