Global Climate Change: Homework Set 4

Massachusetts Institute of Technology Climate Change: Economics, Science and Policy Problem Set #4 GLOBAL POLICY ANALYSIS USING THE TOY IGSM BACKGROUND The ToyIGSM is a very powerful tool. Its economics component is a full version of the research model used in the Joint Program on the Science and Policy of Global Change. The Excel interface allows you to manipulate this model without needing to edit the underlying GAMS code. The Excel program also contains an approximation of the Earth Systems component of the IGSM, generated using regression analysis on 1250 runs of the full model. This homework is divided into two sections: Problem 1 investigates the economic impacts of near term policies: McCain-Lieberman in the US, and the European Trading System (which encompasses the EU and EET in the EPPA definitions). Problem 2 requires an uncertainty analysis of the climate impacts of the business as usual case. PROBLEM 1. THE ETS AND MCCAIN-LIEBERMAN Part A: Independent EU and US Systems. In order to estimate targets for the McCainLieberman bill and the ETS, the following sources were used: “Assessment of US Cap and Trade Proposals” (Paltsev et al., 2007, from your reading), and “An Analysis of the European Trading Scheme” (Reilly and Paltsev, 2005, Joint Program Report #127). The following paragraphs provide guidance for setting up the analysis. Model Procedures. Before implementing constraints in the model, use the Clear Display button on the GHG_Constraints sheet to ensure that all regions outside the US and the EU have no constraints. Also, for the purposes of this problem select the 2020 radio button on the EPPA 4.0 worksheet. This will make the model stop in 2020, keeping runtime to about 3 minutes for simulation runs. The Save Run functionality can be used to save data. However, if you then reopen the run using the “Open Saved Run” functionality, the data from the old run will be loaded into both Ref_Results and EPPA_Results worksheets. This means that any graphs in EPPA_Graphs that compare policy to reference (such as the welfare graph) will not give the right data. You can manually cut and paste the Ref_Results page from a reference run which should correct all the graphs again. Or you can manually copy and paste the graphs and data you want into another worksheet before running the model again. If you do this, don’t forget to delink the graphs (use “paste special – values” for the data), or they will update when the Toy updates. p. 2 Approximating the McCain-Lieberman Bill: The McCain-Lieberman bill requires emissions to return to 2004 levels by 2012, 1990 levels by 2020, and approximately one third of 2000 levels by 2050. These provisions can be approximated by returning to 2005 levels in 2010, 1997 levels in 2015, and 1990 levels in 2020. While the bill is designed to cover all gases, the first part of the exercise will only address CO2. Therefore, we can set the constraints on CO2 as follows: Year US Target (MMT C eq) 2010 1690 2015 1480 2020 1350 Fill these values in on the GHG_Constraints worksheet, and make sure that the CO2 radio button is active in the upper left hand corner. In addition to not addressing non-CO2 GHGs, the approximation also ignores any restriction on covered versus non-covered sources. Nor does it allow for banking between time periods. Approximating the ETS: The initial ETS target is about 1104 MMT carbon equivalents. This policy covers both the EU and EET. In this first period (2010), the allocation to EU nations is about 896 MMT and the allocation to EET nations is 208 MMT. Recent EU proposals suggest that the target by 2020 will be 20% below 1990 levels. We approximate this by the following constraints: Year 2010 2015 2020 EU Target (MMT Ceq) 896 836 776 EET Target (MMT Ceq) 208 208 208 Fill these values in on the GHG_Constraints worksheet. The EU and the EET form one trading system, therefore activate trading in both regions for all time periods. (The ETS covers only CO2 in 2010, but will be extended to other GHGs in later time periods. It actually covers a rather limited number of sectors – about 54% of the total CO2 emissions. The Toy Model does not allow partial sectoral coverage.) Running the Model: Now that constraints have been set, and the EU and the EET having trading enabled (but not the US!), you can run the model. Click the Run EPPA! button. A command window will pop up. The model will take about 3 minutes to run, during which time Excel will be inoperable. Part A: Independent Systems, CO2 Only Policy A1: Report the carbon prices in the US, EU, and EET (listed in the EPPA_Results worksheet). Comment briefly on these prices. A2: Now consider the GDP impacts (the GDP/reference graph in EPPA_Graphs). Which nations (considering both Annex B and Non-Annex B) have the largest p. 3 losses measured by percent GDP, and which nations gain? Briefly explain these results.. A3: If the US decided not to implement a climate policy, how would the results in Part A2 change? Part B: Full Annex B Trading, CO2 Only Policy In Paltsev et al. (2007) it is noted that “For example, linkage with emissions trading with other countries is authorized or anticipated in some of the bills, but may depend on a later judgment by the Administration that the foreign system is sound.” Now we can examine what happens if the US decides to allow trade between McCain Lieberman and the ETS system. Activate trading in the US and run the model again. B1: What are the new carbon prices in the three regions? How did they change from Part A1 and why? Explain the benefits of “where” flexibility. B2: In the EPPA_Analysis Worksheet, the Toy reports how many carbon permits are bought and sold. How much does the EU buy (or sell) in 2010? How does this number compare to the original (pre-trade) reduction of carbon emissions required? B3: The EPPA_Analysis Worksheet also lists GDP/capita (in tens of thousands of US$). What is the cost per person in the US of the policy compared to BAU GDP/capita in 2010, 2015, and 2020? If you take the 2010 cost and assume a 5% increase per year, what would the cost be in 2020? How does this compare to the actual cost? What does this imply in terms of banking? Extra Credit: If you compare the GDP/capita in the trading case to the non-trading case, you will see a non-intuitive result. Explain how this could happen. Reading Babiker et al, The Energy Journal, 25(2), 33-56, 2004 may help with this explanation. Part C: All-Gas Policy. The US policies include all gases. The EU ETS may include all gases in later periods. Therefore it is instructive to examine a case in which all greenhouse gases are constrained. We will use the same total carbon-equivalent reductions, but now covering all gases. Keep trading enabled for all three regions. Make sure the GHG radio button is active in the upper left hand corner of the GHG_Constraints worksheet. Click the Reset Display to Reference button (unfortunately, testing the current Toy seems to indicate that leaving the column blank sets a constraint equal to the CO2 reference). Using “Reference” constraints is an approximation to an unconstrained case, but it isn’t perfect. In the appropriate cells, overwrite the current numbers with the numbers in the following table for the constraints: p. 4 Year US Target (MMT Ceq) EU Target (MMT Ceq) EET Target (MMT Ceq) 2010 2080 1166 281 2015 1895 1124 288 2020 1801 1086 293 C1: What is the new carbon price under the all gas policy? Explain the benefits of “what” flexibility. C2: What is the GDP/capita in the US in 2010 compared to the old policy with trading? C3: Do these results suggest that it is better to implement an all gas policy or a CO2 only policy, and if so why is this the case? (Possible drawbacks of including all gases in one policy will be covered in the April 25th lecture.) QUESTION #2. ECONOMIC AND CLIMATIC UNCERTAINTY Prepare estimates of the uncertain temperature change faced over the 21st Century under no-policy reference conditions. Uncertainty in emissions can be represented by three combinations of the underlying determinants: RR Reference economic growth and energy efficiency in all regions HL High economic growth and Low energy efficiency in all regions LH Low economic growth and High energy efficiency in all regions This analysis will require three emissions projections and analysis of their impacts on temperature, with each run based on different values of the uncertain economic parameters. A range of values for the three key uncertain climate parameters in analysis of the climate response–ocean diffusivity (Kv), climate sensitivity (S) and aerosol forcing (Faer)—are shown in the following table. Kv 1 4 7 S 2 4 Faer 0.5 1 These parameter combinations are not thought to occur with equal probability. Using data and expert judgment, the following prior probability distribution over the three parameters has been developed. They may be treated as statistically independent. p. 5 Kv Prob. 1 0.04 4 0.70 7 0.26 S Prob. 2 0.29 4 0.71 Faer Prob. 0.5 0.31 1 0.69 Part 2A: Temperature Change under Reference Emissions. For the RR emissions projection, construct a probability density function (PDF) of global mean temperature change in the year 2100. Part 2B: Full Uncertainty of Temperature Change. Prepare a PDF of uncertain temperature change to 2100 on the assumption that the probabilities of the emissions projections are RR=.5, HL=.25 and LH=.25. Show your analysis and explain briefly. For both 2A and 2B please submit copies of any worksheets (yours, not those in the ToyIGSM solution) used to perform these calculations and include a brief description of how you put the analysis together. Also turn in the PDF graph. Hints: (1) With advance planning, this problem requires only three runs of the model, one for each emissions projection. (Running time to 2100 under business as usual is about 15 minutes) The attached event tree can be used as a guide when systematically varying the climate parameter. (Check that all of the probabilities sum to one!) (2) For the PDF use a temperature range of 0°C to 7°C with 0.5°C increments. (For example, the 2°C bin will contain values from 1.75°C to 2.25°C.) The probabilities on the y-axis will be the aggregated probability that the occurrence of warming falls into each 0.5°C bin. (3) There should be no constraints on any of the regions. Be sure to clear the reference values (quota constraints) from the spreadsheet. (Should be a handy button to do this.) p. 6 p(Faer = 0.5) 0.31 p(S = 2) 0.29 p(Faer = 1) p(Kv = 1) 0.69 0.04 p(Faer = 0.5) 0.31 p(S = 4) 0.71 p(Faer = 1) 0.69 p(Faer = 0.5) 0.31 p(S = 2) 0.29 p(Kv = 4) 0.7 p(Faer = 1) 0.69 p(Faer = 0.5) 0.31 p(S = 4) 0.71 p(Faer = 1) 0.69 p(Faer = 0.5) 0.31 p(S = 2) 0.29 p(Faer = 1) p(Kv = 7) 0.69 0.26 p(Faer = 0.5) 0.31 p(S = 4) 0.71 p(Faer = 1) 0.69