That is correct- important to remember that while the mean may give an indication of trend, it really is just a mathematical outcome from every input, including any number of unlikely or even anomalous extremes. To illustrate it in an extreme way, it would only take one P with a +20c 850 output to edge up the mean by over half a degree; you get a few wildcards and it doesn't take long for the line to move up. Key here I think is the continuing signal for the mean of outputs to be visibly below the lta.
There is still a stretch of sub -5 means, great in any winter spell if you hope for cold, but as always, not a certainty.
Originally Posted by: Bertwhistle