The 18z and 00z NAE T+0 2-metre temperature fields on Weatheronline look very
odd. At 00z it shows an extensive area of -3 to -6 deg extending from Sussex,
right up through the West Midlands to North Wales. This isn't supported by the
observations which show 00z temps of zero to -2 deg across this area. The 18z
T+0 had a similar, though rather smaller area of incorrectly low temperatures.

I see that there's nothing beyond T+0 for the 00z run so maybe the model has
thrown a wobbly. Or is it me who's got this wrong?

Norman