My usual method gives me an estimate of 17.7°C for the August CET using GFS numbers. Trouble is, that model has been more awry than ever lately with a large positive bias on the daily maximums, partially offset by a mostly (but not always) low bias on minimums.
For example, the 00z of 15th gave an estimated mean CET-to-date of 18.5°C for 20th, but the actual was 18.1°C.
OTOH, with a change to much more unsettled weather, the positive bias ought to reduce. Of late GFS has been predicting too little cloud cover and this has combined with a problem with its soil moisture parameter, making for too much daytime increase in temperature.
With that in mind, what goes on today through Sunday will be important. We're still in a more settled pattern but one prone to stubborn low-level cloud limiting max temps. For example UKV paints a pessimistic picture for today and tomorrow with a lot of cloud widely limiting max temps to the low 20s instead of GFS' mid-20s °C, before a general clearing up during Sunday allowing for mid-20s °C for many. With that scenario I could see the final CET being as low as 17.4°C.
Originally Posted by: Stormchaser