It'll help those with estimates in the 17s+ if the CET stations can exceed the model maximums as much as my locality has been of late.
After reaching 25.6*C today, which even I was surprised by (expected 24s like yesterday given a breeze from the S but even Bournemouth Airport was untroubled with a max of 25.5*C there - Channel cooling cancelled!), the integer FV3 model errors look like this for July so far.
Columns are prediction range (days), minimum temp error (*C), maximum temp error (*C).
| 2 | -1 | -2 |
| 3 | -1 | -3 |
| 4 | -5 | -3 |
| 3 | 2 | -3 |
| 4 | 1 | -1 |
| 3 | 0 | -5 |
| 2 | -4 | -2 |
| 3 | 2 | -2 |
| 4 | -2 | -5 |
| 3 | 1 | -3 |
As you can see, the minimum errors are inconsistent in both magnitude and direction (5 negative, 4 positive, one neutral), but the maximum errors are consistently negative and predominantly between -2 and -3*C. I use integers for easier at-a-glance assessment.
The existence of three-5s, with two at just 3 days range, is laughably sad
.
Originally Posted by: Stormchaser