...
Below we can find some forecast evolution plots for the European records registered on the 19th of July 2022 for 2m minimum temperature. CDFs are plotted for Aberporth and Kenley (UK) and the forecast evolution slides are referred to them as well, in the same order.
3.1 Data assimilation
3.2 HRES
Mean error of the UTCI HRES forecasts of the daily maximum UTCI for the period 17-20 July 2022, at lead times 1 day and 5 days ahead, compared to ERA5:
(e.g. right plot is the mean of the error in the daily max UTCI for the forecasts produced on 12th July valid 17th, 13th July valid 18th, 14th July valid 19th, 15th July valid 20th)
UTCI HRES forecast errors, compared to ERA5, for the maximum UTCI on 19th July 2022 (regardless of timing) at 1, 5 and 7 days ahead:
3.3 ENS
Below we can see the EFI for 2m maximum temperature valid on the 19 July 2022 and at lead times day 1, 3, 5 and 7.
...
ENS probabilities of exceeding 35 degrees in 2 consecutive days (valid on 18-19 July 2022). A sequence of plots every 24h from 00-48h to 312-360h.
ENS probabilities of exceeding 35 degrees in 2 consecutive days (valid on 18-19 July 2022). A sequence of plots every 24h from 00-48h to 192-240h.
ENS probabilities > 99 percentile of model climate in 2 consecutive days (valid on 18-19 July 2022). A sequence of plots every 24h from 00-48h to 168-216h.
Below are example UTCI forecasts for probability of different categories of heat stress, at 7 days ahead of 19th July:
And the same for 1 day ahead of 19th July:
3.4 Monthly forecasts
...
3.5 Comparison with other centres
4. Experience from general performance/other cases
...
In general, very good predictability even 15 days in advance, however maximum temperatures were in general slightly underestimated and there were larger errors in the 2m minimum temperature.
