The Honorable George E. Brown, Jr. Committee on Science House of Representatives Dear Mr. Brown: I have enclosed our response to the two additional questions you posed in your January 18, 1996 letter pertaining to the Committee's November hearing on global climate change models. I appreciate the opportunity to be of assistance on this matter. Sincerely, Peter F. Guerrero Director, Environmental Protection Issues Enclosure Enclosure Responses to Questions For the Record 1. Dr. Michaels suggested in his testimony that climate modelers have not been straightforward in disclosing the uncertainties in their models. Did the GAO find any indication that modelers have failed to appropriately disclose model uncertainties and potential weaknesses? No. During our review of climate change models in the United States, the officials were generally quite candid about the model limitations and areas needing improvement. These shortcomings were also documented in numerous studies which they made available to us. Because of the limitations, the modelers attached levels of certainty, based on their scientific judgment, to various projected outcomes (e.g., global temperature increase and sea level rise) when they published the results of their Modeling Forum. We did not examine models from other countries as part of our review and therefore cannot comment on the extent to which uncertainties have been disclosed from such models. 2. In your opinion, did the USGRP Modeling Forum fairly consider and discuss the views and criticisms of "skeptics" such as Dr. Michaels and Dr. Lindzen who participated in the forum? The USGCRP purposely invited Dr. Michaels and Dr. Lindzen to the Modeling Forum because they were known to be skeptical concerning global warming and the general interpretation of results from climate change models made by other scientists. GAO staff who were present at the Forum observed an open exchange and consideration of views and ideas from all participants although full agreement was not reached on all issues. As a result, the published account of the Forum could be best described as representing the consensus view of the majority of the scientists present. I want to express my appreciation for your participation in the Committee's recent bearing on Global Change modelling. Your testimony was of great value to me and to the Committee. Enclosed, you will find some additional questions which are intended to clarify certain points raised in the hearing and develop additional information for the Committee's use. Your written responses will be included as part of the bearing record. I ask that you provide your responses by February 15. You may contact Dr. William S. Smith of my staff at 202/225-4439 if you have any questions regarding this request. ADDITIONAL QUESTIONS FOR THE RECORD FOR DR. MAHLMAN 1. During the course of the hearing, Dr. Michaels provided a chart which purported to show that a key model used by the IPCC did not provide an adequate representation of temperature data in the 5,000-30,000 ft layer gathered by satellites over the past 20 years. Please provide your own views on the methodology and interpretation used in this chart and its signifigance, if any, to the validity of IPCC models. Some points you may wish to consider include the following: a. Is Dr. Michaels' description of the satellite temperature records in the Southern Hemisphere accurate? b. How would you characterize the model projections of temperature changes in this part of the atmosphere versus the satellite temperature records from the Northern and Southern Hemispheres? c. Do these satellite records of temperature change coincide with the surface measurements of temperature change? Are the two sets of records inherently different or can they be reconciled? d. How accurate have the satellite measurements of temperature been since 1979? Have they improved over time? 2. Dr. Michaels stated that there is an even greater discrepancy between model projections of warming and observed warming in the Northern Hemisphere than in the Southern. He further stated that these climate models project very large warming in the polar region, yet the observed warming has been only slight and occurs prior to 1940, which is before greenhouse gas concentrations changed very much. Since then, there has been very little observed change, according to Dr. Michaels. a. Would you agree with Dr. Michaels that the northern latitude warming occurred before 1940, before the greenhouse effect had changed very much? What reasons are proposed for these latitudes not warming during recent decades? b. Is the temperature record that Dr. Michaels referred to complete enough for reiiable analysis? c. Has there been an observed warming in the northern latitude temperature since !940? 3. Dr. Michaels stated that John Mitchell's new climate model suggests a new, lower estimate of climate sensitivity to a doubling of CO2, i.e., a 1.0-1.5 degree C warming as opposed to the earlier projection of 1-4 degrees. Do you agree with this assessment? What are the reasons that the IPCC projections of global warming appear to have come down somewhat? 4. During the course of the hearing, Dr. Michaels provided a chart which purported to show that all of the temperature change from 1965 to 1994 occurred in one year, a feature which models cannot predict. Please provide your own views on the methodology and interpretation used in this chart and its significance, if any, to the validity of climate models. Some points you may wish to consider include the following: a. Do you agree that this single jump explains the trend in the entire record, or is this an artifact of the method used to present the data? b. Is it reasonable to analyze short temperature trends over unequal and very short segments of time? Are such analyses statistically valid? c. Do you agree with Dr. Michaels assertion that unless we can understand things like the temperature jumping upward in 1976 we do not really understand this entire climate problem? 5. Dr. Michaels testified that the climate models most heavily cited by the IPCC 1992 supplementary report on climate change were known to contain large errors at the time of adoption of the Framework Convention on Climate Change and that such errors were not disclosed with the result that the model's uncertainties were not considered in the debate surrounding this issue. Please respond to this statement. 6. Dr. Michaels testified that prior climate models in 1990 projected erroneously high estimates of warming because computer limitations required modelers to assume that carbon dioxide concentrations doubled immediately, instead of gradually. Since that time, models have incorporated a gradual increase of carbon dioxide concentrations. What differences have been seen between the 1990 model projections and the 1992 projections in terms of the range of temperatures projected out to 2100, in a CO2-doubled world? 7. Dr. Michaels testified that the "so-called skeptics" have been saying for the last seven years that by the year 2100, with a doubling of carbon dioxide, the net warming would be 1.0 to 1.5 degrees C, a smaller increase than is projected by early climate models. a. What was the quantitative basis on which the “skeptics” based such projections of net warming, prior to recent models which incorporate the offsetting influence of aerosols? b. Did the "skeptics" also conduct a series of model projections? Have the skeptics been able to construct a global climate model that, based on rigorous parameterizations, reproduces their projected results? 8. In response to a question from the Chairman, you agreed that projections are based on unvalidated assumptions. Is there any scientific procedure to "validate” assumptions about the future? If not, does the inability to scientifically validate such assumptions render such them useless for the purposes of making projections? Do other models - such as economic models used to project the impact of fiscal or monetary policy -- rely on similar unvalidated assumptions? If the lack of scientifically validated assumptions precludes the use of projections, what other tools are available for policy makers to understand and weigh the future consequences of policy options on complex systems? 9. Dr. Michaels testified that in 1992, in association with the signing of the Rio Treaty, Congress was knowingly misled by witnesses who withheld information regarding known errors in key |