Create a build excluding non-commercial data

[bkeepers]

As Dave Flater points out:

The TICON-4 constants have a CC BY 4.0 license, but they are derived from GESLA-4 time series data, which have mixed licenses that are not fully spelled out.

According to the GESLA website:

For example, users wishing to use the records provided by CV, UZ and CMEMS for consultancy purposes, must contact these organisations to obtain permission first (or in the case of CMEMS the organisations that provided the data to them).

I'd like to review the filtering added in #49 to prioritize free sources, and depending on the result, publish a version of the data excluding non-commercial stations.

[bkeepers]

Some quick analysis: 534 non-commercial stations in the dataset (14.6% of 3,660 TICON stations):

• CMEMS: 532
• CV (City of Venice): 1
• UZ (University of Zagreb): 1

These are almost entirely European — France (111), Netherlands (69), Spain (52), Denmark (49), UK (44), Germany (36), etc.

If deprioritized, 84 CMEMS stations would be replaced by free-license alternatives. The CMEMS stations currently win on record length against nearby stations from refmar, smhi, nhs, ispra, bodc, mi_c, fmi, bfg, etc. Deprioritizing makes those free sources win instead.

The remaining 450 stations that can't be eliminated are concentrated in:

• France: 31 (no nearby refmar equivalent)
• Spain: 30
• Denmark: 27
• Estonia: 20
• Italy: 18
• Greece: 16
• Turkey: 15
• Netherlands: 15
• Latvia: 9, Bulgaria: 6, Norway: 5, Portugal: 5, UK: 4, etc.

Deprioritizing these sources eliminates ~16% of the non-commercial stations (84 of 534). The remaining 450 are the sole coverage for their locations — removing those would create gaps in European coverage, particularly in the Mediterranean, Baltic, and Scandinavian coasts.

At a minimum these need to be marked with commercial_use: false in their source.