ECMWF - Should the model data be freely-accessible

[Su Campu]

Just something that's bugged me for quite a while now. We have a wonderful resource in the ECMWF, which is supported by tax revenue from 30 countries. Maybe THE best model in the world, however, apart from a few basic model fields, the vast bulk of the model data are not freely-available. I know some national agencies (notably Iceland) have at least better time intervals available, but still with only the basic fields (and CAPE, in the case of Iceland).

Compare this to the US, from where we have an endless list of models fields available from the GFS, NAVGEM, etc., freely-available to the World and its mother. Tax dollars pay for these, so why can't our tax euros/pounds/krona/francs give us access to data that we pay for?

Edited January 3, 2015 by Su Campu

[Nick L]

While it is partly funded by the taxpayer, it is also funded by forecasting companies who pay a hell of a lot of money just to get access to some of the data. The ECM is far better than the GFS, so we would either need to increase the amount of tax funding ECMWF or reduce the quality of the model.

 

The fact of the matter is, the vast majority of the public aren't bothered about looking at the data.

[The Eagle]

It's primarily funded by the met offices throughout Europe. Personally I don't believe it should be free as we should at least get some return for the money put in.

 

We get enough data from it anyway imho.

Edited January 6, 2015 by The Eagle

[mountain shadow]

The Euro tax payers should at least get access to a percentage of the output in direct relation to the tax money put in.

 

It is a disgrace what we get out of it.

[Danielvn]

The Euro tax payers should at least get access to a percentage of the output in direct relation to the tax money put in.

 

It is a disgrace what we get out of it.

and how would this be policed? Should we get more than Romania? Should Germany get more or less than Italy?

[Nick L]

The Euro tax payers should at least get access to a percentage of the output in direct relation to the tax money put in.

 

It is a disgrace what we get out of it.

 

But the Met Office gets the data back, and makes public forecasts based on what the taxpayer has paid for.

 

Calling it a disgrace is a bit melodramatic, the fact is that 99.9% of taxpayers would have no interest in seeing the data, nor would they be able to understand it!

[Thundery wintry showers]

It is a difficult topic area.  Giving away more outputs for free encourages more development and spread of knowledge among the general public (e.g. casual weather enthusiasts are more able to develop their own forecasts from the GFS outputs than what little we get from the ECMWF), but it potentially makes funding more problematic which can result either in a loss of quality, or in the public being denied free access to something else (or some of both).  I don't think there is a definitive right or wrong answer here- I see a strong case for both approaches.

 

The GFS approach is most useful for meteorologists who are interested in long-range forecasting and hence the GFS is often used by private weather companies.  For short to medium-range forecasting I don't think the GFS really provides much that the free outputs from the ECMWF and Met Office (notably the FAX charts and the ECMWF ensemble mean) don't.  What we currently have may well be the best of both worlds- the NOAA/GFS giving a lot away for free for the benefit of the experimental long-range forecasters, and the ECMWF keeping most of the advanced stuff behind a paywall and focusing on maximising the relaibility of its short to medium range outputs (some of which are of course freely available).

Edited February 12, 2015 by Thundery wintry showers

[Pennine Ten Foot Drifts]

And in terms of overall performance, just how much 'better' than the GFS model is the ECMWF model anyway ? For example, if the level of data available from the ECMWF had been the same as the GFS, would the Model Output Discussion thread post accuracy, in terms of those posts by the likes of Tamara et al who try and use all available data outputs to produce some kind of mid-range outlook, have been much different ? Personnally I doubt it.

[Yarmy]

I'd prefer it if the published model runs only went out to T144, but in smaller increments (same goes for the UKMO). It would then be easier to compare the divergence between the major models in the semi-reliable time frame. Other than that, it's fine: they are a commercial operation in a competitive environment.

[Gibby]

I think there is more than enough data out there for me to get my head around anyway. Your only likely to find that more data leads to muddier waters and over analystic qualities which we have seen on here and elsewhere already in Winter. I agree with what's said this would have no value other than to the 1% or less of the population that is as fanatical over the elements as we are so could be deemed by the vast majority of the populous as a waste of tax payers money.

[Gibby]

And in terms of overall performance, just how much 'better' than the GFS model is the ECMWF model anyway ? For example, if the level of data available from the ECMWF had been the same as the GFS, would the Model Output Discussion thread post accuracy, in terms of those posts by the likes of Tamara et al who try and use all available data outputs to produce some kind of mid-range outlook, have been much different ? Personnally I doubt it.

 

I publish the verification statistics within my reports each day and since I have started it a month ago only once has GFS come up with better verification rates at any timescale through the 10 day rating. 1st ECM. 2nd UKMO and 3rd GFS