|National Assessment Climate Action Report 2002: Chapter 6 Response to Public and Agency Comments General Comments|
Note: The general comments are numbered and ordered in alphabetical order of the contributing organization or individual, with identifications given at the end of the comment. A response to each comment follows the submitted comment. Where several general comments were provided by a single organization, they are numbered separately to allow a response appropriate to each comment. Where a comment is presented over several paragraphs, it is lumped into a single comment. Where sections of submitted comments do not have a bearing on matters considered in Chapter 6, they have been deleted or ignored. See full listing of submitted comments.
Figure and table numbers as well as page and line numbers in the Chapter are indicated with reference to the version of Chapter 6 made available for public comment . Note that the revised version includes numerous changes, so the page and line numbers cannot be used to reference the final version of the chapter.
General Comment 1
The American Petroleum Institute (API) is a national trade association representing 400 companies engaged in all aspects of the oil and natural gas industry. We are pleased to submit comments to assist in preparing the Third U.S. Climate Action Report in response to the November 15, 2001 Federal Register request for public comments (66 FR 57456).
In many respects, the draft Third U.S. Climate Action Report, being prepared in accordance with U.S. obligations under the UN Framework Convention on Climate Change (UNFCCC), provides a constructive report on U.S. climate action and reflects the National Research Council's (NRC) 2001 review of climate change science. The draft Climate Action Report identifies many of the difficulties in evaluating the potential for climate change and the resulting positive or negative impacts, as well as the implications of these important uncertainties for the development of a constructive climate policy.
Particularly important for a report to the UNFCCC covering the potential impacts on the US, the current draft generally reflects a critical NRC conclusion that “one of the weakest links in our knowledge is the connection between global and regional projections of climate change -- (Chapter 1, page 6, lines 36-37). As discussed in the attached Specific Comments, however, parts of Chapter 6 go beyond this “weakest link in our knowledge -- in projecting regional climate change impacts, and these sections should be revised.
The draft report provides very constructive observations on the apparent variability of climate over the past century and, importantly, the ability of our economy and our society to respond to apparent climate changes or climate variability through adaptation. As noted in the draft report, our ability to adapt depends critically on the strength of our economy. Past behavior demonstrates that adaptation is a realistic and important element in any long-run response to the issue of climate change, and the draft U.S. Climate Action Report's discussion of adaptation is constructive and appropriate.
If you have any questions regarding API's attached specific comments, please feel free to contact me or Russell Jones (202-682-8545 or email@example.com).
American Petroleum Institute
Response: Although limits do exist in projecting how changes in the regional climate will differ from the global and latitudinal projections of climate change in which there is more confidence, these limitations were taken into account in the US National Assessment's analyses. In addition, the geographic features of North America and historical meteorological patterns provide a reasonable basis for making some inferences about the potential large-scale regional patterns of climate change. Even in the absence of detailed regional projections, it is also possible to consider the potential sensitivities and vulnerabilities of the environment and its services and of society and its activities to potential climate change as a means to gain insight about how to enhance resilience and reduce vulnerability to climate variations and change. The chapter already clearly indicates that limits exist to what can be said. Additional explanation has been added concerning uncertainties and to indicate that the scenarios are not predictions, but plausible projections.
General Comment 2.
General Comment on Chapter 6: Overarching Issue: As the text of the Climate Action Report is currently drafted, Chapter 6 is the longest chapter and includes extensive discussion of a wide variety of regional impacts. And while the chapter provides some very useful information on changes during the 20th century and the role adaptation has played and may play in the future, the current extended emphasis on future regional “impacts -- listed in the Clinton Administration's report, the U.S. National Assessment, is inconsistent with a key conclusion of the National Research Council study on climate change, namely, “[the] current analyses are unable to predict with confidence the timing, magnitude, or regional distribution of climate change…. --
The NRC is not alone in concluding that current regional impact assessments are unreliable. According to the IPCC Third Assessment Report (WG III, page 96). “[T]here is considerable uncertainty about the rate of expected change and its manifestations and impacts at the regional and global levels. Science cannot predict the climate and its impacts in Milwaukee, Mumbai, or Moscow half a century ahead very accurately, and it may never be able to do so. -- The conclusions are the same about changes in the oceans. According to the IPCC (WG II, page 352-353), “[I]t is still not possible to assess regional responses to shifts in climate trends, and it is unknown if a general warming will increase or decrease the frequency and intensity of decadal-scale changes in regions where national fisheries occur. --
In fact, important elements in the draft report provide very good examples of the NRC's concern. For instance, Figure 1 illustrates how the Illinois climate might change, but the answer is completely model dependent. As shown in current draft Figure 6.1, according to the Canadian model the climate in Illinois becomes more like that of states South-West of Illinois. However, according to the Hadley model, the climate in Illinois becomes more like that of states due-East of Illinois. Directionally, the models give effectively opposite results.
Similarly, Figure 6.2b indicates that for roughly three-quarters of the regions of the country, the Hadley and Canadian models predict opposite changes in daily precipitation. And for the few regions in which both models project increases in precipitation, there is no agreement as to the amount of the change.
Given the many divergent results from the Hadley and Canadian climate models, it means little that in some circumstances these two models indicate similar impacts. The highlighting of such circumstances was one of the problems with the U.S. National Assessment report. What is important is that both the NRC and the IPCC conclude that current models are unable to predict with confidence the nature of regional climate changes.
Overall, it is important that the Third US Climate Action Plan, which will be submitted to the UNFCCC as an official report in accordance with our treaty obligations, reflect the best available science and not contain a broad listing of specific regional impacts in which there is little confidence.
To achieve this end, Chapter 6 should be significantly shortened and structured around the NRC conclusion. While many of the categories of impacts covered in Chapter 6 might remain to demonstrate the broad and comprehensive U.S. approach to the climate issue, the individual discussions should be significantly shortened. At the same time, the current draft of Chapter 6 contains much good material. The remaining comments on Chapter 6 offer suggestions on how the chapter might be shortened.
American Petroleum Institute
Response: We agree with the National Research Council's (NRC) conclusion that present climate models cannot project regional patterns of potential changes in climate with confidence. We would note, however, that this does not mean that regions will not experience climate change nor that climate models cannot provide some indications of the types of changes that are possible or likely to occur in different parts of North America, even if not with the high degree of confidence that might be desired. While the Assessment could have assumed future changes over the US would have been equal to the global average, both the IPCC and NRC agree that the projected warming will be larger over higher latitudes (and the US reaches to those higher latitudes) and will be larger over land areas than over the oceans (and the US is a land area). Both of these findings suggest that it is therefore appropriate to be considering the latitudinal and continental scale patterns and magnitudes of potential climate change. In addition, as reported in the IPCC Working Group I Third Assessment Report (page 600, Box 10.1, Figure 1), the model results considered by the IPCC are consistent in projecting (except over Alaska in summer for case B2) that the warming will be greater than the global average over all continental areas of the US for both summer and winter and for relatively high (case A2) and relatively low (case B2) emissions, except for Alaska in winter, which the models agree will experience “much greater than average warming. -- While the temperature projections of the models are in general agreement, it is true that the model projections of regional changes in precipitation show significantly less agreement (page 601, Box 10.1, Figure 2), except in Alaska and during winter in western and eastern North America.
To accommodate the range of possible results and to remain consistent with the IPCC and NRC findings that regional-scale predictions (and even projections) cannot be made with confidence, the USNA adopted a three-pronged approach in seeking to understand the potential impacts of climate variability and change and the possibility for considering adaptive measures. Although it has not yet been possible to fully implement this strategy, the findings of the USNA were considered with this three-pronged approach in mind. First, analyses considered what the potential impacts would be if climatic conditions (including extremes) that occurred during the 20th century recurred (this approach was basically equivalent to assuming minimal climate change). Second, a minimum of two differing climate scenarios were used by investigators to try to capture at least some of the range of future possibilities of climate change; that the two primary scenarios had different projections of future changes in, for example, precipitation was helpful in determining which potential changes are likely to pose what types of problems. Use of model scenarios rather than arbitrarily selected changes ensured internal (i.e., meteorological) consistency among variables (e.g., since evaporation of soil moisture helps to limit temperature increases, it would not be internally consistent to consider cases with large increases in both precipitation and temperature). Third, some groups were also able to examine how large changes would need to be to cause an unusual amplification of potential impacts of a particular type (i.e., considering whether a change would led to crossing a threshold or exceeding the likely adaptive capacity). With an indication of where high sensitivities existed, the notion was to then draw upon the paleoclimatic and other indicators of past changes in the climate to get a sense of the potential likelihood of such an occurrence (e.g., although the two primary models projected that the western US would experience an increase in precipitation, the analysis for the western US also considered what would happen if prehistoric droughts were to return).
Unfortunately, this multi-pronged approach to gaining a more robust understanding of potential vulnerabilities of ecosystems and society has apparently caused some confusion among those (including, apparently API) that seem to be imagining that the USNA was attempting to make a specific prediction of future changes on a regional basis. This confusion, in turn, seems to have led to criticisms that are aimed at something that was not attempted -- namely, the USNA specifically did not attempt to make specific predictions, much less specific regional predictions (see, for example the USNA discussion of this on pages 4-5 of the Overview report). Nor did the USNA findings rest solely on model results -- the consideration also reflected changes that had occurred in the past. In addition, this particular chapter for the CAR presents mainly the more robust and more qualitative aspects of the USNA findings, which were carefully considered by the NAST and were not particularly dependent on the specific outcomes of climate models; as such, the results are particularly robust.
The specific API comments are considered below in the context of this response to their general perspective, which seems in conflict with the approach actually taken by the USNA and in this chapter. To make sure that these issues are even better understood, the opening parts of the draft chapter have been rewritten and clarified, and a new introduction now more fully explains the issue of uncertainties.
General Comment 3.
The greatest shortcoming of this document is the nearly complete lack of attribution for data, analysis and conclusions. The area of climate change science is rich with disagreement regarding future projections when it comes to specific predictions. This document fails to treat this appropriately by citing the origin of the findings it includes. As a result, no reviewer can adequately investigate the interpretations. There is nary a paragraph for which this is not the case. Especially egregious is the use of phrases like, “Model-based projections for the 21st century indicate…, -- and “Model projections suggest…. -- with no clarification as to which models are being discussed.
Climate Action Network
Response: This chapter is not intended to actually be a fully documented assessment of potential impacts and adaptation, but is rather intended to draw upon the leading widely reviewed studies of the IPCC and USNA to provide a summary of the more likely impacts. We do note that this is not made clear in the opening section of the chapter and the chapter has been amended to note this so that those seeking to find more information are quickly referred to the reports of the IPCC and USNA that contain more complete documentation.
General Comment 4.
Great efforts seem to be made to always present the advantageous conclusions first whenever two or more outcomes are possible. This is done even when the positive outcome is the least likely. For example in a discussion of forest fires, the idea of trees with thicker bark being more resilient is explored before the more likely loss of species and populations. However this approach also suffers from unconvincing and unsupported presentation (see comment above).
Climate Action Network
Response: The intent was to present a balanced presentation, considering first the more likely impacts and then ones involving a sequence of potential occurrences. Particular sections drew heavily on the USNA Overview report. To the extent that less likely outcomes are featured, this was unintentional and has been addressed in the revision based on the specific comments that were made.
General Comment 5.
Many important habitat types are ignored, including deserts, freshwater systems, arctic/tundra, islands and alpine regions. There are some of the most sensitive systems and they are brushed aside for drawn out discussions of forestry and agriculture. Also missing is the subject of migratory species and non-tree, non-agriculture plant species.
Climate Action Network
Response: While the chapter does include sections on the USNA's sector and regional findings, it is true that a section on the potential impacts on ecosystems generally was omitted. This was inadvertent and resulted because the USNA considered the ecological changes as part of the scenarios rather than as a separate sector. To remedy this, a separate section has been added that provides a summary of findings regarding potential climate interactions with land cover.
General Comment 6.
The forest section seems to ignore public forests. Given the large amount of area covered by public forests in the U.S. there will be a need for adaptation strategies for these forests and it may be Federal gov't money, not market money that pays for it. Some discussion of these lands seems to be needed in this document.
Climate Action Network
Response: While it is certainly true that public forests will be impacted along with private forests, understanding the potential differences in adaptation strategies for forests of different types and ownerships, in different parts of the country, and providing different ecological services remains to be done. Such studies will likely need to be done at the local level, and to promote such efforts, the USFS, in addition to sponsoring the forest sector assessment and ensuring publication of its results in widely available journals, has distributed the USNA Overview report and a summary brochure on the forest sector assessment to all forest managers; similarly, the DOI has, for example, distributed the USNA Overview report to National Park Service ecosystem resource leaders at each national park. What is important to recognize is that the Nation's public lands need to be managed in the context of multiple actual and potential stresses and services (many of which are local), and what will be important is for the management plans to start incorporating consideration of actual and potential changes in climate along with other factors. The distribution of reports to land managers is likely a much more effective strategy for enhancing consideration than attempting to generalize from many specific situations and add a sentence or two in this chapter.
General Comment 7.
The overall focus on the idea of “market solutions -- seems to miss all of the natural resource goals. There is no market incentive to protect most natural resources. Does this mean that they are without value? Emphatically, “no. -- So to protect natural resources from the ravages of climate change we must adopt the “contaminant paradigm. -- In this approach we have realized that the only way to protect natural resources from the impacts of anthropogenic contaminants is to impose regulations and clean-up measures to limit or correct damage. The same approach must be taken with regard to climate change. Greenhouse gases are anthropogenic in source. We must limit their emissions to stop further impacts and help systems adapt to impacts that have already occurred or are committed to occur. For example, market solutions would not have removed lead from gasoline. Nor would they have cleaned up contaminated sites.
Climate Action Network
Response: The added section (mentioned above) dealing with potential climate change interactions with land cover indicates, as the USNA did in its findings (and as this chapter briefly noted in Table 6.2, point 3), that adaptation options are quite limited, and in some cases non-existent, for natural ecosystems. Other matters raised in this comment relate to materials presented in other chapters of this submission.
General Comment 8.
It is important to recognize the many intricate links between human success and the protection of natural resources and systems. This document seems to ignore those links and focus almost exclusively on the direct economic extractions ignoring the crucial support structure that our natural world provides.
Climate Action Network
Response: As indicated above, a separate section on potential climate interactions with land cover has been added.
General Comment 9.
Overall the chapter is heavily weighted towards beneficial impacts of climate change and towards minimizing the importance of potential risks involved. Especially telling is a bias towards assuming that the lower end of a temperature projection is more likely to be true than the higher end when no evidence is given as to why that should be the case. The summary statements at the beginning of the chapter and at the beginning of the sections are often misleading, emphasizing beneficial aspects and either minimizing or calling into question the confidence scientists have in the negative aspects. In addition, while it is assumed that the US will be able to adapt to any changes, no analysis of the potential costs or feasibility of the adaptation is described. Some discussion of the more negative aspects is included in the body of the chapter though not reflected adequately in the summary portions.
Climate Action Network Two
Response: We believe the chapter as presented is generally in agreement with the IPCC and USNA reports which suggest that, given their resources and structure, developed countries will generally be more able to adapt to climate change impacts than developing countries. Addition of the section on potential climate interactions with land cover should make clearer that not all potential impacts can be adapted to, which may help to address at least some of the bias perceived by CAN2. Mention of land cover effects in the chapter Overview should also tend to alleviate the perception that potential negative influences need to be more prominently mentioned to ensure balance.
With respect to the suggestion that the report is considering the lower end of potential climate impacts, the two primary climate model scenarios represent quite well the range of findings of the wider set of models considered by the IPCC, and the results reported are drawn from the USNA analyses.
With regard to the limited discussion of how difficult and costly adaptation may be, the NAST noted that there is very little information on the feasibility and costs of potential adaptation options and called for more research in this area. Without additional information, we agree with the NAST that attempting to determine potential costs in an integrated monetary sense is not possible at this time, that additional options may be identified, that timing will surely affect costs, etc.
General Comment 10.
Thank you for giving us the opportunity to comment on the draft Third National Communication which is being prepared for submission to the United Nations Framework Convention on Climate Change (UNFCCC).
The communication presents an appropriate response to matters related to greenhouse gas emissions and reductions. However, it does not go far enough in discussions of adaptive actions that address climate change being taken by the U.S. We believe that the Executive Summary and Chapter 6 should be modified to broaden the discussion pertaining to adaptation.
The Executive Summary (ES) fails to convey the scope of activities being undertaken by the U.S. that address climate change. The U.S. is engaged in a number of activities that would help address urgent problems of the world that might be made worse due to climate change. Accordingly, these activities would provide more or less immediate benefits while helping reduce vulnerability and enhance adaptability to any future adverse impacts of climate change. Identifying and highlighting these activities would illuminate -- and enable other nations to better appreciate -- the breadth and depth of the U.S. response to climate change. It would also provide leadership in expanding the set of affordable options available to the world community today for addressing the serious issue of climate change.
For example, the potential spread of malaria (a disease which annually kills over 1 million people worldwide) has been proposed as a consequence of climate change, and has been used by some to justify immediate limits on greenhouse gases. The ES is silent on the efforts that have been undertaken by the U.S. government and philanthropies to address this critical issue. Such research includes development of vaccines, and genetic mapping of both the mosquitoes and parasites responsible for this disease. This research could provide benefits in relatively short order even if climate doesn't change, as well as in the long term, should climate change increase the likelihood of malaria.
Similarly, numerous activities presently being undertaken to increase agricultural productivity will help the U.S. respond to the effects of climate and global change. Examples include increasing productivity in poor soil and climatic conditions which might become more prevalent in some areas should climate change; increasing productivity while lowering reliance on synthetic fertilizers and pesticides; reducing water usage for agriculture; and shifting to no-till agriculture. Such activities will help U.S. agriculture to continue to increase productivity to feed the growing human population while at the same time conserving and managing natural resources.
Aggressive actions such as those outlined in the above examples will help solve current known problems that might be made worse by future climate change. They will benefit both the U.S. and the rest of the world. In fact, much, if not most, of these efforts would provide major benefits for developing countries. These existing efforts could also be used to engage developing countries more fully in a dialogue with respect to climate and global change.
Taken as a group, such actions by the U.S. also constitute a coherent and aggressive approach to reducing vulnerability and enhancing adaptability that would expand the set of policies available to deal with climate change. Cumulatively, such actions might increase the level at which atmospheric concentrations of greenhouse gases would be deemed to be “dangerous -- (to use the UNFCCC's formulation of its “ultimate objective -- ). Thus, reducing vulnerability/increasing adaptability would reduce the depth of future emission reductions, and the speed with which they might have to be implemented. These factors, acting together, could result in substantial savings in eventual mitigation costs.
However, at present, the international and U.S. focus is on emission limitations/carbon sequestration, supplemented by research. With such a focus, the only viable options seem to be "Kyoto", "Kyoto lite," or "Kyoto later", supplemented by research. The U.S. should exercise leadership in broadening the menu of policy options to include aggressive adaptation efforts. The current version of the National Communication misses an opportunity for education as well.
Moreover, we see significant omissions in the factors that ought to be considered in devising (or identifying) effective and efficient approaches to addressing climate change. Among the factors we believe ought to be explicitly noted are the following: global benefits ought to exceed costs; actions that provide immediate benefits should be preferred over those whose benefits are delayed; and whether actions are effective with respect to the environment and public health, economically efficient and sustainable, fair, and help solve the urgent problems of today while enhancing our ability to address the problems of the future.
Finally, with respect to the ES, in discussing (on page 4 and elsewhere) U.S. energy consumption and emissions, we would add more context. Specifically, we would note that while our overall energy use is high, our energy use per unit of output, i.e., the energy efficiency of our economy, compares relatively well with the rest of the world.
Department of the Interior
Response: Most of these comments pertain to suggested changes in the Executive Summary of the National Communication. Regarding the suggestion that more needs to be included about potential adaptive measures being taken by the US that might help other countries to better adapt to climate change, we would note that this chapter does not seem to be the place for such discussions. Based on guidance from the UNFCCC, this chapter is supposed to focus on potential impacts and adaptive measures of climate change affecting the US. In that the US does indeed depend on other nations for many vital products and services that might be impacted by climate change (e.g., agricultural products, containment of infectious and vector-borne diseases, etc.), the chapter already includes some indications of how changes outside the US might impact vital sectors within the US, but were the chapter to cover the suggested range of efforts on adaptation measures (done mainly now for other reasons), it would seem that the chapter would also need to summarize the potential impacts for all nations and sectors around the world, and this is clearly beyond the scope of the chapter.
General Comment 11.
Also, we are skeptical of the methodology used to characterize the levels of confidence that are attached to various potential national and regional impacts of climate change for the U.S. The current version uses terms such as "possible," "likely" and "very likely" in association with various potential impacts. The science behind climate change -- or more importantly, its impacts -- are not known with sufficient rigor to allow us to predict what impacts are likely or very likely to occur, especially in various areas of the US. In most instances, at best we can identify what is possible.
The methodology used to distinguish the possible from the (supposedly) "likely" or "very likely" is quite subjective, and seems to ignore that our projections are uncertain not only with respect to various measures of climate change (e.g., temperature, precipitation) but also with regard to our limited ability to model their impacts on species, ecosystems, societies, and so forth. We are also skeptical that one can give much credence to a projection which is not necessarily borne out by historical trends, as seems to be the case for some of the projections. For all these reasons, we urge replacing the “likely -- and “very likely -- by “might, -- “may, -- or “could. --
We are attaching more detailed comments on the Executive Summary and a marked-up copy of Chapter 6. However, with respect to the latter, please note that we have not scrubbed the document of all the instances where “likely -- or “very likely -- , etc., appear, which we believe ought to be done.
If you have any questions on these matters [comments 10 and 11], please contact Indur Goklany at 202-208-4951.
Department of the Interior
Response: Two matters are raised by these comments.
First, follow-on discussions with DOI indicated that the chapter was not sufficiently clear that the uncertainties being described were not being chained together from emissions to climate response to impacts. Were this the case, it would indeed be difficult to suggest impacts with high probability. To clarify what has been done, however, a new Introduction has been prepared that recognizes the full range of uncertainties, but makes clearer that the analysis here is based on potential consequences in the event that a specified climate change occurs. With this approach, a number of potential impacts can be identified with greater certainty.
Second, the issue of how best to report scientific findings in reports to policymakers and the public has been the subject of wide-ranging consideration over the past few years. One tradition common among physical scientists has been not report anything as likely (or even possible) until a high degree of statistical certainty has been achieved (e.g., two-standard deviations or 95% likelihood, etc.), and this has been reflected in statements in reports to the effect that ‘some outcome cannot yet be predicted with confidence' and so one can only say it “might -- or “may -- happen. Among the general public, this has often led to a misinterpretation that therefore the potential outcome is simply uncertain, and that little is understood about the potential outcome when, in fact, much is known, even though statistical certainty cannot be achieved (due to, for example, a short record or the absence of observations, etc.). Discussion in the literature of this issue has then recognized that the traditional, “hypothesis-testing -- approach used by many physical scientists is not the only way, and likely not even the typical way, that policymakers (in both the public and private sectors) consider issues when faced with decisions where the information base is not definitive. For example, in many cases, a relative risk or relative likelihood approach is used, where the results are presented in terms of the more likely outcome based on both data and expert judgment. In addition, discussions have pointed out that interpretations will also depend on such issues as how important and or costly the matter may be, its effect on health, whether an outcome to be experienced is voluntary or involuntary, etc. In any case, the results of these many discussions led to a recognition by the IPCC and for the USNA that the use of vague terms such as “might -- and “may -- by scientists was leading to significant misunderstanding of the potential likelihood of outcomes. For example, the EPA Science Advisory Board did a study a few years ago indicating that readers (even expert readers) gave estimates ranging over several orders of magnitude of potential likelihood when vague terminology was used.
In addition, when terms such as “might -- and “may -- are used, a much fuller range of outcomes (e.g., ranging from 1% or less to 99% or more) must be presented in order to be appropriately self-consistent. When such an approach has been used in the past, this has led some readers to draw very different conclusions -- from worst case to an overplaying of uncertainties. As one example, an early draft of the USNA using the terms “might -- and “may -- led a prominent newspaper columnist to come to conclusions that NAST members considered at significant variance from their intended meaning.
Because of unfortunate experiences with misinterpretations, both the IPCC and the USNA assessment teams developed a lexicon that assigned a relative likelihood to various terms. As part of the IPCC process, a special paper on the subject was prepared in an attempt to encourage uniform terminology across its various WGs. Although not all WGs followed the suggestion, WGs I and II both did adopt a lexicon that assigned specific likelihoods to specific words (e.g., in WG I, “likely -- meant a 66-90% chance, “very likely -- meant 90-99% chance, etc.) and in their reports they very prominently define the terms that they use. Discussions as part of the USNA development process led initially to use of a similar well-defined lexicon in the draft distributed for public comment. In response to a well-argued criticism of the association of specific numerical ranges with specific words, the NAST modified the presentation of its lexicon to recognize that, as this comment notes, these words cannot be used with numerical precision. Instead, the USNA's final lexicon presented the terms as a general ordering of relative likelihood. To make this clearer, the NAST also added significant text to explain its approach and to indicate that their judgments were based on knowledge drawn from scientific studies, theoretical analyses, consideration of the internal consistency of the results, and their personal expert judgment (see page 5 of the Overview report).
In view of this extensive history and of misinterpretations that have resulted from previous uses of terms such as “might -- and “may, -- this comment is not accommodated in the revised version. However, extensive explanatory text has been added, including a new Introduction to the chapter that more fully explains the approach. In addition, follow-up discussions with DOI indicated that a key change that was needed was to make sure that the explanation indicated that the lexicon was referring to relative likelihood or plausibility rather than necessarily referring to an absolute indication of what would happen. Such text has been added, and specific uses of the lexicon in the text have been rechecked.
General Comment 12.
The “2001 U.S. Climate Action Report, -- particularly the section titled “Climate Change Interactions with Coastal Areas and Marine Resources, -- was reviewed. The report is a good evaluation of the threat associated with global warming, specifically sea level rise. The predictions in the report are very consistent with other predictions on sea level rise and are similar to those in a recent report on research done under the International Geosphere-Biosphere Program (IGBP) summarized in the October 30, 2001, issue of “EOS, -- a publication of the American Geophysical Union. The article was written by John Anderson and others.
Response: No response required.
General Comment 13.
The discussion of impacts of climate change in this chapter is based on the US. National Assessment in the 2000 NAST Report. This report remains highly controversial, and was, at that time, the subject of considerable debate with regard to it scientific and policy discussions. The most contentious aspects of the report were (1) the use of the most pessimistic projects of potential climate impacts, with little or no discussion of the uncertainties involved in such projections; and (2) the use of global climate models to predict impacts on regional climates within the US.
Global climate models cannot be used to project changes in regional climate, primarily because of insufficient spatial resolution and their inability to represent regional phenomena such as clouds. As was pointed out when the National Assessment was in review, the regional projections from the two climate models used often provide diametrically opposed projects for regional climate, making it impossible to determine which (or either) is (are) correct. Further, attempts to average the results from the two models are also without scientific basis.
Thank you for your attention to these comments. Please contact Peg Gutmann at 313-594-0400 if you have any questions.
Ford Motor Company
Response: Many of the issues raised by this general comment are covered in the responses to the comments by API and GM. Fundamentally, there seems to be a fundamental misunderstanding in the purpose of the USNA -- it did not attempt to or claim to “predict -- future climate changes on a regional basis (although see the response to comment 2 above indicating the models actually show significant agreement on some aspects of the continental patterns of future climate change), but rather the USNA used several approaches to developing plausible scenarios of what could happen and to then examine the potential impacts that would be likely (or unlikely) to result. As just one example, there is universal agreement that warming will occur -- experience, model results, and logic then lead to an indication that the snow line will be higher in the western US; this in turn will cause impacts to water resources of a magnitude that will require adjustments. This is just one example of how a large-scale result of climate models can be used to investigate and consider a potential regional-scale impact, even though the models do not provide precise information about how much change will occur by exactly when. Ford's comment is roughly equivalent to suggesting that Ford should not make any car until it is sold to a specific individual who has paid every penny -- a business approach that would be unworkable. Although there are uncertainties, individuals and companies work with scenarios all the time in planning their lives, what they produce, purchases, etc. Examples of how scenarios are being used by the public were reported, for example, in a Washington Post article on December 30, 2001 (page F01, “What To Do With Your Life -- ). The chapter text now makes even clearer that the model scenarios are not predictions but are projections.
Ford also fails to explain its claim that this report remains highly controversial, especially with respect to the final version that was revised in response to the public comments. It should be noted that the IPCC, for example, incorporated Assessment findings and subjected those that it included to wide review without finding problems. Similarly, the National Research Council in its report for the President in June 2001 relied on findings of the National Assessment. Points made in other of the suggested critical references seem to rest on misconceptions that are explained again in response to comments submitted in this review. Thus, we do not agree that the National Assessment, when the final version is read in its entirety, is highly controversial.
General Comment 14.
General Motors Corporation respectfully submits these comments in response to the invitation in U.S. EPA's request for public comments (66 FR 57456) on the “2001 U.S. Climate Action Report. -- The 2001 U.S. Climate Action Report provides a comprehensive and objective overview of U.S. national circumstances, and in-depth descriptions of U.S. efforts to address concerns about climate change. As noted in the Report, the U.S. has many efforts underway to address climate change and much progress is being made in reducing the carbon intensity of the U.S. economy. This progress will continue with the sound policies and new initiatives the Administration has announced for addressing climate change in the years and decades ahead.
Most importantly, the U.S. has set forth a set of sound principles for effective public policy that will guide the long-term, global efforts that are needed to meet growing worldwide demands for energy, promote economic growth, and protect the global environment.
Promoting essential global action through bi-lateral agreements and assistance programs, advancing scientific knowledge through the Climate Change Research Initiative, and promoting technological innovation through the Climate Change Technology Initiative provide a solid foundation for actions to address climate change.
At various places, important discussions of climate change science in the Report draw on prior work that is either not supported or has been heavily criticized. In particular, we note that climate science is discussed in two places in the draft “2001 U.S. Climate Action Report. -- The first place is in chapter 1 beginning on page 3. This section is adapted from Climate Change Science: An Analysis of Some Key Questions (NRC, 2001). The first paragraph of this section states that human activities are causing the surface temperature to rise. While this statement is based on a similar statement that appears in the summary of the NRC report, it is noteworthy that there is nothing in that report itself which amplifies or justifies the comment. It does not appear in the report proper, written by the NRC Committee on the Science of Climate Change.
Our concern is that, while there is agreement that the concentration of CO2 in the atmosphere has increased due to man's activities and that the global surface temperature has increased over the last century, the degree to which the CO2 increase has affected this temperature increase remains unknown. This is an important distinction -- and one which the Report should continue to convey.
The second place climate science is discussed is in chapter 6 from pages 1 to 29. This discussion focuses on the predicted impacts of climate change in the U.S. and is based on the 2000 national assessment, Climate Change Impacts on the United States which was prepared and released by the Clinton Administration prior to the 2000 election. The impacts on the U.S. were based on Global Climate Model (GCM) runs to the year 2100 using the Canadian (CGCM) and United Kingdom (Hadley) models. The scientific community has been highly critical of the assessment as a whole, and, in particular, the use of the GCMs to predict regional impacts. These criticisms are summarized in the comments that were submitted to the Office of the U.S. Global Change Research Program in August 2000. Particularly pertinent comments are found in the references listed below. Copies of these references are attached for your review.
Based on the comments in these references, the regional scenarios are at best very speculative and cannot be considered credible. The Report should reflect and acknowledge these shortcomings and consideration should be given to deleting these discussions in chapter 6. Instead, the text should reflect the uncertainty that exists concerning any projection of future climate, particularly regional impacts. In addition, scientific research should be continued to improve the understanding of the climate system, the regional impacts of potential increases in temperature, and the effectiveness of alternative policy actions.
Thank you for this opportunity to comment on this 2001 Climate Action Report. If you have any questions… on the climate science, please call Dr. George Wolff at 313-665-2948.
Response: It is unfortunate that the many citations that are mentioned with regard to the supposed shortcomings of the USNA report refer to the public review draft of the National Assessment made available in June 2000 because many of the comments that were submitted were addressed in the final version of the report that was issued in November 2000. In response to the comments submitted on the draft National Assessment, as explained more extensively above in response to the API comments (see comments 1 and 2), the intent of the USNA report and its approach in using scenarios was more completely explained. In addition, the revisions made clearer that many of the various criticisms cited were critical of a goal and purpose that simply was not intended. For example, the final version of the National Assessment report explained the three-pronged approach to treatment of potential climatic conditions, the use of model results as plausible scenarios (there is no claim that they are accurate predictions), and the manner in which conclusions were evaluated based on the consistency of a combination of available experiments, extrapolation of past trends, theoretical analyses and expert judgments. The NAST also prepared a response to all of the submitted comments (available through the NSF library). It is unfortunate that none of these additional efforts seems to be acknowledged or considered in the criticisms submitted.
The NAST also added a section to its report calling for additional research, as is suggested here as being important by GM, and the Administration is working actively to enhance the research program. However, as the NAST explained in their report, there are quite a number of conclusions that can be drawn based on available information, even in the absence of results that are known with absolute certainty.
Given the dated nature of the references provided by GM and its failure to update its criticisms of a draft report to reflect the changes made in the final report, this general criticism seems moot. However, to make sure that Chapter 6 would stand on its own, additional explanations of these points were added to the chapter.
General Comment 15.
The 2001 U.S. Climate Action Report (USCAR) is, in general, more balanced than analogous compendia, such as the Third Assessment Report of climate change recently published by the United Nations Intergovernmental Panel on Climate Change (IPCC) , or the U.S. National Assessment (USNA) of climate change.
However, certain portions of USCAR, particularly Chapter 6, rely heavily on the USNA or the 2001 report of the National Academy of Sciences (which itself relies heavily on the USNA). Whatever originates from the USNA is highly flawed because the USNA is based upon a true miscarriage of science: it is based upon two models for future projections of climate that perform worse than a table of random numbers when applied to recent climate. The producers of the USNA, mainly the U.S. Global Change Research Program, have ignored this glaring problem, even as it is well-known that they were aware of it. Further, the USNA is based upon a selection of the two most extreme climate models for U.S. temperature and precipitation, for which there is no scientific defense.
Consequently the quality of large sections of the USCAR has been fatally impaired by the acceptance of the nonscientific USNA. This applies mainly to Chapter 6, Impacts and Adaptation. These sections -- and mainly Chapter 6 -- need either to be re-written, or a prominent note needs to be appended detailing the tragic flaws in the USNA.
It is hoped that these comments will force a re-opening of the USNA process, which was headed by the U.S. Global Change Research Program (USGCRP) with a specific investigation into how such a document could have been published, when USGCRP and the associated scientists knew that it was based upon models that simply did not work. As it stands, it is the blackest of marks upon U.S. Environmental Science in recent decades, and the historical credibility of our considerable efforts in this science are at stake. The blatant disregard of science in the USNA (and therefore in the USCAR) will not be noted today or next week. But, in coming decades, academic research carried out in a more dispassionate atmosphere than exists today will surely uncover these flaws and the attempts to cover them up. A responsible Agency would expose them NOW. This review affords that opportunity.
My main comments are directed at Chapter 6; however, there are a few others included below.
Response: These issues are responded to below where the arguments about the models are more fully developed.
General Comment 16.
OVERALL USE OF THE USNA AS BACKGROUND: The essential problem with the USCAR is that it is based upon the USNA. That report is based largely on two climate models, neither one of which, when compared with the 10-year smoothed behavior of the lower 48 states (a very lenient comparison), reduces the residual variance below the raw variance of the data. The one that generates the most lurid warming scenarios -- the Canadian Climate Centre (CCC) Model -- produces much larger errors than are inherent in the natural noise of the data. That is a simple test of whether or not a model is valid (see attached Figures 1 and 2) -- and both of the models used in the USNA fail. All implied effects, including the large temperature rise, are therefore based upon a multiple scientific failure. The USNA's use of those models and that approach is a willful choice to disregard the most fundamental of scientific rules. (And that they did not find and eliminate such an egregious error is testimony to grave bias). For that reason alone, the USCAR should be withdrawn from the public sphere until it becomes scientifically based.
Response: Dr. Michaels comment exhibits a serious misunderstanding of what was being done in the USNA. Basically, he is arguing that a test should be applied to the climate models before they are used, and that the key evaluation point should be to determine if the models can simulate the detailed decadal variations in the climate of the 20th century. However, while this might be a useful test were the USNA attempting to predict the decadal climatic fluctuations of the 21st century, this is not at all what was proposed as an application for the models.
Dr. Michaels is surely aware that decadal variability (which is the variance that his 10-year filter focuses on) is dominantly influenced by such natural factors as the occurrence of volcanic eruptions and air-sea variations (e.g., El Nino/La Nina events and similar oscillations). These fluctuations tend to involve variations in global average temperature of a few tenths of a degree and are basically (along with solar variations) essentially the climatic noise in the context of which the long-term change due to the increase in greenhouse gases is occurring.
With respect to the internal natural variations, Dr. Michaels should realize that the expected amount of decadal variance explained over a century-long simulation, given that the model simulations were started in 1900 or earlier and not tied to the timing of these oscillations in any way, should actually be about zero (with some range, indicating that by chance some runs may explain some of the variance and some may be counter to that -- and any single run could have any of these values).
In that volcanoes are unpredictable and have relatively short-term (few-year) effects,they are not included in the model calculations, so again, the model should not be expected to explain their contribution to any variance. Regarding solar influences, the effects of short-term influences of changes in solar radiation (e.g., sunspot cycles) are generally thought to be small, so again, no explaining of decadal variance should be expected as the details of the ENSO cycling would not be expected to coincide out many decades into the future.
What the USNA was focused on instead was the long-term trend and factors affecting that; indeed, the results in the USNA were typically portrayed as century-long trends and it was indicated that short-term variations, such as decadal variations, should not be considered as realistic as they are controlled by largely random natural factors (e.g., ENSO). Because of the longer-term focus of the USNA, the relevant test to apply is to determine if the models are explaining the long-term trend during the 20th century, and based on this test, as explained in the IPCC WG I report, the models are doing quite well. For example, NCDC examined overlapping averages for 1, 5, 10, 20, and 25 year periods and performed the same analysis of the variance of the residual vs. observations as did Dr. Michaels. This was done for both models, for both the globe and for the lower 48 U.S. states, and for both annual temperature and precipitation. Although shorter averaging periods did indeed result in the residual variance, in general, being SLIGHTLY larger than the variance for the observations, in many instances the longer averaging periods (20 and 25 year) resulted in residual variances being SMALLER than the variance for the observations. Since the longer averaging periods are indicative of longer-term variability and change, which is the driving climate question for this Assessment, then this result provides credence to the use of these two model simulations for providing physically-consistent scenarios of the climate of the 21st century.
Thus, the analyses indicate that the test favored by Dr. Michaels, when applied in a manner consistent with the intended use of the model results (that is to look at long-term trends), does indicate that the models pass the test. By contrast, the failure of the models to pass the test as formulated by Dr. Michaels is a result of a fundamental misunderstanding on his part of how the models were to be used -- namely to provide plausible scenarios for how the long-term climate might change rather than predictions of how the climate is expected to change over the next decade or two.
General Comment 17.
MODEL SELECTION PROBLEMS: Then there is the problem of model selection in the USNA. As shown in Figure 9.3 of the Third Assessment of the United Nations Intergovernmental Panel on Climate Change, the behavior of virtually every General Circulation Climate model (GCM) is the production of a linear warming, despite assumptions of exponential increases in greenhouse forcing. In fact, only one (out of, by my count, 26) GCMs produces a substantially exponential warming -- the CCC model. Others may bend up a little, though not substantially, in the policy-relevant time frame. The USNA specifically chose the outlier with regard to the mathematical form of the output. No graduate student would be allowed to submit a thesis to his or her committee with such arrogant bias, and no national committee should be allowed to submit such a report to the American people.
Even worse, the CCC and Hadley data were decadally smoothed and then (!) subject to a parabolic fit, as the caption for the USNA's Figure 6 makes clear. That makes the CCC even appear warmer because of the very high last decadal average.
One of the two models chosen for use in the USNA, the Canadian Climate Center (CCC) model, predicts the most extreme temperature and precipitation changes of all the models considered for inclusion. The CCC model forecasts the average temperature in the United States to rise 8.1°°F (4.5°C) by the year 2100, more than twice the rise of 3.6°F (2.0°C) forecast by the U.K. model (the second model used in the USNA). Compare this with what has actually occurred during the past century. The CCC model predicted a warming of 2.7°F (1.5°C) in the United States over the course of the twentieth century, but the observations show that the increase was about 0.25°F (0.14°C) (Hansen, J.E., et al., 1999: GISS analysis of surface temperature change. Journal of Geophysical Research, 104, 30,997 -- 31,022), or about 10 times less than the forecast. If the observed ratio continues into the future, the U.S. temperature increase by the year 2100 will be less than 1°F and hardly noticeable. The CCC forecast of precipitation changes across the United States is nearly as extreme. Of all the models reviewed for inclusion in the USNA, the U.K. model predicted more than twice the precipitation change than the second most extreme model, which was the CCC model. The CCC model itself forecast twice the change of the average of the remaining, unselected models. Therefore, along with the fact that GCMs in general cannot accurately forecast climate change at regional levels, the GCMs selected as the basis for the USNA conclusions do not even fairly represent the collection of available climate models.
Why deliberately select such an inappropriate model as the CCC? Tom Karl, a NOAA scientist, told me that that the reason the USNA chose the CCC model is that it provides diurnal temperatures; this is a remarkable criterion given its base performance. Consider the logic: Because we want to include the minute detail of diurnal temperatures, let's select the most extreme climate model in existence, in terms of exponentiality.
Thus the USCAR is driven by a model that 1) doesn't work over the United States; 2) is at functional variance with virtually every other climate model. It is simply impossible to reconcile this skewed choice with the rather esoteric desire to include diurnal temperatures. This reviewer leaves it to everyone else to speculate on the obvious reason.
It is clear that reliance on the USNA in the Climate Action Plan gravely compromises the validity of Chapter 6. It is therefore necessary to write some type of disclaiming footnote showing that the models used there are not valid and are an extreme selection.
Response: These charges by Dr. Michaels all appear to be incorrect, and seem to be based on a misidentification of models by Dr. Michaels in his reading of Figure 9.3 in the IPCC WG I TAR. In Figure 9.3 (page 537) of the IPCC TAR, which the text indicates shows the results of 19 models in the CMIP2 test simulation (not the 26 sets of results that Dr. Michaels says he counted), the model with the highest observed temperature increase (which we believe is the set of results to which Dr. Michaels is referring) is the CCSR/NIES2 model. As the table on the page 538 of the IPCC report indicates, this model is described by Nozawa et al. and is from the Japanese Climate Modeling Centre -- not the Canadian Centre (and we have also checked earlier drafts of this figure appearing in earlier drafts of the IPCC WG I report and do not think changes in the figure are the cause of this mistaken identification).
Based on Figure 9.3, our reading of the curves suggests that the Canadian model appears to be among those in the upper half of the band created by most of the models, just as described in the USNA. In addition, the Canadian model (CGCM1) does not appear to be an unusual or extreme outlier as Dr. Michaels claims [e.g., versions of the GFDL and Hadley (UK) center models appear comparable]. Figure 9.5 in the IPCC report does show that the global average temperature in two versions of the Canadian model do seem to be rising more rapidly than many of the other models in this smaller set showing results of the models on realistic cases, but the GFDL model, for example, appears to be approximately tracking the Canadian model results (although it was not, at that time, run out to 2100), and so the Canadian GCM results do not appear to be unique. With respect to changes in global average precipitation, the Canadian model results also appear to be within the pack of models. Therefore, based on the IPCC figure cited by Dr. Michaels, we do not understand how Dr. Michaels comes to the conclusion that the Canadian model is so much an outlier that its results should be disregarded as implausible, even if the results of the Canadian model do tend to be in the upper group of the pack of models.
That Dr. Michaels is likely to be referring to results from the wrong model in his comment also seems clear from his calculations of the model-simulated temperature changes over the US. Page 36 of the NAST Foundation report includes a table where calculations for the temperature change over the US are done for a number of models and compared to the observed changes (Japanese model results were not available for comparison at the time of the analysis). While the Canadian model does estimate a change in temperature over the US during the 20th century of about 1.05 C (1.9 F) that is higher than the observed rise [estimated by NCDC authors to be from 0.3 to 0.8 C (0.5 to 1.4 F)], this warming is significantly less than Dr. Michaels' claim that the Canadian model indicated an increase of 1.5 C (2.7 F) over the US. We presume this discrepancy arises because Dr. Michaels is using the results of the wrong model.
With respect to the indicated comparisons, it should also be noted that while Dr. Michaels cites warming over the US that is less than indicated in the preceding paragraph, during review of its report when the same comment was submitted, the NAST checked with Dr. Hansen about these estimates and learned that in this paper Dr. Hansen had “failed to take into account important observation bias in the US and that he [was] revising these numbers upward to correct for this mistake and this revision indicates substantial warming over the US. --
Also, Dr. Michaels comment that the use of a parabolic fit of the changes in temperature from the model to match them with the historic observations somehow amplified the overall warming is mistaken. This fit was only to achieve a smooth transition and had only a small influence near the meeting point of the data sets. Thus, this approximation had no significant effect on the overall rate of rise of temperature projected by the models for the 21st century.
With regard to model simulations of projected changes in precipitation over the US, the two models indicate significant differences in their projections of potential changes. In that projections of potential changes in precipitation are less certain than projected changes in temperature, having models with quite different results provided scenarios that tended to span the range of plausible outcomes (it should be noted that the difference is primarily for the summertime over the eastern two-thirds of the US). While Dr. Michaels seems particularly critical of the Canadian model results, it is interesting that the subsequent, presumably improved, generation of the Hadley model (HadCM3) tends to give results more similar over the US to the results of the Canadian model than to the HadCM2 model. In any case, what the two sets of model results provide is a range of outcomes for consideration, properly recognizing that significant uncertainty exists regarding changes in this variable.
With regard to why these particular models were selected as the primary ones for use by the USNA, Dr. Michaels seems to listen more to anecdotes than to seek out the written explanations. The reasons that the results from the Canadian (CGCM1) and Hadley (HadCM2) center models were selected for primary use in the USNA are explained on pages 31-40 of the USNA Foundation report. As explained, there were several factors considered together that only these two models satisfied, including that the models did include a diurnal cycle. (Results from a model with a diurnal cycle are needed by the ecosystem models as a source of estimation of changes in minimum and maximum temperatures, which Dr. Michaels frequently points out have changed in the past by different amounts). Also, these were the only model results that were available for processing when needed, were publicly available and submitted to the IPCC, and had relatively fine spatial resolution, etc.
In addition, Dr. Michaels seems to not understand that the USNA not only used a three-pronged approach to evaluating potential climate impacts (see response to comment 2) but also that a number of the groups were able to use results from more than the minimum set of models so that, for example, issues relating to the Great Lakes were based on the results of 9 models (8 of which were in agreement).
Dr. Michaels (along with other commenters about the use of climate model results in the USNA) also seems to be totally unaware of how the model results were processed for use in carrying out the impact studies (see the explanation on page 39 of the USNA Foundation report). Basically, in recognition of the limits of model representations of regional climates, the observed climatic conditions of the 20th century over the US were used to carry out analyses of changes that may have occurred, and for the 21st century, the model-projected changes in temperature and in the fractional change in precipitation were used to generate the projected conditions over a grid finer than the used by the GCM (thereby helping to generate better estimates of altered conditions in regions of complex orography). This approach (that is, using models to estimate changes rather than absolute values of conditions) had the beneficial effect of canceling out the major systematic problems in the model simulations, allowing a more accurate basis for calculating potential impacts on ecosystems.
In summary, in contradiction to Dr. Michaels' comments, the projected changes in climatic conditions over the US were representative of the full set of models and were normalized to better match the actual fine-scale climatic conditions over the US. While there remain uncertainties, the approach provides quite plausible scenarios of potential changes in climate over the US. That some of the changes seem unprecedented compared to our experiences during the 20th century properly characterizes the situation. The chapter now includes a somewhat more extensive explanation of the ways in which the models are being chosen, but other of Dr. Michaels' comments are not relevant to how the models were used and he should go back and more fully read the US National Assessment reports.