In the wake of COP21, I thought that it would be interesting to compare the respective pathways of China and the U.S (and others) based on official data. I still plan to post on this topic, but obtaining official data on the pathways proved much more difficult than I anticipated. Leading into the COP21 conference on October 31, 2015, the UNFCCC Secretariat published its “Synthesis report on the aggregate effect of the intended nationally determined contributions”, the terms of reference of which were described as follows:
The UNFCCC Synthesis Report was 66 pages long. However, it contained zero information on the commitments of the individual countries or even regions. Searching for such information will be the topic of today’s post.
The only relevant quantitative information in the UNFCCC Synthesis Report was on global emissions for several past years and projected levels in 2025 and 2030 as follows:
Not a shred of data anywhere in the Synthesis Report on commitments of individual countries.
The historical figures in Synthesis Report paragraph 191 shown above tie closely to the corresponding numbers in the IPCC WG3 Figure TS.1 for 1990, 2000 and 2010 (38, 40, 48 Gt) as shown below. So while there are a variety of accounting definitions of GHG emissions and one has to take care to compare apples to apples, the UNFCCC seems to have had the same unit in mind as IPCC WG3.
Figure 1: IPCC AR5 WG3 Technical Summary Figure TS.1. Total annual anthropogenic GHG emissions (GtCO2eq / yr) by groups of gases 1970 – 2010: carbon dioxide (CO2) from fossil fuel combustion and industrial processes; CO2 from Forestry and Other Land Use 4 (FOLU); methane (CH4); nitrous oxide (N2O); fluorinated gases5 covered under the Kyoto Protocol (F-gases). At the right side of the figure, GHG emissions in 2010 are shown again broken down into these components with the associated uncertainties (90 % confidence interval) indicated by the error bars. Total anthropogenic GHG emissions uncertainties are derived from the individual gas estimates as described in Chapter 5 [126.96.36.199]. Emissions are converted into CO2-equivalents based on Global Warming Potentials with a 100-year time horizon (GWP100) from the IPCC Second Assessment Report (SAR). The emissions data from FOLU represents land-based CO2 emissions from forest and peat fires and decay that approximate to the net CO2 flux from FOLU as described in Chapter 11 of this report. Average annual GHG emissions growth rates for the four decades are highlighted with the brackets. The average annual growth rate from 1970 to 2000 is 1.3 %. [Figure 1.3]
Note that “GHG” emissions reported as “CO2 equivalent” by the IPCC here also included methane, nitrous oxide, CFCs and CO2 emissions attribute to land use, land use change and forestry (LULUCF), shown above as “CO2 FOLU” (Forestry and Other Land Use). Emissions for the other gases are converted to “CO2 equivalent” according to their Global Warming Potential as estimated by IPCC.
The above totals tie very closely to historical data on emissions published by the EU – their Emissions Database for Global Atmospheric Research (EDGAR) – see here. This dataset includes historical data by country back to 1970, including separate estimates of CO2 from fossil fuel and industrial processes separately and an estimate of GHG emissions (as CO2 equivalent). The most recent data for overall GHG emissions is here; results to 2012 were published in November 2014. Data on CO2 emissions from fossil fuel and industrial processes, the largest component, is more up-to-date, with the most recent results by country up to 2014 here.
In the above graphic, I’ve overplotted the EDGAR v4.2 GHG results (red +- signs). This shows that, up to minor amendments, the IPCC TS.1 figure is showing the same thing as the EDGAR v4.2 GHG series (thus permitting the use of the EDGAR v4.2 data for more detailed analysis).
I’ve also estimated 2013 and 2014 GHG values from the corresponding CO2 (FFIP) data (magenta +-signs). The estimated 2014 value is 54.8 Gt (CO2 eq). In other words, virtually all of the agreed GHG increase to 2025 from the INDCs (to 55.2 Gt COE eq) appears to have already occurred by 2014.
UNEP Pledge Pipeline Dataset
The only recent collation of INDC pledges that I’ve been able to locate is published by UNEP – see here. Its current update (December 22, 2015) says that “160 INDCs (187 countries) are now submitted to UNFCCC”, up “from 119 until 1st October” when the Synthesis Report was issued. It links to a spreadsheet of commitments. Although the historic total for 2010 (48.9 Gt COE eq excluding LULUCF; 49.8 Gt COE eq including) reconciles more or less to EDGAR v 4.2 values, UNEP values for individual countries do not reconcile at all. In addition, there were many missing values, even where EDGAR had 2010 values.
There were some peculiar errors. For example, 2020 emissions from the Congo were projected to equivalent to India’s and only slightly smaller than the EU’s. However, this was due to a conversion error from Mt COE eq to Gt COE eq, in which UNEP incorrectly divided by 100, rather than 1000. Given that this particular error was the same order of magnitude as the projected yield of the INDCs, one naturally wondered whether it had any impact and, if not, how the corrective plug worked.
A plugged line “Not Accounted For” particularly caught my eye – this ought to ring red bells for any data analyst and one would have expected that one of the 40,000+ COP21 would have paid attention to it. The values of this were surprisingly high: over 11 Gt COE eq in 2010 (over 20%) of the total – an amazingly high value given availability of EDGAR v4.2 data. The values were suspiciously variable: why would the values decline by over 50% going from 2020 to 2030. Nor did the formulas for the plugged terms make any sense: I’ve highlighted the formula for 2020 incl LULUCF, which shows a plugged term of 10000 Mt CO2 eq less seven terms for individual countries, none of which made any sense. The corresponding term for 2020 with low reductions was completely different. (The values shown below come from a screenshot of the spreadsheet which shows selected rows and columns).
A couple of days ago, I notified Joergen Fenhann of UNEP of the Congo error, receiving a short thanking acknowledgement. I followed up on December 21 with questions about the plugged values:
The line item for Not accounted for has a plugged value from which some country values are deducted. Where does this plugged value come from? Why are some countries deducted but not others?
Fenhann did not reply to my question on December 21. However, this morning, I noticed that UNEP made major changes to their spreadsheet – without any notice of that a prior error had been corrected and, needless to say, without any acknowledgement. The 2020 MtCO2eq emissions (incl LULUCF) are now 63944 MT versus the former 58524 Mt, an overnight increase of 5.4 Gt.
The plugged values no longer contain the peculiar deductions for seemingly arbitrary countries and are now “explained” as “Data added in order to get the same totals as in UNEP Gap Report 2014”. In a future post, I’ll try to determine the impact of this plug.
It seems really odd to me that the UNFCCC Synthesis Report did not contain supporting information showing emissions by country. UNEP has a spreadsheet purporting to do so, but unfortunately it contains large plugged values, country errors and does not reconcile immediately to published historical data.
Nonetheless, it can be used as a basis for comparing the pathways between countries of interest, which I will do in a forthcoming post for China relative to the U.S. and EU, a post in which I’ll consider some of Naomi Oreskes’ attempt to divert attention away from Chinese coal consumption.