Prior to an Economic Treatment of Emissions and Their Uncertainties under the Kyoto Protocol: Scientific Uncertainties that Must be Kept in Mind

Transcription

1 Prior to an Economic Treatment of Emissions and Their Uncertainties under the Kyoto Protocol: Scientific Uncertainties that Must be Kept in Mind M. Jonas and S. Nilsson (IIASA) International Workshop on: Uncertainty in Greenhouse Gas Inventories: Verification, Compliance and Trading Warsaw, Poland September Sept

2 Contents: 1. Motivation 2. Working within a FGA Uncertainty Verification Framework Step 0: Setting the Stage Step 1: Bottom-up vs Top-down: Verification of Emissions Step 2: Bottom-up/Top-down vs SD: Verification of Emission Changes Step 3: Effectiveness vis-à-vis Compliance: Statistical Significance vs Detectability Step 4: Effectiveness vis-à-vis Credibility: Uncertainty in the Accounting Matters Step 5: SD: Different Techniques Different Findings 3. Summary and Conclusion 24 Sept

3 Where this presentation will take you: M Uncertainty KP Economic Studies FCA FGA KP Verification Economic World Physical Scientific World 24 Sept

4 M Economic studies placed into a physical scientific framework so what? Motivation: Two questions matter: 1. How credible are the emissions that I am paying for? 2. How will the emission price develop in the future? At the end of our presentation, you will know the answer to Question 1! 24 Sept

5 Contents: 1. Motivation 2. Working within a FGA Uncertainty Verification Framework Step 0: Setting the Stage Step 1: Bottom-up vs Top-down: Verification of Emissions Step 2: Bottom-up/Top-down vs SD: Verification of Emission Changes Step 3: Effectiveness vis-à-vis Compliance: Statistical Significance vs Detectability Step 4: Effectiveness vis-à-vis Credibility: Uncertainty in the Accounting Matters Step 5: SD: Different Techniques Different Findings 3. Summary and Conclusion 24 Sept

6 Scientific quality attached to Moss & Schneider s uncertainty concept: S0 Plausibility Consensus (Soft Knowledge) Observations (+ Accounting) Verification Theory Validation Model Results: Diagnostic Modeling Prognostic Modeling Validation Plausibility 24 Sept

7 Accounting vs Diagnostic and Prognostic Modeling: S0 Net Emissions U Account U Model Accounting Prognostic Model Mode Diagnostic Model Mode Past Base Year Future Commitment Year Time 24 Sept

8 The Applied Uncertainty Concept: S0 Uncertainty Range Quantified Systematic Error Or Measured Bias Random Errors True Systematic Error Or Unknown Bias No Knowledge Greater Accuracy Smaller Precision Accepted Value True Value Smaller Accuracy Greater Precision 24 Sept

9 Our ACDb Experience: Uncertainty Classes S0 24 Sept

10 Contents: 1. Motivation 2. Working within a FGA Uncertainty Verification Framework Step 0: Setting the Stage Step 1: Bottom-up vs Top-down: Verification of Emissions Step 2: Bottom-up/Top-down vs SD: Verification of Emission Changes Step 3: Effectiveness vis-à-vis Compliance: Statistical Significance vs Detectability Step 4: Effectiveness vis-à-vis Credibility: Uncertainty in the Accounting Matters Step 5: SD: Different Techniques Different Findings 3. Summary and Conclusion 24 Sept

11 PGA under the KP: S1 Net Storage in the Atmosphere Sphere of activity under the Kyoto Protocol FF Industry Kyoto Biosphere Non-Kyoto? Biosphere Country X 24 Sept

12 Can the Kyoto Protocol be verified? S1 Verification requires (following science theory!) bottom-up/top-down FGA 1. Atmospheric view 2. Completeness resolving countries! The basket approach under the KP also includes gases that have both anthropogenic and natural sources (sinks). However, these are most crucial because Lesson 1: the KP cannot be verified if the biosphere is split up into a Kyoto biosphere and a non-kyoto biosphere! This is because an atmospheric measurement that can meet this discrimination requirement is not available. 24 Sept

13 Contents: 1. Motivation 2. Working within a FGA Uncertainty Verification Framework Step 0: Setting the Stage Step 1: Bottom-up vs Top-down: Verification of Emissions Step 2: Bottom-up/Top-down vs SD: Verification of Emission Changes Step 3: Effectiveness vis-à-vis Compliance: Statistical Significance vs Detectability Step 4: Effectiveness vis-à-vis Credibility: Uncertainty in the Accounting Matters Step 5: SD: Different Techniques Different Findings 3. Summary and Conclusion 24 Sept

14 S2 But the KP focuses on emission changes: The KP requires that net emission changes (emission signals) of specified GHG sources and sinks, including those of the Kyoto biosphere but excluding those of the non-kyoto biosphere, be verified on the spatial scale of countries by the time of commitment, relative to a specified base year. The relevant question then is whether these emission signals outstrip uncertainty and can be verified (correctly: detected). 24 Sept

15 But the KP focuses on emission changes: S2 Atmosphere F net 2ε ε = const Time t 1 t 2 24 Sept

16 Total and Trend Unc Concepts of the IPCC: S2 Net Emissions Total Trend RC Time for Achieving Reduction Commitment KT t 1 t 2 Time 24 Sept

17 It is the total uncertainty in the commitment year/period that matters if we ever want to place SD meaningfully in a bottom-up/topdown verification context. S2 F net 2ε ε = const Time t 1 t 2 Lesson 2: The temporal detection of emission changes cannot be placed meaningfully in a bottom-up/top-down verification context if SD does not acknowledge total uncertainty. 24 Sept

18 Contents: 1. Motivation 2. Working within a FGA Uncertainty Verification Framework Step 0: Setting the Stage Step 1: Bottom-up vs Top-down: Verification of Emissions Step 2: Bottom-up/Top-down vs SD: Verification of Emission Changes Step 3: Effectiveness vis-à-vis Compliance: Statistical Significance vs Detectability Step 4: Effectiveness vis-à-vis Credibility: Uncertainty in the Accounting Matters Step 5: SD: Different Techniques Different Findings 3. Summary and Conclusion 24 Sept

19 Emission Signal: Statistical Significance vs Detectabilty S3 x ε 1 ε 2 x a) VT > t 2 ε 1 x t 1 t 2 Time b) VT = t 2 ε 2 ε 1 t 1 VT Time ε 2 c) VT < t 2 Statistical significance does not imply detectability! VT t 1 t 2 Time 24 Sept

20 X ε 1 x 1 -x 2 = δ KP x 1 Emission Signal: How to grasp its detectabilty S3 ε 2 t 1 t 2 X Time ε x,2 t 1 t 2 x 1 -x 2 = δ KP x 1 Time 24 Sept

21 Emission Signal: Statistical Significance vs Detectabilty S3 Lesson 3: The knowledge of total uncertainty at only two points in time without considering the dynamics of the emission signal permits investigating its statistical significance but not its detectability. 24 Sept

22 Contents: 1. Motivation 2. Working within a FGA Uncertainty Verification Framework Step 0: Setting the Stage Step 1: Bottom-up vs Top-down: Verification of Emissions Step 2: Bottom-up/Top-down vs SD: Verification of Emission Changes Step 3: Effectiveness vis-à-vis Compliance: Statistical Significance vs Detectability Step 4: Effectiveness vis-à-vis Credibility: Uncertainty in the Accounting Matters Step 5: SD: Different Techniques Different Findings 3. Summary and Conclusion 24 Sept

23 Uncertainty in the accounting matters: a) PCA(FF) b) PCA(FF+LUCF) S4 F net F net FF Signal FF Signal FF+LUCF Signal ε FF ε FF+LUCF VT Time VT Time c) PCA(FF) d) PCA(FF+LUCF) F net F net FF+LUCF Signal FF Signal ε FF ε FF+LUCF VT Time VT Time 24 Sept

24 Uncertainty in the accounting matters: S4 Net GHG Emissions Net GHG Emissions Base Year Commitment Year/Period Time Base Year Commitment Year/Period Time 24 Sept

25 S4 Uncertainty in the accounting matters: Lesson 4: Without uncertainty, an effective (credible) emission trading system cannot be established. 24 Sept

26 Contents: 1. Motivation 2. Working within a FGA Uncertainty Verification Framework Step 0: Setting the Stage Step 1: Bottom-up vs Top-down: Verification of Emissions Step 2: Bottom-up/Top-down vs SD: Verification of Emission Changes Step 3: Effectiveness vis-à-vis Compliance: Statistical Significance vs Detectability Step 4: Effectiveness vis-à-vis Credibility: Uncertainty in the Accounting Matters Step 5: SD: Different Techniques Different Findings 3. Summary and Conclusion 24 Sept

27 Signal Detection Basics: S5 We distinguish between Preparatory SD addresses the question: How well do we need to know net emissions if we want to detect a specified emission signal after a given time? Midway SD SD in Retrospect No what-if type of prognostics involved! Preparatory SD is also an excellent monitoring tool! 24 Sept

28 SD Und Concept: F net Assuming that S5 Base Year Level Reduction Level Risk Undershooting x t,1 -x 1 ε 1, x t,2 -x 2 ε 2 and ρ 1 = ρ 2, we find with risk α: x t,2 (1-δ KP ) x t,1 Time x x t 1 t 2 ( 1 δ ) 1 ( 1 2α ) ( 1 2α ) ρ ρ δ mod Und { δ + 2( 1 2α )( δ ) ρ} 2 KP 1 KP KP 24 Sept

29 Und Concept: Accurate Results/ Emission Reduction S5 The Und Concept runs counter to the spirit of the KP! Annex I countries committed to emission limitation: NZ, RU, UA; NO; AU; IS. 24 Sept

30 SD Und&VT Concept: Assuming that x t,1 -x 1 ε 1, x t,2 -x 2 ε 2, ρ 1 = ρ 2 and demanding detectability, we find with risk α: S5 x t,2 (1-δ crit ) x t,1 2 ( ) Und x1 1 ( 1 δcrit ) 1 δkp + UGap + 1 2α 1 δcrit ρ x α ρ { ( )( ) } δ mod 24 Sept

31 Und&VT Concept: Accurate Results/ Emission Reduction S5 The Und&VT Concept runs counter to the Kyoto policy process! Annex I countries committed to emission limitation: NZ, RU, UA; NO; AU; IS. 24 Sept

32 SD: Different Techniques Different Findings: Lesson 5: Signal detection techniques differ; each has its pros and cons. A discussion on which technique to select has not even started! Economists must be aware that the risk of compliance, i.e., that the countries true emissions in the commitment year/period are above their true Kyoto targets can be grasped (although the countries true net emissions are unknown) and thus be priced. We believe that not evaluating the countries emission signals in terms of risk and detectability will miss economic reality. S4 24 Sept

33 Summary We have 1) step-by-step specified the relevant conditions for carrying out temporal signal detection under the Kyoto Protocol and identified a number of scientific uncertainties that economic experts must keep in mind; and 2) answered the crucial question of how credible are tradable emission permits. Conclusion Our specific intention was to provide a basis for economic experts to carry out useful emission trading assessments and to specify the validity of their assessments from a physical scientific point of view. Our general intention, however, was that we see a clear need for an intense interaction (both ways) between physical scientists and economic experts if we ever want the KP to become successful. We must begin to talk to each other! 24 Sept