I noticed that trying to merge two revisions with data downloads git fails and leaves a mess that I think has to be cleaned by hand which is impossible for 20.000+ files in the archive. So I had to use a fresh clone of the repository. So whenever downloading the data make sure that your working copy is up to date and commit right after that, else you'll end up with a mess.

Some changes I found compared to the version downloaded in summer 2017.

The following countries have new or changed data:

• Bosnia and Herzegovina: new data for 2012 and 2014 for individual gases and for aggregate Kyoto gases complete new timeseries from 2001 to 2014. Data unchanged from 1990 to 2000
• Chile: 2010, Kyoto GHG, category 2 datapoint that seemed erroneus in the last version has been corrected.
• Dominican Republic: Data for Year 2010 has been added.
• Ecuador: Data for 2012 has been added. Historical data has been changed (some data points have been removed). 1994 emissions are much higher than emissions for subsequent years.
• Guinea-Bissau: Data for Year 2010 has been added.
• Israel: Data for 2014 and 2015 has been added. Historical time series have been updated slightly.
• Republic of Korea: Data for 2013 and 2014 has been added. Historical CH4 time series have been updated and amended.
• Saint Lucia: Data for 2005 and 2010 has been added. Year 200 data has been adjusted. 1995 emissions still very high compared to other years (originating from CH4, category 6 (waste)).
• Madagascar: Data for 2005 - 2010 has been added.
• Paraguay: Data for 2005,2012 has been added. 1994 data has been revised.
• Saudi Arabia: Data for 2012 has been added.
• Solomon islands: Data has been added for up to four years depending on gas and category. Only few data available for individual gases.
• Thailand: Data for Year 2014 has been added. For aggregate Kyoto GHG a complete timeseries from 2000 to 2013 has been added.
• Viet Nam: Data for Year 2013 has been added.

Small changes in individual sectors might be missing as countries where only investigated in detail where changes we visible in CATM0EL Kyoto GHG data.

General problems

• Belize: Aggregate Kyoto GHG CATM0EL data does not coincide with calculated sums for years 2003, 2006,and 2009. (not a change actually)

https://unfccc.int/process/transparency-and-reporting/greenhouse-gas-data/ghg-data-unfccc states

In response to the mandate received from the SBSTA, the secretariat has made available the following six modules of the GHG data interface reflecting data reported by Annex I Parties and, to the extent possible, information provided in the national communications and biennial update reports of non-Annex I Parties as of 27 May 2018:

cc @AnnisG @AnnGuenther @mljeffery

Before doing a new release (#6) i'd like to rewrite the code to take the used years from the data and not prescribe them.

Thanks for providing this nice and very helpful data package!

I have discovered one problem concerning the CO2 emission from the German category 1.A.2.g: Its only sub-category with reported emissions (1.A.2.g.vii) does not match the emissions reported under 1.A.2.g. The discrepancy is quite substantial e.g. in 2019 1.A.2.g reports 74947.20 kt while 1.A.2.g.vii only 3450.54 kt.

The numbers from the database you provide match those from the UNFCCC data interface, thus I assume a reporting error. Or do I miss something?

Robbie Andrew pointed out that direct CSV downloads from the interface have fewer digits than the one provided in this dataset.

This is due to the raw original JSON data having these extra digits, e.g.

"data":[{"id":3,"name":"Australia","rows":[{"id":8787,"name":"2.A.1  Cement Production","unitId":5,"cells":[{"column":0,"numberValue":3462.87171200000010},

https://github.com/JGuetschow/unfccc-detailed-data-by-party/blob/master/archive/annex-one/2.A%20Mineral%20Industry___2.A.1%20Cement%20Production___CO%E2%82%82.json

Should maybe rounded as in the downloadable CSVs.

cc @JGuetschow

The last year of the data (2016, in this case) is not shown in the final csv file.

The API returns 2016 and the year is present in the json files but not in the final csv file.

I guess at one point the API call returned "Last Inventory Year (2016)" and index[indexed_keys[-1]][-5:-1] transformed this to simply "2016". Now it seems the API returns "2016" and this operation transforms it to "201". In my local version I've commented out this line and everything works fine. Not sure if this behavior can be verified and changed in the source code here?

Also later in the file there is this line:

ordered = [
            "Party",
            "Parent Category",
            "Category",
            "Gas",
            "Unit",
            "Base year",
        ] + list(filtered.columns[5:-1])

removing -1 from the last line seems to fix this.

Has been updated ...
https://unfccc.int/process/transparency-and-reporting/greenhouse-gas-data/ghg-data-unfccc

Will do an update run.

/cc @JGuetschow

The secretariat has completed the upgrade of all nine modules of the GHG data interface on 8 November 2019.

How are global warming potentials for aggregate gas time series treated. Do the data from the UNFCCC interface use the same GWPs for all countries (and if so which)? If different GWPs are used how is this indicated in the interface and in this dataset?