Modeling climate change and its impacts: Law, policy, and science

Texas Law Review

Volumn 86, Issue 7, 2008, Pages 1655-1699

  Farber, Daniel A ^a  

^a UNIVERSITY OF CALIFORNIA   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 50049112706     PISSN: 00404411     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Conference Paper

References (268)

1. For example, one section of the Clean Air Act requires the use of "photochemical grid modeling" unless EPA certifies some other technique to be "at least as effective." 42 U.S.C. §7511a(c)(2)(A)(2000).
2. Daniel C. Esty, Environmental Protection in the Information Age, 79 N.Y.U. L. REV. 115 164 (2004).
3. Id.
4. Id.
5. Modeling defects are the subject of ORRIN H. PILKEY & LINDA PILKEY-JARVIS, USELESS ARITHMETIC: WHY ENVIRONMENTAL SCIENTISTS CAN'T PREDICT THE FUTURE (2007). Although their skepticism of modeling is somewhat extreme, they do effectively point out the risks of relying on models as infallible forecasts.
6. See id. at 182 (describing the failures of mathematical models to "predict the outcome of natural processes on the earth's surface"). For another perspective on the use of modeling in environmental law, see generally James D. Fine & Dave Owen, Technocracy and Democracy: Conflicts Between Models and Participation in Environmental Law and Planning, 56 HASTINGS L.J. 901 (2005). According to Fine and Owen: Models can process reams of data and represent mathematically complex chemical, physical and social relationships, allowing modelers to make predictions and test assumptions in ways that otherwise would not be possible. Not surprisingly, models have become essential and ubiquitous planning tools, our dependence upon them making their abandonment all but unthinkable. Nevertheless, models are incomplete representations of reality and suffer from many sources of uncertainty. Every model, strictly speaking, is an approximation, for no model can be an exact representation of the real world. Air quality planning models, like any physically based simulation, also are unverifiable, limiting modelers' ability to assess their reliability. Approximations, subjective choices, and errors in both design and application are common in any modeling effort, but the complexity of current state-of-the-science models hinders assessment of their certainty. Id. at 904-05 (footnotes omitted).
7. For an overview of climate modeling, see generally David A. Randall et al., Climate Models and Their Evaluation, in INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE, CLIMATE CHANGE 2007: THE PHYSICAL SCIENCE BASIS, CONTRIBUTION OF WORKING GROUP I TO THE FOURTH ASSESSMENT REPORT OF THE INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE 589, 589-662 (S. Solomon et al. eds., 2007), available at http://www.ipcc.ch/pdf7assessment-report/ar4/wgl/ar4wgl-chapter8.pdf. A good discussion of uncertainties in models can be found in Gerald A. Meehl et al., Global Climate Projections, in INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE, supra, at 747, 754-60, 797-810. A more detailed introduction to modeling can be found in KENDAL MCGUFFIE & ANN HENDERSON- SELLERS, A CLIMATE MODELLING PRIMER (3d ed. 2005), although it should be noted that the authors' statement that the book requires "basic high school mathematics," id. at xiv, is true only for those readers whose basic high school courses included vector calculus, and some details of the discussion presume some post-high school exposure to physics. Among other useful features, Appendix B of the book contains a glossary of technical vocabulary, most helpful for those of us whose daily conversation does not include terms like "advection" or "adiabatic." Id. at 257-74. The basic concepts underlying climate modeling are explained in JOHN HARTE, CONSIDER A SPHERICAL COW: A COURSE IN ENVIRONMENTAL PROBLEM SOLVING (1988). For a general overview of the current state of knowledge about climate issues, see PETER J. ROBINSON & ANN HENDERSON-SELLERS, CONTEMPORARY CLIMATOLOGY (2d ed. 1999).
8. This Article focuses on the uncertainties in predicting the degree of harm from climate change. There are also large uncertainties about the costs of mitigation measures. Oddly, those who view uncertainties about harm as a basis for ignoring the impact of climate change tend to place a great deal of credence in economic models of mitigation cost-in their minds, uncertainty apparently is relevant when considering scientific models but not economic models. See Philippe Tulkens & Henry Tulkens, The White House and the Kyoto Protocol: Double Standards on Uncertainties and Their Consequences 10 (Fondazione Eni Enrico Mattel, Working Paper No. 89.2006, June 2006), available at http://ssrn.com/abstract=910811 ("These[] misleading facts and figures show that a balanced approach in dealing with the two uncertainties on the science and on the economics of the issue was not followed in the White House communication about the Kyoto Protocol and the existence of anthropogenic climate change.").
9. ELIZABETH KOLBERT, FIELD NOTES FROM A CATASTROPHE: MAN, NATURE, AND CLIMATE CHANGE 99-101 (2006).
10. Id. at 101.
11. Id. at 100-02.
12. Hervé Le Treut et al., Historical Overview of Climate Change Science, in INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE, supra note 7, at 93, 98. For a detailed discussion, see generally id. at 93, 93-127.
13. Id. at 98.
14. Id. at 112.
15. MCGUFFIE & HENDERSON-SELLERS, supra note 7, at 177.
16. Id. at 9.
17. Id. at 48. Early models turned out to be much too sensitive to perturbations. Id. at 53.
18. Id. at xiii.
19. Id. at 6.
20. Id. at 58.
21. Id. Perhaps ironically in light of the origin of the models, there is now some tension between the views of climate scientists and those of meteorologists about matters such as the relationship of climate change to hurricane intensity. This topic is explored in depth in CHRIS MOONEY, STORM WORLD: HURRICANES, POLITICS, AND THE BATTLE OVER GLOBAL WARMING (2007).
22. See MCGUFFIE & HENDERSON- SELLERS, supra note 7, at 208 (explaining the decreasing need for flux adjustment as modeling improves).
23. Id. at 27.
24. Id. at 29.
25. Randall et al., supra note 7, at 589, 592.
26. Id. at 602.
27. Id. at 604.
28. Id. at 605.
29. MCGUFFIE & HENDERSON-SELLERS, supra note 7, at 72. For a discussion of efforts to model land use, see generally Edwin van der Werf & Sonja M. Peterson, Modeling Linkages Between Climate Policy and Land Use (Fondazione Eni Enrico Mattel, Working Paper No. 56.2007, May 2007), available at http://ssrn.com/abstract= 989968.
30. M. Bruce Beck, How Best to Look Forward?, 316 SCIENCE 202, 202 (2007). For an assessment of recent improvement in climate-related data and areas for future improvement, see generally COMM. ON STRATEGIC ADVICE ON THE U.S. CLIMATE CHANGE SCI. PROGRAM, NAT'L RESEARCH COUNCIL, EVALUATING PROGRESS OF THE U.S. CLIMATE CHANGE SCIENCE PROGRAM: METHODS AND PRELIMINARY RESULTS 51-97 (2007).
31. Le Treut et al., supra note 12, at 93, 102.
32. Id. at 116.
33. MCGUFFIE & HENDERSON-SELLERS, supra note 7, at 32.
34. Randall et al., supra note 7, at 589, 613.
35. MCGUFFIE & HENDERSON-SELLERS, supra note 7, at 199.
36. The demands posed by these models will only increase. To deal with needed model improvements, "the United States will need at least a thirtyfold increase in high-performance computing resources within the next five years." COMM. ON STRATEGIC ADVICE ON THE U.S. CLIMATE CHANGE SCI. PROGRAM, supra note 30, at 110.
37. MCGUFFIE & HENDERSON-SELLERS, supra note 7, at 4. 38. Id. at 9.
38. Id. at 56. Since there are 360 degrees of latitude and longitude, this means dividing the earth's surface into roughly 2,500-20,000 pieces.
39. Id.
40. See id. at 118 (explaining the relationship between parameterization and the potential for long-term projections).
41. For a detailed discussion of EMICs, see id. at 117-62.
42. Id. at 150.
43. Randall et al., supra note 7, at 589, 592.
44. Id.
45. Id. at 643.
46. Id. at 644.
47. Meehl et al., supra note 7, at 747, 797.
48. Id.
49. Id. at 797, 800.
50. Id. at 797.
51. Id.
52. MCGUFFIE & HENDERSON-SELLERS, supra note 7, at 12.
53. -2." Id. at 3 (footnote omitted).
54. Richard A. Kerr, Scientists Tell Policymakers We're All Warming the World, 315 SCIENCE 754, 754 (2007); see also id. at 755 ("[T]he IPCC authors weren't impressed by the contrarian argument that the warming is just an 'urban heat island effect' driven by increasing amounts of heat-absorbing concrete and asphalt.") (quoting Intergovernmental Panel on Climate Change, supra note 54, at 1,5).
55. Id. at 755 (quoting Intergovernmental Panel on Climate Change, supra note 54, at 1, 5).
56. See id. ("[N]ew analyses have largely settled the hullabaloo over why thermometers at Earth's surface measured more warming than remote-sensing satellites had detected higher in the atmosphere.").
57. See id. ("But all of these natural variations in combination have not warmed the world enough, fast enough, and for long enough in the right geographic patterns to produce the observed wanning . . ..").
58. Id.
59. Intergovernmental Panel on Climate Change, supra note 54, at 1, 3 & n.7.
60. Randall et al., supra note 7, at 589, 591.
61. Le Treut et al., supra note 12, at 93, 114.
62. Id.
63. Id.
64. Id. at 117.
65. Id. at 117-18.
66. Peter Cox & David Stephenson, A Changing Climate for Prediction, 317 SCIENCE 207, 208 (2007) (citations omitted).
67. Id. As a result, the most reliable forecasts are for lead times between thirty and fifty years. Id. The authors suggest that better data about current conditions could help pin down the parameters. Id.
68. Le Treut et al., supra note 12, at 93, 111.
69. Id. at 121.
70. Randall et al., supra note 7, at 589, 629.
71. Id.
72. Id.
73. Id. at 633.
74. Id. at 635. The IPCC's summary on the relationship between clouds and current models is as follows: Despite some advances in the understanding of the physical processes that control the cloud response to climate change and in the evaluation of some components of cloud feedbacks in current models, it is not yet possible to assess which of the model estimates of cloud feedback is the most reliable. However, progress has been made in the identification of the cloud types, the dynamical regimes and the regions of the globe responsible for the large spread of cloud feedback estimates among current models. Id. at 638.
75. See id. at 600 ("One source of confidence in models comes from the fact that model fundamentals are based on established physical laws, such as conservation of mass, energy and momentum, along with a wealth of observations.").
76. Id. at 593.
77. Id.
78. Id. at 594.
79. Id.
80. Id. at 595.
81. Id. at 600. "Skill" is an indication of a model's ability to predict patterns of events rather than averages. Robert D. Hetland, Short Communication: Event-Driven Model Skill Assessment, 11 OCEAN MODELING 214,215 (2006).
82. Randall et al., supra note 7, at 589, 623.
83. Id. at 627.
84. Myles Alien et al., Scientific Challenges in the Attribution of Harm to Human Influences on Climate, 155 U. PA. L. REV. 1353, 1364 (2007).
85. Id. at 1366.
86. Meehl et al., supra note 7, at 747, 754; see also MCGUFFIE & HENDERSON-SELLERS, supra note 7, at 230 ("[T]he model group mean (after excluding unreasonable results/outliers) outperforms any one model, where performance is measured against observational data.").
87. James M. Murphy et al., Quantification of Modelling Uncertainties in a Large Ensemble of Climate Change Simulations, 430 NATURE 768, 769 (2004).
88. Id. at 770 fig.3.
89. MCGUFFIE & HENDERSON-SELLERS, supra note 7, at 16.
90. See id. at 240 (noting that "integrated assessment models" have only been under development for about ten years, as opposed to forty years for climate models).
91. See Richard Wolfson & Stephen H. Schneider, Understanding Climate Science, in STEPHEN H. SCHNEIDER ET AL., CLIMATE CHANGE POLICY: A SURVEY 3, 46, 31-41 (2002) (exploring scientists' consensus over projected global warming and its consequences, and concluding that "uncertainty ... may temper but not destroy our confidence in scientific projections").
92. For a list, see MCGUFFIE & HENDERSON- SELLERS, supra note 7, at 242.
93. Id. at 240-43. The treatment of uncertainty is tabulated in id. at 242 tbl.6.6.
94. NICHOLAS STERN, THE ECONOMICS OF CLIMATE CHANGE: THE STERN REVIEW 166-67 (2007).
95. Id. at 146-47.
96. Id. at 148.
97. Id.
98. WILLIAM D. NORDHAUS & JOSEPH BOYER, WARMING THE WORLD: ECONOMIC MODELS OF GLOBAL WARMING 76 (2000).
99. Id. at 51. Models using such "history-based projections" are vulnerable to "structural breaks" and "radically new technologies" that can disrupt trends. S.J. Tol, Carbon Dioxide Emission Scenarios for the USA 9 (Fondazione Eni Enrico Mattel, Working Paper No. 117.2006, Sept. 2006), available at http://www.ssrn.com/abstract=932508.
100. NORDHAUS & BOYER, supra note 99, at 87-88. While this Article was in final publication stages, a new book, WILLIAM NORDHAUS, A QUESTION OF BALANCE: WEIGHING THE OPTIONS ON GLOBAL WARMING POLICIES (2009), was published. It contains a more explicit treatment of uncertainty than the earlier work.
101. STERN, supra note 95, at 186. For an appraisal of The Stern Review and its critics, see Rick Baker et al., The Stern Review: An Assessment of Its Methodology (Austl. Gov't Productivity Comm'n, Staff Working Paper, Jan. 2008), available at http://www.pc.gov.au/ data/assets/pdffile/0005/74723/sternreview.pdf.
102. Jason F. Shogren & Michael A. Toman, How Much Climate Change is Too Much?, in CLIMATE CHANGE ECONOMICS AND POLICY: AN RFF ANTHOLOGY 35,42 (Michael A. Toman ed., 2001).
103. Id.
104. Id. at 43.
105. STERN, supra note 95, at 269.
106. Id. at 271.
107. See Olivier Deschenes & Michael Greenstone, The Economic Impacts of Climate Change: Evidence from Agricultural Profits and Random Fluctuations in Weather, 96 AM. ECON. REV. 354, 355-57 (2007) (arguing that climate change would in fact marginally benefit U.S. farmers overall, although a few states would suffer considerable agricultural losses). But note that this study excludes possible impacts of increases in extreme events such as storms and droughts. Id.
108. Wolfram Schlenker et al., The Impact of Global Warming on U.S. Agriculture: An Econometric Analysis of Optimal Growing Conditions, 88 REV. ECON. & STAT. 113, 120-23 (2006).
109. A good overview of modeling issues can be found in J.C. Hourcade et al., Estimating the Costs of Mitigating Greenhouse Gases, in INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE, CLIMATE CHANGE 1995: ECONOMIC AND SOCIAL DIMENSIONS OF CLIMATE CHANGE, CONTRIBUTION OF WORKING GROUP III TO THE SECOND ASSESSMENT REPORT OF THE INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE 263, 263-96 (James P. Bruce et al. eds., 1996). Of course, in the decade since this Report, models have improved in their capacity to handle these issues. The MIT Integrated Global System Model seems to be particularly well-regarded. For information about this model, see MIT Integrated Framework: Economics and Anthropogenic Emissions, http://web.mit.edu/globalchange/www/eppa.htmltfpub.
110. NORDHAUS & BOYER, supra note 99, at 98. This does not, however, impede Nordhaus and Boyer from issuing confident policy pronouncements. See id. at 95-98.
111. STEPHEN J. DECANIO, ECONOMIC MODELS OF CLIMATE CHANGE: A CRITIQUE 130-37, 137tbl.5.4(2003).
112. See Matthew D. Zinn, Adapting to Climate Change: Environmental Law in a Warmer World, 34 ECOLOGY L.Q. 61, 81-84 (2007) (noting that adaptation may not be successfully managed to minimize ecological or other impacts).
113. Indeed, they may have been because the book describing the model is coauthored. NORDHAUS & BOYER, supra note 99, at xi. The level of complexity of the economic models can be gauged from the fact that the Nordhaus model is designed to run on a PC, id. at 56, while today's most advanced climate models require weeks to run on a supercomputer.
114. Id. at 17.
115. Id. at 20.
116. Id. at 47.
117. Id. at 53.
118. Additional examples include disclaimers that: (1) "there are no established methodologies for valuing catastrophic risk," id. at 71; (2) findings regarding climate impacts are "highly conjectural" and it is difficult "to make solid estimates of the impacts of climate change," id.; and (3) "[g]iven the lack of any comprehensive estimates, the authors have made rough estimates here of the extent to which the economy and other institutions are vulnerable to climate change," id. at 86.
119. Id. at 96.
120. Id. at 127.
121. Id. at 129.
122. Id. at 174.
123. Id. at 177.
124. For an example of similar overstatement, see Richard S.J. Tol, The Marginal Damage Costs of Carbon Dioxide Emissions: An Assessment of the Uncertainties, 33 ENERGY POL'Y 2064 (2005). Tol says: "One can therefore safely say that, for all practical purposes, climate change impacts may be very uncertain but it is unlikely that the marginal damage costs of carbon dioxide emissions exceed $50/tC and are likely to be substantially smaller than that." Id. at 2073. A more accurate way of introducing his findings would have been: "While all existing models are imperfect, an assessment of their overall predictions and uncertainties suggests that. . . ." He might also have done better to unpack the terms "unlikely" and "likely." Considering peer-reviewed studies only, Table 3 of his article shows a mean marginal cost of $50, a 10% probability that the cost is over $125, and a 5% probability that the cost is over $245. Id. at 2071. If we used confidence intervals (the 5% to 95% range), the confidence interval for the $50 per ton estimate would go from -$9 to +$245. One might even say that, for all practical purposes, we are essentially clueless about the right number. It also turns out that much of the uncertainty is actually based on ethical disagreements, depending on how the model accounts for impacts on the poor (equity weighting") and on future generations ("discount rates"). Id. at 2072-73. So another restatement of Tol's results would be: "At this point, climate change is primarily an ethical rather than economic issue."
125. Tulkens & Tulkens, supra note 8, at 8.
126. See id. at 16 (noting that studies comparing pre- and post-implementation cost estimates for regulatory policies have a "predominant tendency to overestimate the cost ex ante").
127. For a good overview of adaptation issues, see generally Zinn, supra note 113.
128. Intergovernmental Panel on Climate Change, Summary for Policymakers, in INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE, CLIMATE CHANGE 2007: IMPACTS, ADAPTATION AND VULNERABILITY, CONTRIBUTION OF WORKING GROUP II TO THE FOURTH ASSESSMENT REPORT OF THE INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE 7, 19 (Martin Parry et al. eds., 2007) (footnotes omitted).
129. See id. at 724-27 (assessing adaptation costs and benefits).
130. Google Earth, http://earth.google.com.
131. For further discussion of the application of GIS to environmental issues, see generally Robert J. Goldstein, Putting Environmental Law on the Map: A Spatial Approach to Environmental Law Using GIS, in LAW AND GEOGRAPHY 523, 523-38 (Jane Holder & Carolyn Harrison eds., 2003).
132. IAN HEYWOOD ET AL., AN INTRODUCTION TO GEOGRAPHICAL INFORMATION SYSTEMS 246-47, 250 (1998).
133. Id. at 182-84.
134. Id. at 184-86.
135. Jody Freeman & Daniel A. Farber, Modular Environmental Regulation, 54 DUKE L.J. 795, 813(2005). The authors explain: [T]he division of responsibility for environmental regulation and natural resource management by media, among agencies at the same level of government, and among different levels of government can be extremely complex. The different mandates, culture, management structures, and budgetary priorities of the variety of agencies involved in environmental regulation and natural resource management can present, to put it mildly, significant challenges for coordination. Id.
136. HEYWOOD ET AL., supra note 133, at 177-78.
137. On the general problem of educating judges and juries about scientific issues, see ELAINE L. SPENCER, USE AND MISUSE OF TECHNICAL DATA: TELLING THE SCIENTIFIC STORY TO SCIENTIFIC VIRGINS (2006), available at http://www.grahamdunn.com/go/articles/use-and- misuse-of-technical-data-telling-the-scientific-story-to-scientific-virgins.
138. Randall et al., supra note 7, at 589, 591.
139. Id. at 592.
140. 509 U.S. 579 (1993).
141. Id. at 582.
142. Id. at 583.
143. Id.
144. Id. at 585.
145. Id. at 587.
146. Id. at 592-95.
147. Id. at 594-95.
148. Id. at 593-94 (citations omitted).
149. 522 U.S. 136(1997).
150. Id. at 146.
151. 526 U.S. 137(1999). For a general discussion of these opinions and their implications, see the multi-volume resource DAVID L. FAIGMAN ET AL., MODERN SCIENTIFIC EVIDENCE (2005). These decisions have not been without their critics. For instance, a well known environmental scholar contends that they have prompted lower courts to reject sound cases: It is now clear after more than a decade's experience with Dauben that the lower courts have applied it quite vigorously to screen out not only "junk science" but also a good deal of "sound science" as well. Since the plaintiff ordinarily has the burden of proof in tort litigation, this aggressive invocation of the judge's new role as guardian of the purity of scientific evidence has clearly had a disproportionate impact on plaintiffs. A plaintiffs attorney must come to court prepared not only to establish the expert's qualifications, but also to demonstrate to a skeptical trial judge that the testimony forms scientifically reliable conclusions based upon reliable data and that those conclusions "fit" the legal requirements for establishing cause-in-fact. Thomas O. McGarity, Our Science is Sound Science and Their Science is Junk Science: Science-Based Strategies for Avoiding Accountability and Responsibility for Risk-Producing Products and Activities, 52 U. KAN. L. REV. 897, 910 (2004). For an excellent treatment of the application of the Daubert test to climate models, see Matthew W. Swinehart, Note, Remedying Daubert 's Inadequacy in Evaluating the Admissibility of Scientific Models Used in Environmental Tort Litigation, 86 TEXAS L. REV. 1281(2008).
152. Kumho Tire, 526 U.S. at 142-45.
153. Id. at 152.
154. FED. R. EVID. 702 advisory committee's note (amended 2000).
155. FED. R. EVID. 702.
156. FED. R. EVID. 702 advisory committee's note.
157. Id.
158. Randall et al., supra note 7, at 589, 591.
159. Id. at 593.
160. Id.
161. Id. at 594.
162. MCGUFFIE & HENDERSON-SELLERS, supra note 7, at 229.
163. See, e.g., Gen. Elec. Co. v. Joiner, 522 U.S. 136, 141 (1997) (stating that abuse of discretion is the proper standard for reviewing a district court's evidentiary rulings); Stokes v. Ga.-Pac. Corp., 894 F.2d 764, 767 (5th Cir. 1990) (stating that a trial judge's ruling on the admissibility of evidence is reviewed for abuse of discretion).
164. See CARL F. CRANOR, TOXIC TORTS: SCIENCE, LAW AND THE POSSIBILITY OF JUSTICE 18-25, 328-30 (2006) (suggesting that the variability is partly due to the difficulty, even among experts, in interpreting nuanced scientific results).
165. See, e.g., Alien v. Pa. Eng'g Corp., 102 F.3d 194, 198 (5th Cir. 1996) (affirming the district court's decision to exclude an expert's opinion that was based on the "weight of the evidence" rather than "rigorous scientific study"); Wright v. Willamette Indus., Inc., 91 F.3d 1105, 1107-08 (8th Cir. 1996) (explaining that an expert doctor's testimony on the cause of the plaintiffs injuries should have been excluded at trial under Daubert because the doctor's opinion was based on a preponderance of the evidence rather than scientific knowledge).
166. See Russellyn S. Carruth & Bernard D. Goldstein, Relative Risk Greater than Two in Proof of Causation in Toxic Tort Litigation, 41 JURIMETRICS J. 195, 200-01 (2001) (analyzing the efficacy of adopting a bright-line rule of relative risk to establish causation in toxic-torts cases).
167. FED. R. EVID. 702 advisory committee's note.
168. Id.
169. 127 S. Ct. 1438 (2007).
170. Id. at 1451.
171. Id. at 1449.
172. Massachusetts v. EPA, 415 F.3d 50, 55, 58 (D.C. Cir. 2005).
173. Massachusetts v. EPA, 127 S. Ct. at 1455.
174. Id. at 1455 (citations omitted) (quoting 68 Fed. Reg. 52,930 (Sept. 8, 2003) and COMM. ON THE SCI. OF CLIMATE CHANGE, NAT'L RESEARCH COUNCIL, CLIMATE CHANGE: AN ANALYSIS OF SOME KEY QUESTIONS 16 (2001)).
175. Id. at 1456.
176. Id.
177. See Green Mountain Chrysler Plymouth Dodge Jeep v. Crombie, 508 F. Supp. 2d 295, 310-33 (D. Vt. 2007) (analyzing in depth the qualifications of the various expert witnesses called at trial under Daubert and its progeny).
178. Id. at 320.
179. Id. at 313-16.
180. Id. at 317.
181. Id.
182. Id. at 318.
183. Id.
184. Id. at 325.
185. Id. at 321-22.
186. Id. at 322.
187. Id.
188. Id. at 323-24.
189. Climate skeptics may also attempt to present their own modeling results in court. Given the lack of peer review, general rejection by the community of climate scientists, and the other Daubert factors, such testimony should arguably be inadmissible.
190. See generally Timothy W. Cerniglia, Computer-Generated Exhibits-Demonstrative, Substantive, or Pedagogical-Their Place in Evidence, 18 AM. J. TRIAL ADVOC. 1 (1994) (describing the different types of computer-generated evidence and the confusion that they have engendered in the courts); Gregory P. Joseph, A Simplified Approach to Computer-Generated Evidence and Animations, 43 N.Y.L. SCH. L. REV. 875 (1999) (analyzing issues of trustworthiness and hearsay with regard to computer-generated models). On the concern that visual evidence might be given undue weight, see Hampton Dellinger, Words Are Enough: The Troublesome Use of Photographs, Maps and Other Images in Supreme Court Opinions, 110 HARV. L. REV. 1704, 1749 ( 1997). Dellinger remarks: [T]he Court's use of photographs, maps, and other attachments has been both rich and problematic: rich in that attachments have accompanied opinions in a number of critical cases, and problematic in that their presence has been more distracting than illuminating. At worst, as with the "giant" cross in Capitol Square, the visual attachments freeze a badly atypical, and thus misleading, version of reality. The foregoing analysis of the Supreme Court's use of attachments reveals clear failings. And there is little reason to be sanguine about the future. Egregious examples of the misuse of attachments have continued to appear in recent years. With developing computer technologies, attaching sight-based objects will only get easier, while the susceptibility of such objects to manipulation will only increase. Id. at 1749 (emphasis added).
191. For general discussions of EPA modeling and its evaluation by courts, see Fine & Owen, supra note 6, at 935-36 (discussing judicial review of the use of models), and Thomas O. McGarity & Wendy E. Wagner, Legal Aspects of the Regulatory Use of Environmental Modeling, 33 Envtl. L. Rep. (Envtl. Law Inst.) 10,751 (2003) (drawing conclusions from numerous instances of substantive review of agency models).
192. 784 F.2d 224 (6th Cir. 1986).
193. Id. at 226.
194. The court described the model in some detail: The CRSTER model is a single source model designed for application to hot, buoyant, stack effluents of the kind commonly produced by power plant and furnace chimneys. After factoring in background pollution, the CRSTER model is actually capable of modeling up to 19 sources, but only if they are in close proximity to one another-multiple stacks at a single plant, for example. It treats all sources being modeled as occupying the same location. The CRSTER model is a "gaussian plume dispersion" model, which means that it describes how a plume of dirty air emerging from a smoke stack will spread upward and outward according to the principles of fluid dynamics. The model consists of a "preprocessor subroutine," which translates input data such as meteorological conditions and source characteristics into a format suitable for the model, and a set of subroutines that perform the actual modeling. These subroutines use data translated by the preprocessor to calculate how pollution from the source will spread out in light of the conditions (e.g., wind speed and direction) embodied in that data. In so doing, the model makes a number of simplifying assumptions about such factors as the constancy of wind speed and direction, the uniformity of emission, the inability of the plume to cross a low atmospheric layer called the "mixing level," the absence of vertical wind shear, the nonreactivity of the effluent, and the degree of diffusion of the plume. Id. at 228 (citation omitted).
195. Id. at 229.
196. Id. at 230. The court reaffirmed its position in a follow-up opinion, Ohio v. EPA, 798 F.2d 880, 882 (6th Cir. 1986).
197. See McGarity & Wagner, supra note 192, at 10,761 (calling the decision "[o]ne of the most infamous and arguably aberrational model cases"). By five years later, the case was being cited in its own circuit merely for the proposition that air models are used to predict ozone dispersion. See, e.g., Wall v. EPA, 265 F.3d 426, 428 (6th Cir. 2001) (citing Ohio v. EPA, 784 F.2d at 228-29). The court's earlier suspicion of those models seemed to have been forgotten. See Sierra Club v. EPA, 375 F.3d 537, 540 (7th Cir. 2004) (reading Wall as allowing flexibility over the use of modeling methods). Of course, modeling methods improved after the case was decided, but the models remain imperfect: Models have improved since the early days of air quality planning. As simulation models have evolved, they have become increasingly complex; this complexity has reduced the extent to which models oversimplify the real world. Current models use more monitoring data, greater computing power, improved and expanded algorithms, and more efficient, accurate mathematical solution methods than did their predecessors. Nevertheless, models' skill in simulating ozone concentrations has not increased commensurately, and significant limitations remain for both the mathematical representations and applications of air quality models. Additionally, due to the time and effort involved in developing or updating an air quality model, usually no single "state-of-the-science" model contains a formulation depicting the most modern scientific understanding. As a result, the difficulty of modeling complex processes remains a source of uncertainty. Fine & Owen, supra note 6, at 924 (citations omitted).
198. See McGarity & Wagner, supra note 192, at 10,759, 10,757-61 (noting that "[successful challenges to EPA's modeling exercises have almost uniformly involved unhelpful and nonspecific Agency responses" to criticisms of EPA's choice of models).
199. 28 F.3d 1259, 1264-65 (D.C. Cir. 1994).
200. See, e.g., Richard J. Pierce Jr., Two Problems in Administrative Law: Political Polarity on the District of Columbia Circuit and Judicial Deterrence of Agency Rulemaking, 1988 DUKE L.J. 300, 323-24 (emphasizing the "long accepted" principle that "courts should defer to agency policy decisions made in conditions of uncertainty").
201. 462 U.S. 87 (1983).
202. Id. at 103.
203. Id.
204. McGarity & Wagner, supra note 192, at 10,752.
205. Id. at 10,768.
206. Id.
207. Information Quality Act, Pub. L. No. 106-554, 114 Stat. 2763 (2000).
208. See, e.g., Letter from Christopher C. Horner, Competitive Enter. Inst, to Info. Officer, EPA, (Feb. 10, 2003), available at http://www.epa.gov/quality/mformationguidelines/documents/ 7428.pdf ("seeking a substantive response to" a petition to halt dissemination of a climate-action report).
209. Wendy E. Wagner, Commons Ignorance: The Failure of Environmental Law to Produce Needed Information on Health and the Environment, 53 DUKE L.J. 1619, 1713 n.338 (2004). McGarity provides more details about the dispute: In early 2003, the Competitive Enterprise Institute, a think tank that has historically opposed national and international efforts to abate greenhouse gasses, filed an IQA challenge in three agencies demanding that they "withdraw" the National Assessment on Climate Change (NACC), an interagency report on the role that greenhouse gasses play in global warming. Although the Report had received extensive peer review and public vetting, CEI nevertheless launched a classic corpuscular attack on various aspects of the Report that were not, in CEI's view, based on "sound science." After the White House Office of Science and Technology Policy (OSTP) denied CEI's petition, CEI sued President Bush and the Director of OSTP seeking a judicial ruling on the merits of its IQA challenge. The case subsequently settled when the federal government agreed to place a disclaimer on the NACC advising that it had not been prepared in accordance with the requirements of the IQA. McGarity, supra note 152, at 925.
210. McGarity, supra note 152, at 913 (quoting the Information Quality Act, 114 Stat. at 2763 A-153-54).
211. Recent OMB efforts to implement the statute by mandating peer review should not impede agency use of climate models, given the prevalence of peer review and other institutional review procedures. For a discussion of OMB's initiatives, see Patrick A. Fuller, Note, How Peer Review of Agency Science Can Help Rulemaking: Enhancing Judicial Deference at the Frontiers of Knowledge, 75 GEO. WASH. L. REV. 931, 940-43 (2007). One concern of scientists is that they will be exposed to fishing expeditions into their data and methods, but a recent case indicates that the IQA does not provide industry with a judicially enforceable right to such information. Salt Inst. v. Leavitt, 440 F.3d 156, 159 (4th Cir. 2006).
212. Allen et al., supra note 85, at 1361.
213. Id. "There is considerable debate over the extent to which currently available models span the range of plausible real-world responses." Id.
214. Id. at 1375.
215. See supra subpart II(C).
216. See supra subpart II(C).
217. Meehl et al., supra note 7, at 747, 797 (citations omitted).
218. Id. at 754.
219. Id.
220. Id.
221. Id. at 798.
222. Id. at 798-99.
223. Id.
224. Id. at 799.
225. Id.
226. Id.
227. Id. at 805 (citations omitted).
228. See Myanna Lahsen, Seductive Simulations? Uncertainty Distribution Around Climate Models, 35 SOC. STUD. SCI. 895, 898 (2005) (describing the limitations of, and problems with, climate models).
229. See J.E. Hansen, Scientific Reticence and Sea Level Rise, 1 ENVTL. RES. LETTERS (ISSUE 2), Apr-June 2007, Letter 2, at 1-2, available at http://www.iop.org/EJ/article/1748- 9326/2/2/024002/erl7_2_024002.pdf (labeling this phenomenon "scientific reticence").
230. For an excellent general discussion of the treatment of uncertainty in policy analysis, see M. GRANGER MORGAN & MAX HENRION, UNCERTAINTY: A GUIDE TO DEALING WITH UNCERTAINTY IN QUANTITATIVE RISK AND POLICY ANALYSIS (1990). They observe that many approaches exist regarding different applications: As we have seen, there exists an almost overwhelming variety of different methods for representing, propagating, and analyzing uncertainties. They have been developed and applied in a wide range of modeling domains, in policy and risk analysis as well as more traditional areas of science and engineering. These include financial investment planning, physics, chemical engineering, nuclear engineering and safety analysis, ecological modeling and atmospheric modeling. Id. at 211. The legal literature, to the extent that uncertainty is considered in any rigorous way, seems to rely almost exclusively on the standard economic equation of risk with outcome variance. Granger and Henrion note that in settings "such as nuclear safety analysis"-and one might add climate change here-"where the tails of the distributions are of particular interest," reliance on the mean and variance as a basis for evaluating uncertainty can entail "a serious deficiency." Id. at 213. They prefer Monte Carlo sampling
231. The discussion in the next two paragraphs draws heavily from Daniel A. Farber, Probabilities Behaving Badly: Complexity Theory and Environmental Uncertainty, 37 U.C. DAVIS L. REV. 145 (2003).
232. Id. at 155.
233. Coupled human and natural systems may be even more prone to "nonlinear dynamics with thresholds, reciprocal feedback loops, time lags, resilience, heterogeneity, and surprises." See Jianguo Liu et al., Complexity of Coupled Human and Natural Systems, 317 SCIENCE 1513, 1513 (2007) (reviewing interdisciplinary studies of feedback between natural and human systems).
234. Farber, supra note 232, at 153.
235. Id.
236. Id.
237. Id. at 155.
238. See BENOIT B. MANDELBROT, THE FRACTAL GEOMETRY OF NATURE 337-38 (1983) (introducing the notion of an infinite-variance syndrome in the price-modeling context).
239. Farber, supra note 232, at 155.
240. Id.
241. Id.
242. Thomas C. Schelling, Climate Change: The Uncertainties, the Certainties, and What They Imply About Action, 4 ECONOMISTS' VOICE (ISSUE 3), July 2007, at 4, available at http://www.be press.com/ev/vol4/iss3/art3.
243. Id.
244. See Kenneth J. Arrow, Global Climate Change: A Challenge to Policy, 4 ECONOMISTS' VOICE (ISSUE 3), June 2007, at 2, available at http://www.bepress.com/ev/vol4/iss3/art2.
245. Martin L. Weitzman, The Role of Uncertainty in the Economics of Catastrophic Climate Change 17 (AEI-Brookings Joint Ctr., Working Paper No. 07-11, 2007), available at http://ssrn.com/abstract=992873. Weitzman shows that even a small degree of uncertainty about a single model parameter can become magnified into substantial economic risk. Id. at 12-19.
246. 42 U.S.C. §§ 4321-437 of (2000).
247. 40 C.F.R. § 1502.22(b) (2007).
248. Id.
249. Michael Funk & Michael Paetz, Environmental Policy Under Model Uncertainty: A Robust Optimal Control Approach 2 (CESifo, Working Paper No. 1938, March 2007), available at http://www.ssrn.com/abstract=971461.
250. Id. at 1-2. In essence, robust optimal control requires the policy maker to consider alternative models that are in some sense close to the preferred model but that produce more drastic predictions-thus, it could be viewed as a way to consider plausible worst-case scenarios. Id. at 2.
251. David G. Groves, New Methods for Identifying Robust Long-Term Water Resources Management Strategies for California 123-25 (Sept. 2005) (unpublished Ph.D. dissertation, Pardee RAND Graduate School), http://www.rand.org/pubs/ rgs_dissertations/2006/RAND_RGSD196.pdf.
252. Id. at 125. See also David G. Groves & Robert J. Lempert, A New Analytic Method for Finding Policy-Relevant Scenarios, 17 GLOBAL ENVTL. CHANGE 73, 75-76 (2007) (describing the advantages of RDM).
253. Groves, supra note 252, at 125.
254. What modelers "do best" has been described as follows: [Mobilizing the current science base (uncertainties, warts, and all) into a computational model; thus to assess the plausibility of the imagined futures, under gross uncertainty; thus to identify what we would most like to understand much better, in the here and now, about the uncertainties (be they elements of the science, the policies, or the technologies) that are crucial to discriminating between the reachability of our worst fears or our greatest hopes. Nothing stands still: not images of the future, nor our qualitative models, nor our quantitative models. Call this living in a "recursive predictive world" .... Beck, supra note 30, at 202.
255. For a description of the development of OMB's role in regulatory oversight, along with sensible suggestions for improving cost-benefit analysis, see Daniel H. Cole, Towards 'Best Practice' Standards in Environmental Benefit-Cost Analysis, 23 RES. LAW&ECON. 1, 8 (2007).
256. Exec. Order No. 12,291, 46 Fed. Reg. 13,193 (Feb. 17, 1981).
257. Id. at 13,193.
258. Id. at 13,193-94.
259. Exec. Order No. 12,866, 3 C.F.R. 638 (1994), reprinted in 5 U.S.C. § 601.
260. See Ellen Siegler, Executive Order 12866: An Analysis of the New Executive Order on Regulatory Planning and Review, 24 Envtl. L. Rep. (Envtl. Law Inst.) 10,070, 10,070 (1994) (stating that the Order sought to eliminate useless regulations and to streamline the regulatoryreview process).
261. See Stuart Shapiro, Unequal Partners: Cost-Benefit Analysis and Executive Review of Regulations, 35 Envtl. L. Rep. (Envtl. Law Inst.) 10,433, 10,437-38 (2005) (arguing that centralized review may lead to increased political influence on outcomes).
262. Professor McGarity describes the Bush Administration's efforts as follows: [T]he battle continued to rage when President George W. Bush appointed John D. Graham, a strong proponent of cost-benefit analysis from the Harvard Center for Risk Analysis, to head the Office of Information and Regulatory Affairs (OIRA) in the Office of Management and Budget. This nomination was strongly supported by regulated industries and equally strongly opposed by public interest groups. From that position, to which he was confirmed in late 2001, Graham has overseen the centralized review process for health and safety regulations. To the chagrin of public interest groups and the joy of industry-funded think tanks, OIRA greatly stemmed the flow of health, safety and environmental regulation during the Bush Administration. Although EPA promulgated several important regulations, most of which were required by statute, OSHA did not promulgate a single significant health standard during the entire four years. Thomas O. McGarity, The Story of the Benzene Case: Judicially Imposed Regulatory Reform Through Risk Assessment, in ENVIRONMENTAL LAW STORIES 141, 169 (Richard J. Lazarus & Oliver A. Houck eds., 2005).
263. For a vigorous critique of the way cost-benefit analysis treats environmental and health benefits, see Frank Ackerman & Lisa Heinzerling, Pricing the Priceless: Cost-Benefit Analysis of Environmental Protection, 150 U. PA. L. REV. 1553 (2002). Some economists advocate the use of "contingent valuation" studies to measure how much people are willing to pay for nonuse values. Contingent valuation is essentially a survey technique. People are given information about an environmental issue and then asked if they would be willing to pay a certain amount to solve the problem. There is a great deal of dispute about whether contingent valuation, even if done carefully, provides a genuine measure of preferences. Cass Sunstein, for example, finds many contingent-valuation analyses difficult to take seriously. He stresses what he describes as the "astonishing and devastating fact" that willingness to pay seems constant regardless of the scale of the environmental problem. CASS R. SUNSTEIN, FREE MARKETS AND SOCIAL JUSTICE 142-43 (1997). For an environmentalist critique of contingent valuation, see John M. Heyde, Is Contingent Valuation Worth the Trouble?, 62 U. CHI. L. REV. 331, 343-53 (1995) (arguing that contingent valuation is ineffective because the "values being measured... are so hazy and ill defined that people's responses are much more subject to biases and variability" than with tangible goods). Another approach for valuation of nonmarket costs and benefits is the concept of ecosystem services. See generally Gretchen C. Daily, et al., Protecting Ecosystem Services: Science, Economics, and Law, 20 STAN. ENVTL. L.J. 309, 309-13 (2001) (defending the value of the ecosystem-services approach); James Salzman, Creating Markets for Ecosystem Services: Notes from the Field, 80 N.Y.U. L. REV. 870, 877-88, 911-32 (2005) (explaining the concept of ecosystem services and sharing lessons learned from the ecosystem of a market for water quality in Australia).
264. For discussions of the discounting issue, see Daniel A. Farber, From Here to Eternity, 2003 U. ILL. L. REV. 289, 332 (contending that "exponential discounting does not ultimately rest on an illegitimate privileging of the present over the future"); Lisa Heinzerling, The Temporal Dimension in Environmental Law, 31 Envtl. L. Rep. (Envtl. Law Inst.) 11,055, 11,067-71 (2001) (arguing that a temporal analysis "undermines both the practicability of discounting and its theoretical underpinnings"). Over longer time periods, the results of changes in discount rates are even more dramatic, as Cass Sunstein explains: If an agency chooses a discount rate of 2%, the outcome will be very different from what it would be if an agency were to choose a discount rate of 10%; the benefits calculation will shift dramatically as a result. If a human life is valued at $8 million, and if an agency chooses a 10% discount rate, a life saved 100 years from now is worth only $581. "At a discount rate of 5%, one death next year counts for more than a billion deaths in 500 years." Cass R. Sunstein, Cost-Benefit Default Principles, 99 MICH. L. REV. 1651, 1711 (2001) (quoting DEREK PARFIT, REASONS AND PERSONS 357 (1984). On the central role of discounting in current disputes over climate change among economists, see Daniel H. Cole, The Stern Review and Its Critics: Implications for the Theory and Practice of Benefit-Cost Analysis 29-35 (2007), available at http://ssrn.com/abstract=989085 (highlighting the importance of the discount rate and evaluating the 0.1% pure discount rate used in the U.K. Treasury's cost-benefit analysis).
265. See Avinash Dixit, Investment and Hysteresis, 6 J. ECON. PERSP. 107, 110-13 (1992) (mathematically illustrating the effects of waiting for new information in decision making). The basic point is that "[wjhere there is uncertainty, there may be learning." W. Kip Viscusi & Richard Zeckhauser, Environmental Policy Choice Under Uncertainty, 3 J. ENVTL. ECON. & MGMT. 97, 108 (1976).
266. See Dixit, supra note 266, at 116 ("[C]urrent revenues have to rise to nearly double the level that ensures a positive net worth before waiting ceases to be optimal."); see also id. at 117, 120 (providing other examples of the magnitude of hysteresis effects). Sometimes, we may be uncertain about the degree of irreversibility itself, and here too the possibility of learning must be taken into account. See Viscusi & Zeckhauser, supra note 266, at 107-08 ("When confronted with real policy choices, we should attempt to learn about the values of the parameters that describe the likelihood that the potential irreversibilities will occur, and the consequences when they do.").
267. See Graciela Chichilnisky & Geoffrey Heal, Global Environmental Risks, 1 J. ECON. PERSP. 65, 76-79 (1993). Chichilnisky and Heal argue the importance of waiting for more information: [B]enefits will accrue in the future from the preservation of a resource, but these are currently unknown. If the resource is preserved into the future, then in the future the decision about whether to preserve it can be reconsidered in the light of better information then available about the benefits from its existence. If it is not preserved, then there is no chance of reconsideration when we have better information. Id. at 77. See also Anthony C. Fisher & W. Michael Hanemann, Option Value and the Extinction of Species, in 4 ADVANCES IN APPLIED MICRO-E CONOMICS: RISK, UNCERTAINTY, AND THE VALUATION OF BENEFITS AND COSTS 169, 184-85 (V. Kerry Smith ed., 1986) (concluding that where a decision to develop in the first of two periods is irreversible and where the benefit of choosing to develop in the second period is unknown at the start of the first period, the expected benefit of a decision to wait is greater where information about the value of such benefit becomes available at the start of the second period).
268. On this issue and its potentially large effects on sea-level change, see Robin E. Bell, The Unquiet Ice, SCI. AM., Feb. 2008, at 61, 67 (observing that current climate models do not consider "such major features as ice streams, and none of them incorporates an accurate representation of the bottom of an ice sheet," so "predicting future sea-level change from the current climate models greatly underestimates the future contribution of the polar ice sheets to sea-level rise").