Uncertainty Analysis: Technique 1 & 2

Technique 1

Inventoried emissions of GHGs are uncertain, and this uncertainty translates into a risk that true emissions are greater than those estimated and reported. Undershooting helps to limit, or even reduce, this risk. Techniques are available to analyze uncertain emission changes (emission signals) from various points of view ranging from signal quality (defined adjustments, statistical significance, detectability, etc.) to the way uncertainty is addressed (trend uncertainty or total uncertainty).

Figure 1 illustrates the undershooting [Und] concept which accounts for uncertainty at, and correlates uncertainty between, two points in time. The Und concept thus follows the footsteps of statistical significance in quantifying the aforementioned risk. It allows factoring in a change in uncertainty, which can be due to learning and/or result from structural changes in the emitters. However, so far it is assumed that our knowledge of uncertainty stays constant over time in relative terms (first-order approach). This is because researchers are only beginning to diagnose ‘observed’ changes in the uncertainty of CO2 emissions from fossil fuel burning and separate their causes.

Fig. 1:Technique 1: Undershooting [Und] concept. Undershooting helps to limit, or even reduce, the risk that true emissions are greater than those estimated and reported (illustrated with the help of normal probability density functions). The Und concept accounts for uncertainty at two points in time, here with the second time point referring to 2010 when a country has to comply with its Kyoto (emissions) target [KT]. To correlate uncertainties, a factor of 0.75 is used which is typical for currently reported emission uncertainties. Sources: Jonas et al. (2007: Fig. 3); Jonas et al. (2008: App. C).


Technique 2

Inventoried emissions of GHGs are uncertain, and this uncertainty translates into a risk that true emissions are greater than those estimated and reported. Undershooting helps to limit, or even reduce, this risk. Techniques are available to analyze uncertain emission changes (emission signals) from various points of view ranging from signal quality (defined adjustments, statistical significance, detectability, etc.) to the way uncertainty is addressed (trend uncertainty or total uncertainty).

Figure 1 illustrates the combined undershooting and verification time [Und&VT] concept which accounts for uncertainty at a specified point in time. The Und&VT concept thus follows the footsteps of signal detection in quantifying the aforementioned risk. It allows factoring in a change in uncertainty, which can be due to learning and/or result from structural changes in the emitters. This would be important if the selected point in time is in the (near-term) future. However, so far it is assumed that our knowledge of uncertainty in the future will be as good as our today’s knowledge in relative terms (first-order approach). This is because researchers are only beginning to diagnose ‘observed’ changes in the uncertainty of CO2 emissions from fossil fuel burning and separate their causes.

Fig. 1:Technique 1: Combined undershooting and verification time [Und&VT] concepts. Undershooting helps to limit, or even reduce, the risk that true emissions are greater than those estimated and reported, but also to ensure that an emission signal becomes ‘detectable’. The Und&VT concept accounts for uncertainty at a specified point in time, here referring to 2010 when a country has to comply with its Kyoto (emissions) target [KT]. Depending on how detectability (here expressed via δcrit, the critical emission limitation or reduction) and KT relate to each other, four cases can be distinguished (two of which are shown in the figure). Case 1 reflects detectability while Case 2 does not. In contrast to case 1, case 2 requires an initial obligatory undershooting which is introduced to ensure that detectability of emission reductions, not increases, is given before Annex B countries are permitted to make economic use of potential excess emission reductions. Sources: Jonas et al. (2007: Fig. 4); Jonas et al. (2008: App. D).





Print this page

Last edited: 22 July 2013

CONTACT DETAILS

Matthias Jonas

Senior Research Scholar

Advanced Systems Analysis

T +43(0) 2236 807 430

Impacts of Uncertainty

Go back to the main page for

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