b) This means that the world is stuck with a sea level rise of about 18 cm (7 in)/ yr, just what was observed during the past century. And there is nothing we can do about it, any more than we can stop the ocean tides.

c) Careful analysis shows that the warming of the early 1900s actually slowed this ongoing SL rise, likely because of increased ice accumulation in the Antarctic.

The bottom line: Currently available scientific evidence does not support any of the results of the NACC, which should therefore be viewed merely as a "what if" exercise, similar to the one conducted by the Office of Technology Assessment in 1993.5 Such exercises deserve only a modest amount of effort and money; one should not shortchange the serious research required for atmospheric and ocean observations, and for developing better climate models.

The NACC should definitely NOT be used to justify irrational and unscientific_energy and environmental policies, including the economically damaging Kyoto Protocol. These policy recommendations are especially appropriate during the coming presidential campaigns and debates. I respectfully request that an expanded exposition 6 be made part of my written record. [The Executive Summary is in the appendix, the whole document can be found at: //www hoover.stanford.edu/publications/epp/102/102complete.pdf]

The CHAIRMAN. Thank you, Dr. Singer. In other words, you reject the findings of the Assessment practically in its entirety.

Dr. SINGER. I think these are interesting exercises, what-if exercises, but I do not think they should be taken seriously.

The CHAIRMAN. Dr. Schmitt, in the climate change you have noted in your findings, what is the impact on the ecology of the oceans, such as the effect on reef life, et cetera?

Dr. SCHMITT. Well, I am hardly an expert, but there are very significant impacts on fisheries. I know the cod fishery in New England has changed a lot. It is difficult to sort out whether it is due to overfishing or just changes in the North Atlantic Oscillation, because the water off Labrador is so much colder now than it was 10, 15 years ago.

In other areas warming in tropical areas has a great impact on the life of corals. There is a phenomenon called coral bleaching, which basically kills a coral reef, and I believe that occurs if the water gets too warm. In other areas the stocks of salmon have been correlated with these climate phenomenon such as the North Atlantic Oscillation and the Pacific Decadal Oscillation.

These phenomena, with their long time scales-they are 5, 10, 15 year cycles-hold out the hope of predictability because the ocean has this long memory of the heat content. It has enormous heat content. It has 99.9 percent of the heat content of the climate system, and we need to be doing a much better job on monitoring that heat content.

The CHAIRMAN. Mr. Karl, do you have a response to Dr. Singer's views?

Mr. KARL. Yes. I have-I do not know where to begin, to be quite honest.

Dr. SINGER. Just start anywhere.

Mr. KARL. I guess I would first point out that what we did in the assessment was draw on the published referenced literature. In fact, I think if you look at the references in the assessment there is probably over 95 percent are from papers that have been peer

4 S.F. Singer. Hot Talk, Cold Science: Global Warming's Unfinished Debate. (The Independent Institute, Oakland, CA. (second edition, p. 18)).

5 Office of Technology Assessment. "Preparing for an Uncertain Climate." Govt. Printing Office, Washington, DC. 1993.

6S.F. Singer. "Climate Policy-From Rio to Kyoto: A Political Issue for 2000-and Beyond." Hoover Institution Essay in Public Policy No. 102. Stanford, CA, 2000.

reviewed. The other 5 percent are reports that often were used because we needed to obtain the data from those reports.

What I would just want to point out is that the position of Dr. Singer, although I very much respect his opinions, is quite at odds with the scientific published literature. I would just point out a few egregious examples of what I have heard.

50 percent of the rise, or more than half of the rise in sea level is due to the expansion of ocean waters. As temperatures increase, the ocean density increases, and it has nothing to do with the melting of ice glaciers.

The other aspects that I heard which I would completely disagree with, and that is the warming in the U.S. record. It is very clear, in fact, especially in the last decade or two, the U.S. was lagging behind global temperature increases up till the early 1980's, and since the mid-1980's, and particularly during the 1990's, the U.S. has virtually captured the rest of the globe.

That is not to be unexpected. In fact, if you look at one area in the country, the Southeast part of the U.S., it is where we have not seen much of an increase in temperature. In fact, there have been very small changes in temperature, but again if you look at the 1990's in the Southeast, we are almost now as warm as we were back in the 1930's.

And again, I might point out that in the Arctic we have had record low ice extents in the Arctic. In fact, if you look at the latest IPCC report that is up for review, it is documented that we also see reduced snow cover extending across the northern hemisphere. So it is not just the temperature records that we use to deduce the fact that the globe is warming. There are many, many other ancillary pieces of information that are used as well.

So those are just a few of the things I would like to point out. The CHAIRMAN. Dr. Janetos, can you comment on the Science Magazine article which claims that the two models used in the report, the Hadley Center and the Canadian, are not intended, or capable of predicting future impacts of climate change on a regional basis?

Dr. JANETOS. Mr. Chairman, in the assessment we tried to be very careful to say what we have not done is try to predict exactly what the future will be like. Each of these models, each of these general circulation models was selected after a careful review of the criteria that we set a priori in order to understand the potential consequences.

They had to have saved the right data, they had to have used an emissions scenario that was already well understood, and they had to be documented in the scientific review literature.

What we have tried to do is essentially ask the question, what if the models are correct? Since we cannot distinguish between them on scientific bases, we need to be able to understand the implications of the different plausible futures that they hold for the U.S.

The CHAIRMAN. Thank you. We will be submitting written questions that we hope you will be able to respond to. I apologize for the short-circuiting of this hearing. We will be asking the second panel to come back. We thank you for taking your time to come before the Committee.

You have added a lot to this very important discussion, and I want everyone to be very aware that we will continue to pursue this with further hearings. I think that it is an issue that is extremely important for us to seriously consider, and I thank you for being here. I thank you for your continued efforts, and I hope I have the opportunity to personally visit with all the members of the panel as we explore this very complex and difficult situation. I thank you.

Unfortunately, this hearing is adjourned.

[Whereupon, at 10:10 a.m., the Committee adjourned.]

APPENDIX

RESPONSE TO WRITTEN QUESTIONS SUBMITTED BY HON. JOHN MCCAIN

TO THOMAS R. KARL

Question 1. Can you explain the process used in the report to address the differences between the results of the two computer models and how this process is used to identify new research areas?

Answer. A lexicon was developed to communicate scientific uncertainty related to the scenarios from the two climate models used in the National Assessment as well as other models, data, information, and state of knowledge. This lexicon conveyed areas of uncertainty by linking words with probabilities. For example, if the National Assessment Synthesis Team (NAST) assessed the about even odds for an event the word 'possible' was used. On the other hand, if the NAST was fairly certain about an event, then words like "very likely" or "very probable" were used to indicate that there was more than 90% chance of occurrence. Where both models agreed, projections were seen as more certain. In cases where model results differed, both possible future scenarios were examined and results were characterized as less certain. Model results were not merged.

Whenever the NAST encountered instances where there was considerable uncertainty about the outcome these areas were then identified in a 'Research Needs' section of the report. In our Research Needs section we recommend a number of measures that are required to improve our confidence in modeling future climates. Question 2. If the report explicitly does not "merge" the results of models that disagree, can this assessment be considered a fair analysis of climate change? Furthermore, when the two models diverged, where these results downplayed in the report versus when they concurred?

Answer. In response to the first question, it is difficult to understand why the assessment would not be considered fair if the two primary models were not merged. As indicated above, the NAST did not merge the scenarios from the models, but rather the NAST reflected the uncertainty related to several different possible outcomes and expressed this lack of confidence through use of the lexicon.

In response to the second question, the answer is no. Again, the NAST painstakingly used the lexicon to express its confidence in any projected changes for the 21st Century. Projected scenarios from all relevant models were discussed. This included both of the two primary models as well as the secondary models used in the Assess

ment.

Question 3. Dr. Schmitt has raised several issues concerning the impact of the oceans on the climate modeling results. How sensitive are the climate change impacts on the U.S. to changes within the ocean water temperatures?

Answer. Dr. Schmitt's remarks refers to improving climate forecasts from climate models that are dependent upon initial conditions. These deterministic climate model forecasts require information about the current state of the oceans. Clearly, it is very important to have comprehensive high-quality real-time ocean observations available to properly initialize these models.

The Global Climate Models used in the National Assessment do not require realtime initial conditions. They are self-contained models and generate their own ocean climate. Changes in ocean temperatures can have a large impact on the climate of the U.S. An obvious example relates to the changes of ocean temperatures in the tropical Pacific related to the El Niño southern oscillation and its effect on the temperature and precipitation in the U.S. Another example relates to hurricane formation. Water temperatures significantly less than 80 degrees Fahrenheit do not provide enough energy to the atmosphere to spawn powerful hurricanes. And as a result, hurricane formation is highly seasonal dependent.

Question 4. In the past few years, the U.S. experienced some distinctive weather patterns, namely El Niño and la niña. Can you discuss how these and other warm ocean water related weather patterns factors into your modeling efforts?