CET

[Hiya]

Just an idea but is it possible to use the GFS data from each run to calculate a figure for the average temperature within the CET zone?

[VillagePlank]

First thing, I guess, is to note, as you do, that GFS is forecast data, and CET is observed data, apart from that, I see no reason why you couldn't produce a forecast CET using GFS data, as long you use the methods as outlined originally by Manley and his subsequent revisors.

The trick is getting hold of the numerical data without figuring out how to decode a GRIB file. An nice XML format would be nice, perhaps GML, but, alas, it's unlikely anywhere in the near future.

[Shiny_Bottom_1]

Hi Hiya.

This is the Net-Wx 16 day ave temp, which I believe is generated from the GFS:

http://www.netweather.tv/index.cgi?action=tmp;sess=

It is for the UK, rather than the CET zone, but the best indications I know of.

Also worth bearing in mind that, as the METO CET zone is more southern biased (in its stations), then it is likely (but not always the case) going to be a bit above those in the link (A classic example was of late when the CET Zone/ Southern UK had colder temps than Scotland).

[Hiya]

What I was really suggesting was that perhaps this forecast figure could be be coded into a feature on extra.

[shuggee]

That's something Paul or Karl would be able to help with - I'll PM Karl to get his view on the techie side

[Karl]

http://www.netweather.tv/index.cgi?action=cet;sess=

that's our cet calculation type thing but based off recorded data

i'd imagine it might be misleading to calc a cet on only 16 days of forecast data - that time when the temperature shoots up in march april could be quite confusing..

correct if i'm wrong here... maybe something for the cfs

[shuggee]

Well I suppose a 16 day moving projection is sufficient. Then mid month each month, there's a full month projection.

CFS would be rubbish - not enough detail and it's a bit too flaky.

[Hiya]

Well I suppose a 16 day moving projection is sufficient. Then mid month each month, there's a full month projection.

Thats exactly what I'm talking about.

I think it would give a good handle to judge a run, or follow a trend in runs.