glenh

The above chart from 'I'm dreaming of' looks remarkably similar to this: That was from further north, hence the colder temps, but a similar look relative to the 30 year mean

I think this is a misunderstanding of how deterministic numerical weather models work. They don't produce output based on historical data.

Using a location central to the whole UK (hopefully avoiding N/S arguments), the GFS ensemble mean is for the most part below average for the whole forecast period after Saturday:

GFS ensembles certainly seem to be firming up on a noticeable cool (and settle) down mid month: