good point @SimonLeeWX . Using different regime definitions allows somewhat to deal with flow-dependent predictability as we advocate in our @ECMWF article https://www.ecmwf.int/en/newsletter/165/meteorology/how-make-use-weather-regimes-extended-range-predictions-europe (1/6) https://twitter.com/SimonLeeWx/status/1328458477342167047
Thinking in NAO alone (=EOF1) would mislead you in a rather certain NAO+ situation. Z500 ensmean mostly reflects climatology due to the cancelling out of anomalies of the contrasting forecast scenarios. EOF1/2 space as well as 4 regimes unveil the competing BLO projections. (2/6)
Check out: https://apps.ecmwf.int/webapps/opencharts/permalinks/500-hpa-z:-weekly-mean-anomalies-40470 via @ECMWF (3/6)
Check out: https://apps.ecmwf.int/webapps/opencharts/permalinks/weather-regimes-probabilities-98370 via @ECMWF (4/6)
The latest 7WR forecast, which I unfortunately can not share here, shows that the ensemble competes between a group of members tending to Atlantic Trough/Scandinavian Trough or Greenland/Scandinavian/European blocking, thus the mild or cold scenarios for Northern Europe. (5/6)
Regarding the generally low skill > week 2 let us bet not too much
(6/6)
