Forecast verification

Len

unread,

Apr 2, 2024, 9:58:13 AMApr 2

to Weather and Climate

As a follow-up to Graham's observation that the UKMO forecast temperatures for the Southwest are often too low by a couple of degrees I monitored them for max temp. over

7 days last week at my postcode PL9

Forecast issued at 1638 GMT Mon. 25th March

Forecast Measured

25th Mon. 10°C 10.1°

26th Tues 11°C 11.9°

27th Wed. 8°C 10.5°

28th Thurs. 11°C 12.5°

29th Fri. 11°C 13.1°

30th Sat. 12°C 14.1°

31st Sun. 13°C 14.1°

So the case is proven over this period.

The forecast of max temperature 24 hrs or more ahead is between one and 2 degrees below actual.

Measurements taken in Stevenson screen.

Len

Wembury, SW Devon

Freddie

unread,

Apr 7, 2024, 3:42:30 AMApr 7

to Weather and Climate

The correlation with wind direction would be interesting. Perhaps modelled sea temperatures are on the low side?

Graham Easterling

unread,

Apr 8, 2024, 7:23:49 AMApr 8

to Weather and Climate

Yes, I think the modelled sea temperatures are a bit of a problem. Currently the SST off Cornwall is 11-11.5C whichever buoy you look at (Porthleven, Looe, Penzance, Perranporth.) SST anomaly all around the SW coast is around 1.5C

As far is Penzance is concerned the big errors occur with an offshore (W--NW wind), and sunshine (if only!) in spring/early summer. The models seem to significantly understate the effect of the wind being offshore, perhaps with the peninsula being so narrow here? Also, with the manned station at Camborne being on the north coast perhaps there is little appreciation of how much warmer the south coast is under these conditions. Though I would have thought that would only affect a 'human' forecast.