Final Report DE-FG03-94ER61944

Transcription

1 Final Report DE-FG03-94ER61944 Title: "Learning and Enhanced Climate Representation in integrated Assessment Models" PI: Prof. Charles D. Kolstad, Department of Economics, UCSB Start Date: 9/94 End Date: 5/97 0 bject ive: The objective of the project is to enhance capabilities for integrated-assessment modeling in two major areas: learning/r&d/information acquisition and the nexus between climate dynamics and climate impacts. In the first of these areas, our objective is to improve the way in which economic models deal with learning (endogenous and/or exogenous) within an economy. This would obviously include the R&D process, whereby knowledge about climate change (and many other things) is acquired over time and influences regulatory actions. The work in climate dynamics is focused in part on incorporating the regional climate-change results from equilibrium and transient general circulation model (GCM) simulations into our simplified integrated-assessment model. While the work is generic and therefore applicable to any integrated-assessment model, it is done in the context of a standard Ramsey growth model. Thus the work involves theoretical conceptualization, empirical implementation in an integratedassessment model, and analysis using that model. Approach: This project is divided into four separate tasks. (1) Expand Characterization of Learning This task is one of the most fundamental parts of this project. The basic problem is how to incorporate learning (information acquisition) into an optimal-growth model. Previous work of the PI has focused on exogenous learning -the star-shaped spreading of beliefs. This task involves extending this approach and developing several additional representations of learning, particularly Bayesian learning about climate sensitivity. (2) Enhanced Climate Model. This task involves improving the representation of climate within simple optimal growth models. Two criteria are particularly important: tractability and realism. To deal with C02 accumulation, we have sought to reduce the dimensionality of C02

2 DISCLAIMER This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employe#, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recornmendation. or favoring by the United States Government or any agency thereof. The views and opinions of authors expnssed herein do not neccssariiy state or reflect those of the United States Government or any agency thereof.

3 DISCLAIMER Portions of this document may be illegible electronic image products. images are produced from the best available original document.

4 accumulation equations used in typical IA models. We have also sought to use historic data to calibrate the climate relationships within our integrated assessment model. (3) Refined Pollution Control and Damage. One of the major avenues of damage in integrated assessment may arise from an inability to adapt to climate change due to a imperfect knowledge that a climate change has occurred. For instance, a farmer may fail to realize the climate has changed, continue with old practices and in the process incur significantly more damage than if the farmer had known the change had occurred. This task will seek to better characterize damage associated with delayed learning. (4) Analysis. In order to tie the three preceding tasks together, one or two analyses will be conducted of key climate change issues. The two issues we are currently planning to consider are: (a) the effect of learning on reaching an agreement on climate control; and (b) the effect of learning with the possibility of a climate catastrophe. Other analyses may also be conducted, including the effect of uncertainty on climate sensitivity on abatement levels. Results to Date: Our most significant result has been to incorporate a new representation of learning into integrated assessment models. We have successfully incorporated Bayesian learning about climate sensitivity into a DICE-like integrated assessment model. This has involved representing the standard optimization-type integrated assessmenvoptima1 growth model as a stochastic dynamic program. This allows the state of knowledge about climate sensitivity to be represented as a state variable which evolves over time according to Bayes rule and learning occurs regarding climate sensitivity. The empirical implications of our examination of climate sensitivity is that it may take years to resolve uncertainty, based purely on the stochastic temperature record. Learning about climate damage (which we are currently examining) appears to take an even longer period of time. Another result of our work has been the development of a new algorithm for solving dynamic programs of the complexity of an integrated assessment model. The method involves using neural networks to approximate the value function in the dynamic program. We have also worked to improve the efficiency of this method. An ancillary result of our work is that we have successfully demonstrated the use of dynamic programming in solving integrated assessment models. This step forward is very important to be able to look at uncertainty and many types of learning. Another result is that we have been able to deduce the underlying distribution on climate shocks, using a GCM and the instrumental temperature record. This work is important for our integrated assessment activities but it is also important in and of itself. We have backed sulfate aerosols and tropospheric ozone out of the temperature record, leaving an induced distribution on climate sensitivity to greenhouse gas forcing. 2

5 Deliverables: Papers prepared: 1. M.E. Schlesinger, N. Ramankutty, C Kolstad, "Reduced-Form Climate Models for Economic Studies," University of Illinois Working Paper, October, M.E. Schlesinger and N. Andronova "Observationally Determined Climate Sensitivity and Its Probability Distribution," University of Illinois Working Paper, June, C. Kolstad, "Learning, Irreversibilities, and Climate," UCSB Working Paper in Economics #15-96(1996) 4. D. Kelly and C. Kolstad, "The Climate Change Footprint: Will We See It Before It Is Upon Us?," forthcoming, J. Economic Dynamics and Control. 5. D. Kelly and C. Kolstad, "Malthus and Climate Change: Betting on a Stable Population," UCSB Working Paper in Economics #9-96(1 996)[submitted to American Economic Review]. 6. C. Kolstad, "Deja Vu All Over Again: What's New and What's Not in Integrated Assessment Modelling of Climate Change," UCSB working paper (1 996). 7. D. Kelly and S. Spear, "Endogenous Strategic Business Cycles," forthcoming in Journal of Economic Theory. 8. D. Kelly and J. Shurish, "Stability of Functional Rational Expectations Equilibria," working paper (1 996),forthcoming, J. Economic Dynamics and Control. 9. C. Kolstad, "Uncertainty, Learning, Stock Externalities and Capital Irreversibilities," forthcoming, Proc., NATO Workshop, Wageningen, Netherlands. 10. C. Kolstad, "Learning and Stock Effects in Environmental Regulation: The Case of Greenhouse Gas Emissions," J. nv. con. and Mgmt., 31:l-18(1996). 11. D. Kelly, C. Kolstad and G. Mitchell, "Adjustment Costs from Enviornmental Change Induced by Incomplete Information and Learning," UCSB working paper (1 997). 12. D. Kelly and C. Kolstad, "The Climat Change Footprint: Will We See It Before It is Upon Us?, in N. Nakicenovic, W. Nordhaus, R. Richels and F. Toth (Eds), 3

6 Climate Change: Integrating Science, Economics and Policy, IIASA WP , Laxenburg, Austria (1996). 13. D. Kelly, On Kuznets Curves Arising From Stock Externalities, UCSB Working Paper. 14. D. Kelly and D. Steigerwald, An Economic Model of Conditional Heteroskedasticity, UCSB Working Paper. 15. D. Kelly, Unit Roots in the Climate: Is the Recent Warming Due to Persistent Shocks? UCSB working paper, under review at JEEM. Presentations: 1. D. Kelly, UCSB School of Environmental Science and Management, April, C. Kolstad, Workshop on Climate Change: Integrating Science, Economics and Policy, March 1996, IIASA, Austria. 3. C. Kolstad, EMF 14 Meeting #4: Integrated Assessment of Global Climate Change, March 21-23, 1996, IIASA Austria. 4. C. Kolstad, Keynote Speaker, Seventh Annual Conference of the European Association of Environmental and Resource Economists, June, 1996, Lisbon. 5. C. Kolstad, Timing the Abatement of Greenhouse Gas Emissions, CIRED, Paris, June, D. Kelly, Third Occasional California Workshop on Environmental and Resource Economics, Santa Barbara, May, C. Kolstad, Department of Economics, University of Michigan, Ann Arbor, March C. Kolstad, Snowmass Workshop on Climate Change, July, C. Kolstad, NBER Conference on Economics and Policy Issues in Global Warming: An Assessment of the Intergovernmental Panel Report, Snowmass, CO, July 23-24, C. Kolstad, International Association of Energy Economics, Delhi, Malthus and 4

7 Climate Change: Betting on a Stable Population, January, C. Kolstad, Department of Agricultural Economics, University of California, Berkeley, Adjustment Costs from Enviornmental Change Induced by Incomplete Information and Learning, May C. Kolstad, NOAA Workshop on Climate, Duke University, Adjustment Costs from Enviornmental Change Induced by Incomplete Information and Learning, May D. Kelly, Electric Power Research Institute, Bayesian Learning, Pollution-and Growth, Palo Alto, CA (May 1997). D. Kelly, Department of Economics, University of Wyoming, Bayesian Learning, Pollution and Growth, February Collaborations: Prof. Michael E. Schlesinger, Dept. of Atmospheric Sciences, University of Illinois, Urbana, Illinois Postdocs Supported: Dr. David Kelly (formerly with Carnegie-Mellon) Graduate Students Supported: Aran Ratcliff e, Edward Balsdon, Youngbae Moon, Glenn Mitchell 5