Starting to wonder if I should have waited for the October thread to open before sending my prediction - again we've seen a shift of model output past the end of the preceding month.
Back on 30th Sep the model runs were giving me mid-month CET estimates in the 10.3 to 11.0°C range, about 1.0 to 1.5°C below average.
Yesterday's GFS 06z came in at 11.9°C, about average, while today's 12z is way up at a balmy 13.0°C, about 1.2°C above the norm!
Still, I'm not actually too concerned at this stage. My near-average estimate hides a cool-warm-cool anticipated pattern for the month.
Can only wait to see if the cool final part actually happens, but in the meantime, with a lot of anticyclonic weather forecast including light winds, along with maritime airmasses bringing ample moisture, I'd not be surprised to see lingering fog keeping temperatures down in places by Sunday.
After all, GFS has a strange habit of keeping temperatures too high by night, then too low by day, which means it can fail to pick up on the inversions that support fog and, sometimes, lead to it having a high error on the day temps too.
It's a bit of a close call with a seriously warm regime, though - the high would only have to shift a few hundred miles southeast to put the CET region in a steady southwesterly flow with impressively high temps either by night or day, depending on cloud amounts.
So, I'm not resting easy, just not too concerned... at least for now 
.