Impact of Uncertainty on Compliance and Emission Trading

This exercise allows an user to get a grip on the uncertainty in the emissions of GHGs and understand its impact at the scale of countries on compliance and the amount of emission permits that can be traded under the Kyoto Protocol.

Background Information

The starting point for this educational exercise is the Kyoto Protocol to the United Nations Framework Convention on Climate Change [UNFCCC]. The exercise centers on a tool which allows assessing changes in the national emissions of greenhouse gases [GHGs] in lieu of their uncertainties.

What is this educational exercise good for? 

The assessment of GHG emissions and removals is high on political and scientific agendas. Policy-makers use GHG inventories to develop strategies and policies to reduce emissions and to track the progress of these measures. However, GHG inventories contain uncertainties with far-reaching scientific, political and economic implications. At present, Parties to the UNFCCC are encouraged, but not obliged, to report uncertainties associated with GHG emission estimates. Inventory uncertainty is monitored, but not regulated, under the Kyoto Protocol.

Given this setting, the purpose of this exercise is very specific: It shall help a user get a grip on uncertainty and understand its impact on I) compliance and II) the amount of emission permits that can be traded under the Kyoto Protocol.

How does the tool work? 

The tool allows a user to analyze by means of two techniques emission changes in lieu of their uncertainty. Spatially, the changes in emissions refer to the country scale; and temporally to the changes I) that countries have agreed to meet by 2008/12, the first commitment period of the Kyoto Protocol [Track I: compliance mode]; and II) that countries report annually with reference to 1990, the base year under the Kyoto Protocol [Track II: monitoring mode].

Which emissions-change-versus-uncertainty analysis techniques are used? 

Irrespective of the track (compliance or monitoring mode) followed, two analysis techniques are used. The two techniques differ but they follow the same idea: Inventoried emissions of GHGs are uncertain, and this uncertainty translates into a risk that true emissions are greater than those estimated and reported. To compensate for, or even reduce, this risk a safety margin (or undershooting) is considered. For a given risk, the safety margin differs depending on the selected analysis technique. The first technique follows the undershooting [Und] concept which accounts for uncertainty at two points in time; while the second follows the combined undershooting and verification time [Und&VT] concept which accounts for uncertainty at one point in time. The major characteristics of the two techniques are listed below and their mathematical background can be found here.

Taken into account by the techniqueUndUnd&VT
Trend uncertainty 
Total uncertainty 
Emissions difference between t1 and t2 
Emissions gradient between t1 and t2 
Risk with reference to the concept of significance 
Risk with reference to the concept of detectability 

Under which conditions is the tool used? 

The most important assumption for using this tool is that no gap exists in accounting GHG emissions bottom-up and top-down. This condition follows the concept of dual-constrained full GHG accounting. Of the many sources of GHG emissions, emissions of CO2 from fossil fuel burning are the largest and most important in terms of impact on the climate, and they have the lowest quantitative uncertainty. How certain, then, are our best numbers of GHG emissions? Information on the uncertainty of CO2 emissions estimates can be found here. Other conditions under which the tool is used are listed below.

With reference toCondition of use
Countries/groups of countriesas listed in Annex B to the Kyoto Protocol, including the ‘old’ EU Member States collectively (EU-15)
GHGsas listed in Annex A to the Kyoto Protocol, but only collectively (in CO2-eq) and not individually
Not taken into account1) Land use, land-use change, and forestry [LULUCF]
2) the so-called Kyoto mechanisms
Uncertaintiesrefer to an equal-sided confidence interval of 95% (following the concept of a normal distribution) and are classified in relative terms according to the following intervals (in %): 0 , 5[; [5 , 10[; [10 , 20[; [20 , 40[;  and >40%
To facilitate calculations, only the medians of these intervals are used (exception: Class 5 is represented by 40%, its lower boundary).
Data fromEuropean Environment Agency (EEA)
Note: The EEA produces Community-wide GHG inventories on a rolling basis and releases these officially with a time lag of two and more years (see also here). This explains a time lag of about three years in the data that we use in our emissions-change-versus-uncertainty analyses.

Print this page

Last edited: 14 December 2017


Matthias Jonas

Senior Research Scholar

Exploratory Modeling Of Human-Natural Systems Research Group

T +43(0) 2236 807 430

International Institute for Applied Systems Analysis (IIASA)
Schlossplatz 1, A-2361 Laxenburg, Austria
Phone: (+43 2236) 807 0 Fax:(+43 2236) 71 313