American Enterprise Institute

June 3, 2008

[Edited transcript from audio tapes]

12:45 p.m.	Registration
1:00	Introduction:	Christopher DeMuth, AEI
1:10		Panel I: The Science of Geoengineering
	Introduction:	Samuel Thernstrom, AEI
	Presenter:	Tom Wigley, National Center for Atmospheric Research
	Discussants:	Kerry Emanuel, Massachusetts Institute of Technology
		Vaughan Turekian, American Association for the Advancement of Science
	Moderator:	Samuel Thernstrom, AEI
3:10		Panel II: The Implications of Geoengineering for Climate Policy
	Introduction:	Lee Lane, AEI
	Presenter:	Scott Barrett, Johns Hopkins University
	Discussant:	Fred Iklé, Center for Strategic and International Studies
	Moderator:	Lee Lane, AEI
5:00	Adjournment

Proceedings:

Panel One

Samuel Thernstrom:  Thank you, Chris.  Thanks very much.

When the writer, Arthur C. Clarke, died in late March, I was struck by the following passage in his obituary.  The notions of the -- "His notions of the future remained unswervingly radical.  Sir Arthur knew that outlandish ideas often become reality."  "But they provoked," he wrote, "three stages of reaction.  First, 'It is completely impossible.'  Second, 'It is possible, but not worth doing.'  Third, 'I said it was a good idea all along.'" 

At the moment, I think the idea of changing the earth's environment in ways that would counteract the effects of global warming lies probably somewhere between those first two stages of popular opinion.  Most people probably believe that it is impossible; although most scientists in the field increasingly believe that they are wrong. 

Consequently, critics are retreating to the second idea that geoengineering may be feasible, but that it is undesirable.  That may be true, but only time and further scientific investigation and policy analysis will tell.  But, we are certainly altering our environment right now in a massive, uncontrolled, and certainly unintentional experiment, and the solutions proposed so far do not seem to be effective. 

While the science is still in its infancy, it is clear that geoengineering offers a potential solution to the worst aspects of global warming in that it could be fast, effective, and affordable.  Those are three powerful virtues in a climate policy that mitigation, at the moment, cannot claim.  Given those facts, I think it is unavoidable that people will eventually embrace this idea if the science holds up.

Geoengineering is without a doubt the most original and potentially important idea in climate policy that I have heard of since I first became interested in the general climate issue more than twenty years ago.  It is also certainly the most controversial.  There are many important and complex questions about both the science and policy implications of geoengineering and scholars have really only just begun to explore them.  That is, of course, the purpose of our conference today, and of the series of papers that we intend to commission over the next couple of years.

Now, geoengineering is not actually a new idea.  It has been discussed as early as the mid-1960s, but it was never taken particularly seriously until now.  That is starting to change as a set of facts is becoming increasingly clear.  Unless we can radically curtail global greenhouse gas emissions, atmospheric concentrations of greenhouse gases will rise dramatically over the course of this century.  Warming could be as little as one degree Celsius or as much as six degrees.  That is a broad range and there is certainly significant uncertainty involved in such long-term projections.  But, they do give a meaningful sense of what future scenarios policymakers are considering. 

Optimists may choose to believe that it is not only possible, but likely that the world will find ways to effectively phase out the use of fossil fuels in the coming -- in the next few decades.  Realists, however, are increasingly recognizing that the odds of that happening are not very good.

We have two decades of experience now with climate policy and very few meaningful signs of progress.  We still face the same fundamental problems that have plagued this process all along.  The lack of clean energy technologies that can be produced on a mass scale at a cost that even developing nations such as China can afford and the lack of a global political consensus to develop and deploy these technologies.  Those facts seem unlikely to change in the near future.

Voters, of course, are generally supportive of actions to address warming while remaining very reluctant to bear the actual costs of doing so.  We see this ambivalence among our elected officials as well.  Congress, of course, as everyone knows has finally taken -- is finally debating a climate bill this week and although it is not expected to pass, I do think that it is likely that legislation will be enacted of some sort within the next couple of years.  And it is also possible that a successor agreement to Kyoto will be negotiated in the same time.  But, if history is any guide, the environmental effects of these efforts are likely to be very modest.

American and European emissions may slow somewhat in the coming years, but they will not cease.  Meanwhile, China will continue to build a coal fired power plant every week and global emissions will continue to rise.  Against this backdrop of policy failures, we have a growing sense of alarm from some -- from a number of climate scientists. 

NASA Scientist, Jim Hansen, for instance, and a number of colleagues published a paper at the end of March calling for stabilization of greenhouse gases at 350 parts per million; a target that would significantly be more ambitious than the more commonly discussed 450 parts per million.  The authors of that paper note that if the present overshoot of the 350 part per million target is not brief, there is a possibility of seeing irreversible catastrophic effects.

Just days after Hansen's article was published, Tom Wigley, who is with us today, and two colleagues of his, Roger Pielke, Jr. and Christopher Green, published another important article in the magazine Nature that looked into the assumptions behind the IPCC's assessment of how much emissions will need to be cut in the future.  Their conclusions were quite shocking.  Although the IPCC has been accused by its critics of alarmism, their assumptions about how much natural progress in energy efficiency will occur are surprisingly sanguine. 

It is responsible, of course, to assume that there will be improvements in energy efficiency and carbon intensity in the coming years driven simply by the desire of companies and consumers to cut costs and reduce waste.  The goal of climate policy naturally is to accelerate that trend, but understanding the baseline of that natural trend is key to knowing how ambitious our policy targets need to be in order to be effective. 

The authors of this paper found that the IPCC had assumed that two-thirds or more of all of the energy efficiency improvements and decarbonization of energy supply required to stabilize greenhouse gases will occur naturally absent any climate policies at all to promote them.  Unfortunately, the data does not seem to support the IPCC's assumptions.  The IPCC, for instance, assumes that carbon intensity will drop this decade.  So far, of course, it is not doing so.  It is increasing.  It further assumes that energy intensity will improve at a rate greater than one percent per year, which the authors believe may be neither realistic nor achievable even with a substantial policy effort. 

To take one specific and particularly important example, the IPCC has assumed that China's emissions will grow at a rate of only two-and-a-half to perhaps five percent a year this decade.  While the actual data at the moment shows that the correct rate is likely to be at least 11 percent, if not 13 percent.

The authors' conclusion is really quite serious.  They write, "The world is on a development and energy path that will bring with it a surge in carbon dioxide emissions; a surge that can only end with a transformation of global energy systems.  We believe that technological transformation will take -- we believe that such technological transformation will take many decades to complete even if we start taking far more aggressive action on energy technology innovation today."

That dismaying conclusion, therefore, is even -- is that even with current policy proposals, we are likely to fall far short of the emissions reductions needed to make these efforts successful.  The question of the impact on warming -- the impacts of warming is far too complicated to explore in detail today, but I think there is a simple point that we could probably all agree on.  We do not really know how much warming to expect, one degree or six, but there is not much question that there is some real risk that the consequences of warming may prove to be significant and possibly even catastrophic. 

For instance, Harvard economist Martin Weitzman has gotten a lot of attention for a paper that he published in February that explores the risk of possible worst case scenarios of catastrophic climate change such as a rapid melting of the polar ice caps that would produce dramatic sea level rises.  Weitzman acknowledges, of course, that these are low probability scenarios, but his point is that the risk is not zero by any means and that the public policy ought to take those risks into account.  And indeed it certainly ought to, but the question is, "How?" 

We could say, of course, that we should simply redouble our efforts to cut emissions and certainly some people will do that, but I have already discussed the limitations with that approach.  So, a prudent response, it seems, would be to also consider what our options might be if that does not work, particularly because emissions reductions are very slow to take effect given the persistence of greenhouse gases in the atmosphere.  If the polar ice caps start melting quickly it will be too late then through mitigation.

If mitigation fails, at the moment, our only other option is adaptation, but if the effects of warming are more severe than we can reasonably adapt to, what can we do?  A number of distinguished scientists believe that geoengineering, that is technologies that would change features of the earth's environment to protect us from warming, may be able to provide a crucial safety valve should other mitigation policies fail to sufficiently curb warming. 

The various technologies and visage for geoengineering are obviously untested and the science, it should be acknowledged, is still in its infancy.  But, the research conducted to date is quite promising.  The National Academy of Sciences, for instance, examined this idea in 1992 and concluded that geoengineering is feasible, economical, and capable.

The most likely techniques would involve reflecting about two percent of the incoming sunlight, thereby blocking just enough of this energy to cancel out the warming that is expected to occur during the remainder of this century.  There are various technologies that could achieve this -- there are various different ways of doing this using relatively simple technologies.  And there are other ways of achieving the same general effect, and Tom can certainly talk about a number of those specific examples. 

But, the simplest, most popular idea is to artificially reproduce the cooling effect that was observed after the 1991 eruption of Mount Pinatubo, which cooled the planet by roughly a half of degree Celsius for two to three years.  The paper from Tom Wigley that is in your packet describes the potential effect of artificially reproducing roughly half of a Mount Pinatubo eruption every year.

Now, volcanoes, of course, are very crude ways of doing this.  An engineered system would be much more precise and would use very fine particles rather than thick volcanic ash.  But that is, of course, by no means the only technique.  Scientists are, for instance, studying alternatives such as spraying sea water to increase the reflectivity of clouds over the ocean, which would have the same ultimate effect.

Naturally, these ideas may seem quite farfetched to people who have never heard of them before, if not frightening, but a growing number of leading climate scientists take them very seriously.  The National Academy of Sciences, NASA, and the Department of Energy have all studied geoengineering and concluded that the science is credible enough to warrant a more systematic study of the field, yet the Bush Administration has declined to support such research, although the federal government currently spends in the range of $2 to $3 billion a year on climate science R&D, depending on how you count the figures.

Given the administration's relatively poor public standing on this record, I think it is probably just as well that it did not embrace the idea.  But, obviously, Congress and the next administration will have the opportunity to consider funding it and the question is whether the fear factor will be too strong for them, as well.

Certainly, there are a number of plausible objections and questions that come to mind when considering geoengineering.  The question is whether there are, upon reflection, reasonable answers to those questions.  And so, this is the first of a series of conferences that we intend to host to explore them with leading scholars. 

Critics have two main concerns about geoengineering.  First, that it would have negative side effects such as ozone depletion or the disruption of rainfall patterns, and I am going to leave that question to our panel to discuss since they are the scientists.

But, the second critical question is the so called "moral hazard" question; the fear that discussion of geoengineering will undermine support for mitigation, and I would like to talk about that just for a minute.

Scientists naturally as a rule -- as a general rule are exceptionally strongly committed to the pursuit of knowledge.  "Find the facts and let the chips fall where they may."  So, the fact that some scientists have a rather different approach to geoengineering is somewhat extraordinary.  Geoengineering is an idea that according to some is simply too dangerous to talk about and has met with a surprising degree of hostility within the scientific and environmental communities. 

When Harvard University and the Council on Foreign Relations recently held conferences on this subject, for instance, they did so behind closed doors.  I have also heard reliable reports that at least one other major university was simply afraid to hold any meeting on the subject at all.  I have heard and read reports of scientists who fear that even any discussion of geoengineering, much less real research and experimentation, could undermine the motivations for mitigation.  Geoengineering, it is feared, would be seen as an excuse to simply continue emissions forever. 

With all due respect, I think this concern is seriously misplaced.  Geoengineering is a remarkable idea with great potential, but it is not the silver bullet.  It is not a permanent solution to warming and it is not a perfect solution to warming.  Obviously, there are risks involved in geoengineering and limitations also on what it can do.  I find it hard to imagine that there be any serious political support for policy of "Geoengineering forever and mitigation never."  It would be foolish, I think, to do -- to consider that and I do not know of anyone actually who advocates that.

Aside from the risks of geoengineering itself, at least one important environmental effect of high CO2 concentrations in the atmosphere, namely ocean acidification, cannot be corrected by geoengineering.  Geoengineering, therefore, if it is ever undertaken, must be considered a complement to, rather than a substitute for, a long-term program to transition to a zero emissions economy.  But, what geoengineering does is it potentially buys us time, several decades, to make mitigation work.  And, consequently, in fact, it may be the key to mitigation's success, not its undoing.  Time, above all, is what we need to make mitigation work. 

I do believe that it is possible to phase out the use of fossil fuels, but the prospects for doing so successfully improve dramatically if the goal is to do it over the course of a century rather than doing it in half that time or less.  Rather than being the death of mitigation, therefore, geoengineering may be the vital linchpin that makes it possible.  There are many challenges to mitigation, but every one of them can be potentially addressed with time.  Time coupled with sufficient dedication and resources is what will produce new clean energy technologies at a price that the developing world can actually afford. 

If we can solve the technology problem, we can solve the cost problem, and from there the political problems will fall away, I believe. 

I am particularly pleased that Tom Wigley could join us today since he is the author of a seminal article outlining this idea, which is included in your packet as I mentioned.  I just want to highlight the key conclusion of that article, which I think is quite striking.  "A combined mitigation-geoengineering approach to climate stabilization has a number of advantages over either alternative used separately.  A relatively modest geoengineering investment could substantially reduce the economic and technological burden on mitigation.  More ambitious geoengineering, when combined with mitigation, could even lead to the stabilization of global-mean temperatures at near present levels and reduce future sea rise to a rate much less than that observed over the 20th Century, aspects of future change that are virtually impossible to achieve through mitigation alone."

I should be clear that the question for policymakers today is not whether to deploy a geoengineering system immediately or even to make it a primary focus on the U.S. climate policy.  Rather, it is whether to make a serious investment in the research and development so that we might understand whether it is really a feasible option and might develop the means to deploy it if we find that we actually need it.  Part of that research, in fact, as I hope we will discuss with Tom, should be the question of how long we can afford to wait before geoengineering -- before experimenting with some degree of geoengineering. 

I do not consider myself an advocate of geoengineering, nor I think would any of our panelists today, and if my remarks may have given that impression so far, it is because I assume that most people on hearing this idea need to be persuaded that they should take it seriously.  It does seem rather outlandish to most people on first blush.  But, I am an unabashed advocate of conducting a rigorous research program into the science and engineering questions and in the coming months we intend to do that.

Without further ado, let me just introduce our panel.  Tom Wigley is a senior scientist at the National Center for Atmospheric Research.  He is one of the country's most distinguished climate scientists and one of the leading scientists in the world researching geoengineering.

Mr. Wigley has published widely in the field of climatology and related sciences and is the author of more than 250 referred journal articles and book chapters.  He has been a contributor to the -- all of the Intergovernmental Panel on Climate Change (IPCC) assessments and he developed the well-known MAGICC coupled gas-cycle/climate model used to produce temperature and sea-level projections given in IPCC reports.  He is also the former director of the Climate -- Climatic Research Unit at the University of East Anglia in the United Kingdom.

We are privileged to have two superlative discussants join us today to respond to Mr. Wigley.  Kerry Emanuel, sitting next to Tom, is a professor of atmospheric sciences at the Massachusetts Institute of Technology, where he has been on the faculty since 1981. 

He is one of the nation's leading atmospheric scientists.  Geoengineering is not his specialty itself, but he is a top climatologist and he has been generous enough to share his time today to offer his perspective on that question.

Professor Emanuel is the author or coauthor of more than a hundred peer-reviewed scientific papers.  His most recent books include Divine Wind: The History and Science of Hurricanes and What We Know about Climate Change, MIT Press, 2007.

Closest to me here, we have Vaughan Turekian, who is the chief international officer for the American Advancement -- American Association for the Advancement of Science.  Previously, Mr. Turekian served as special assistant for the under secretary of state for democracy and global affairs, where he was lead adviser on a variety of scientific issues, including the environment, technology, clean energy, sustainable development, and climate change.

Mr. Turekian also worked at the National Academy of Sciences, where he was the study director for the 2001 White House-requested report on the study of climate change science.  He is published widely on a range of topics.  I would particularly recommend to you his 2007 Foreign Policy article coauthored with Paul Sanders -- Saunders entitled “Why Climate Change Can't Be Stopped.”

We will follow our usual format for this conference this afternoon.  Mr. Wigley will present.  Mr. Emanuel and Turekian will comment.  I will give the panelists an opportunity to question each other, and then I will open up the floor to questions from all of you. 

Thanks very much.  Tom.

Tom Wigley:  All right.  Do you want me to talk from over there?

Samuel Thernstrom:  You can either do it there or [inaudible].

Tom Wigley:  With -- first, I would like to thank Lee Lane and Sam here for inviting me to this important event to discuss a subject that is quite close to my heart.  And I would also like to thank Sam for giving such a comprehensive overview really.  I mean I am just going to fill in some of the details and I hope that -- and I appreciate the fact that you have introduced a number of topics very comprehensively that make my task a lot easier.

Okay, so, ‘Science of Engineering’ is the title of my talk.  That is not my title.  That was the title that I was asked to address, and I am a scientist, so I am going to try and talk about the scientific aspects more than the technological and the policy and political aspects.  But, of course, those things will spin off from what I have to say.

Just to summarize the issues that I am going to try and cover here, I want to define a few terms, first of all.  I am going to concentrate on a particular type of geoengineering that is sometimes called Solar Radiation Management, and it deals with trying to change the amount of incoming solar radiation.  And there are different ways one can do that and, specifically, one can use a kind of artificial volcano idea by putting sulfate aerosols or some other particles in the upper atmosphere, the stratosphere, where they stay around for a few years, and they will reflect incoming solar radiation and cause a cooling that would perhaps compensate to some degree for the warming due to greenhouse gas emissions.

I am going to say something about the mitigation issue per se.  And then, as Sam says, in my little science paper, I talk about combining both geoengineering and mitigation and the need -- the essential need for mitigation no matter what we do, and I will say a little bit more about that.  Then I will talk about the sea-level and temperature -- consequences of these combined strategies, and finally draw some conclusions or summarize some of the results.

Okay, so mitigation -- I am sure most of you know the words, but I will define them nevertheless very briefly.  So, mitigation essentially means reducing greenhouse gas emissions.  Carbon dioxide is the most important and that can include things like reforestation or carbon capture and disposal and so on; just reducing the net emissions of carbon dioxide and the other greenhouse gases into the atmosphere.

Adaptation is trying to reduce the impacts of those climate changes proactively by developing adaptive mechanisms.  And then, geoengineering and, specifically, Solar Radiation Management, is the deliberate modification of the earth's short-wave radiation budget to try to reduce the magnitude of global warming or climate change.

Up until a few years ago, everybody realized that we were going to have to both mitigate and adapt.  We could not just do one or the other, and the reason why adaptation is important is that the climate system has a lot of inertia.  And so, no matter what we do, the continued global warming and the need for people to adapt to those changes in climate that are inevitable no matter what we do.  And more recently people have realized that perhaps a third corner of this triangle might be geoengineering, which should be combined with the other two strategies. 

Some examples of geoengineering in the sense of Solar Radiation Management that I am going to concentrate on here are the injection of aerosols or aerosol precursors.  So, for example, if we put sulfur dioxide in the stratosphere and there is enough water vapor there, then the sulfur dioxide will be oxidized and hydrolyzed to form sulfuric acid droplets, and it is those droplets that reflect back incoming solar radiation.  There are other ways of doing that, but I am going to concentrate on that particular method.

Another method that I mentioned is changing the reflectivity of clouds, and I have got some nice little pictures that show you that strategy.  And I think that that is a complementary approach that may also need to be explored because if it works then it is relatively inexpensive, and I do not think we should have a portfolio of just one strategy.  We need to have a full portfolio of a number of strategies because no one method is going to solve the problems that we face.

And then, there are other things like painting roofs white, and changing vegetation albedo, and so on, that I consider to be aspects of geoengineering.  I am not going to say any more about that, but in a sense they are examples of Solar Radiation Management.

Oh -- that is interesting -- oh, it is just a little slow response.  This is a picture just showing that if you increase the number of cloud condensation nuclei, then you do change the droplet size distribution in clouds and you change the reflectivity of clouds. 

So, this is one of the geoengineering approaches -- and this is demonstration that in principle this approach should work in a very elegant way of changing cloud condensation nuclei – and has been proposed by some English scientists.  And this is a very clever boat that they imagined might be moving about the world's oceans in the areas where there are the right sorts of clouds, where this is a self-propelled boat. 

It has very strange looking sails here that rotate and allow the boat to be moved around by the wind and the rotation also pumps water -- sea water up into these columns and sprays it up into the clouds and those small droplets then act as a cloud condensation nuclei, change the droplet size distribution in the clouds, and thereby reduce their albedo -- increase their reflectivity.  So, this is a rather interesting and elegant approach that has many, or if not all of these geoengineering ideas -- it is okay on paper, but is yet to be demonstrated as a practical consideration.

Okay, so why should we consider geoengineering at all?  You know, hopefully in an ideal world, we ought to be able to solve the climate change problem just through reducing emissions and through adapting to the inevitable changes that will occur.  Now, the real difficulty here is that if we are to follow Article 2 of the Framework Convention, and that is try to avoid what is called "dangerous interference" of the climate system, even though that is a term that is not well-defined, then there must be some stabilization target for carbon dioxide that will give a high probability of avoiding dangerous interference of the climate system. 

A general opinion is that that CO2 stabilization level is about 450 parts per million.  And as Sam says, Jim Hansen and other colleagues have suggested that it may be much less than that.  It might be 350 parts per million.  And, in fact, there are even arguments to suggest that we might have to go back to below the pre-industrial level to stabilize the rise of sea-level. 

The real question is -- you know, let us take this upper limit of the bound for avoiding dangerous interference at 450 parts per million.  The big question is, "Is that achievable?"  I mean do we have the technology level and do we have the political will?  But, do we have the technology to achieve levelization of 450 parts per million?

Okay, so there are two issues here that need to be thought about seriously.  Firstly, suppose that we decide that 450 parts per million is a reasonable stabilization target and then we find that we cannot get there from here, that the technology is not available, the political will is not there, but particularly the technology is not there to stabilize at 450 parts per million.  What might happen is that we might have to follow what is called an overshoot pathway where the concentration of carbon dioxide goes well above 450 parts per million, and then is brought back down by appropriate mitigation strategies back to 450 as an eventual target.  And that is a case that I am going to consider the overshoot pathway.

We might find out, as Jim Hansen believes, that the climate system is more susceptible to increasing greenhouse gas concentrations and that 450 parts per million is too high a target.  So, you know, how would we -- if we cannot even achieve 450 parts per million, then how would we ever get down to a lower level?  And that is where -- in both of these cases where we have to think seriously about geoengineering. 

Unfortunately, geoengineering has a number of potential risks associated with it.  I am going to consider later a scenario where I think those risks are minimized, but many of the recent research papers on this issue have asked the question, "What would be the risks associated with trying to offset the amount of warming associated with a doubling of carbon dioxide with geoengineering alone?" 

Okay, so that question can be posed in a slightly different way and one can say, "What if we wanted to cool the world by three degrees Celsius using geoengineering?"  We would have to implement a certain -- a magnitude of geoengineering and that magnitude would have associated with it certain risks.  So, what are those risks?  And most recent research has been trying to address those particular issues.

I will not go through all of the risks associated possibly with geoengineering or with this rather extreme case of geoengineering.  But, one of the big uncertainties is what might happen to the ozone layer, and it happens that ozone is depleted more [audio glitch] on the surfaces of particles -- or the surfaces of particles are necessary for the ozone depletion in the stratosphere. 

So, if we put more particles in the stratosphere then we might slow down the rate of ozone depletion or we might even reverse the situation where ozone would be increased temporarily and then the slow down would be delayed by many decades.  I do not think that is a serious problem if one considers a realistic geoengineering strategy, but in an extreme geoengineering strategy it is a very real possibility. 

Another area of considerable concern and uncertainty is, "What is the effect of geoengineering on the climate system?  Would geoengineering completely balance the effects of greenhouse gases on the climate system?"  Well, geoengineering is a change in the short-wave radiation balance.  Greenhouse gas forcing is a change in the way of long-wave radiation balance.  And we know that the climate effects -- even for the same amount of radiation change, the effects on the climate are quite different. 

So, we cannot just do geoengineering to completely balance out the effect of global warming associated with greenhouse gases, and we need to know more about what those counterbalancing and differential effects are.  So, those are just a couple of the issues that people are considering in doing research on possible geoengineering strategies.

One of the other issues associated with geoengineering is cost.  And again there have been no definitive estimates for the cost of geoengineering, but it does appear that a significant amount of geoengineering could be done for much less than the cost of mitigation.  So, again that is an uncertain conclusion, but Paul Crutzen estimated that one could offset most of the global warming at the cost of $50 billion a year, which is a relatively small amount.  Again, I said that is a very uncertain quantity, but it does seem that geoengineering -- you know, once we get over the development of the technology to do this, could be relatively inexpensive. 

We do not have the technology to perform these geoengineering experiments yet.  And this is a little profile  -- latitudinal height profile through the atmosphere, and it shows the tropopause and the boundary between the stratosphere and the troposphere there as that -- those little, curved red lines. 

We have research aircraft -- an end car that flies in the upper tropical troposphere and actually in the stratosphere at higher latitudes, if we desire that to be the case.  There are research aircraft that can fly at about the level where people have suggested the injection of sulfur dioxide or some sulfur compound to produce aerosols.  So, we have got the technology available on a small scale, but do we have it available at a scale sufficient to be able to implement geoengineering in a way that will have some substantial effect on the climate system?  So, we are part-way there, but we are not -- certainly not fully there in terms of the technology.

This is a point I made right at the start, and Sam has made it, and I cannot stress this too forcefully, and that is that geoengineering cannot be looked at as a replacement for mitigation.  Mitigation is absolutely essential, if for no other reason than the fact that increasing carbon dioxide increases the acidity of the oceans, and if that trend goes too far then the whole of the ocean biosphere is in jeopardy.  Simply because at the bottom of the food chain shell producing organisms, which require a certain pH level in the ocean to be able to produce those shells, if we increase acidity, they will not be able to produce shells.  They will die, and then the food chain will be disrupted, so that is not a consequence we really want to contemplate. 

So, we have to mitigate.  We have to reduce CO2 levels to somewhere around to 450 to 550 parts per million to avoid this pH-related catastrophe.  So, for that reason alone, we should consider both mitigation and geoengineering as complementary approaches to solving the climate change problem.

I mentioned some of the risks already associated with geoengineering.  One that has been mentioned is that if you are putting sulfur compounds in the stratosphere they are going to fall out.  If they fall out then what did that add to acid rain?  Well, the amount of sulfur that one has to put in the stratosphere is very small compared with how much sulfur dioxide we are producing in the troposphere now -- you know, 10 percent, 20 percent or so, so that definitely is not a serious issue.

Sulfate falling from the stratosphere into the higher troposphere will affect high-level clouds.  And again that is something that has not been really examined at all, at least not in the geoengineering context, but that does not seem to be a very important issue. 

If we inject a lot of sulfur dioxide and produce a lot of aerosols in the stratosphere that would certainly slow down the recovery of the ozone layer.  So, the issue there is just how much of the material we put into the stratosphere and I will come back to that one later.

I have already mentioned the uncertainty and the patterns of climate change.  For example, Alan Roebuck and other people have suggested that if we were to counterbalance global warming exactly with geoengineering in the global-mean sense, then that would cause global rainfall to decrease simply because rainfall is more sensitive to short-wave energy balance changes than it is to long-wave energy balance changes.  And he suggested that the Indian monsoon or the monsoon systems of the world might become less active.  Now, I mean -- I think that is a -- not a sensible suggestion simply because the amount of reduction of the monsoon that might occur is a few percent and the interannual variability of the monsoon is plus or minus 30 percent.  So, whatever signal might occur there is completely masked by the noise of interannual variability. 

And, furthermore, I am not sure that decreasing the intensity of the monsoon would be a bad thing, because flooding is one of the major problems of the monsoon, at least in India and Bangladesh.  I visited Bangladesh one time and the whole country was covered in water and I think a less intense monsoon might reduce the probability of those sorts of events.  So, there are a lot of unresolved issues here, not only in climate science, but also in the implications of these scientific experiments.

I want to say a little bit now -- some background on mitigation before I move into the details of the geoengineering combined with mitigation.  And there are some standard concentration pathways that have been used in the literature for carbon dioxide stabilizing at different levels in the future, and this diagram shows those pathways.  And essentially they are rather arbitrary stabilization levels of 350, 450, 550, 650, and 750 parts per million, and a smoothly varying transition away from a no-climate-policy baseline eventually stabilizing at those different levels.  You can see we are already above the 350 pathway, so that is an overshoot case, and I am going to consider another overshoot case, which stabilizes at 450 parts per million.

Now, what is important about stabilizing concentration is that it is not the same as stabilizing emissions.  In order to stabilize concentrations, we have to reduce emissions substantially and for a very long period of time, and these are the corresponding emissions pathways for those stabilizing levels.  And if you look at the 350 case, you can see that one would almost certainly have to go to some period of negative CO2 emissions in order to stabilize at 350 parts per million.  To stabilize at 450 parts per million, we have to reduce emissions immediately below the baseline -- the no-climate-policy baseline that would occur.  In the absence of policy, if we were to stabilize at 550 parts per million, we might be out a way to a few years to a decade before we had to move away from that no-policy pathway. 

So, these are very, very challenging emissions targets that one would have to meet to stabilize at any of these particular levels.  And for 350, you know, one could almost say that it was impossible; although, there are technologies out there on the table for removing CO2 directly from the atmosphere in small amounts.  So, in principle, if we have another 75 years to wait, we might be able to develop the technology to extract CO2 directly from the atmosphere and have negative emissions.  So, one should never say "impossible" given the ingenuity of the human species.

I might just skip through those numbers and just say something about this stabilization challenge as a concentration challenge, because it is really a technology challenge.  And as Sam mentioned, my colleagues, Roger Pielke and Chris Green, and I wrote a little commentary in Nature that was meant to show exactly what the magnitude of the technology challenge is.

Now, in the literature, there are a lot of no-climate-policy emissions pathways into the future.  They are producing a book called The Special Report on Emissions Scenarios and all of those pathways they -- no-climate-policy.  So, these pathways of emissions into the future that will happen without any policy -- there is a huge range of uncertainty there because it depends on economic growth, population growth, technology and so on. 

So, there is a range of possibilities, but all of these no-climate-policies’ trajectories assume that a large amount of technological change towards carbon-neutral technology, away from fossil fuels, will occur spontaneously in the absence of policy.  And the blue bars there are the spontaneous reductions in cumulative emissions, things that are meant to happen in the absence of policy.  The red bars are the remaining policy-driven changes in cumulative emissions.

And as Sam said, you know, roughly two-thirds of the changes that are necessary to stabilize the concentration of carbon dioxide at around 500 parts per million -- roughly two-thirds are expected in these emissions scenarios to occur spontaneously automatically.  We have reason to believe, and many economists agree with this, that that is an optimistic scenario. 

So, what that means is that the true technology challenge is not just the red bit there, but it is the -- it is from the top of the panel down to the yellow.  The true amount of innovative technology that has to be introduced is really very, very large.  Just how much will occur in the absence of policy and how much will have to be policy-driven is highly uncertain, but it is a huge challenge.

I have just got a couple of quotes here from some leading economists in this area.  Carman Difiglio works for the -- or used to work for IEA -- works for the Department of Energy now, and I will read this out.  He says that, "A 450 stabilization scenario requires a complete transformation of investment in the electric power sector by 2012” -- that is not very far off.  And he quotes from the World Industry Outlook published in 2007, "exceptionally strong and immediate policy action would be essential for the 450 stabilization scenario to happen, and the associated costs would be very high."

And then Jeffrey Sachs, in a nice little piece in Scientific American a couple of months ago, says this.  "  current technologies cannot support both a decline in carbon dioxide emissions and an expanding global economy.  If we try to restrain emissions without a fundamentally new set of technologies, we will end up stifling economic growth --."  And the word "fundamentally" is important -- "fundamentally new set of technologies," you know, that is really putting it out there that there is a huge and generally unappreciated technological challenge.

And another colleague of mine, Marty Hoffert, has said many years ago that this technology challenge really requires a Manhattan Project to develop and deploy the technologies required.  It is like going to the moon again only we are not going to the moon, we are trying to save the planet.  There is no inclining that we are heading in that direction at all at the moment.  And if not -- I mean if you look at China and India, we are basically going in the opposite direction.

Now, many people believe that geoengineering should only be used as a last resort and that raises the question, "What do you mean by last resort?"  And there are two ways to look at that question.  One is the standard way, and that is that geoengineering should be a last resort in the climate context.  In other words, if we realize and see that we are heading towards some really dangerous so called "tipping point" in the climate system such as Greenland rapidly decaying and melting and raising sea-level, if we can see that on the horizon with high confidence, then geoengineering becomes a last resort strategy that we might have to implement. 

The other last resort aspect is the technology aspect and that is one that I think is really important.  If we see that we do not have the technology, we are not developing the technology to mitigate at some safe level of future climate change, then that is also a last resort.  If the technology is not there then maybe we need to think seriously about other approaches like geoengineering.

Okay, so what geoengineering allows us to do is depart more slowly from the no-climate-policy CO2 pathway.  And in that sense, give us more time to develop carbon-neutral technologies that are appropriate for mitigation at a sensible level like 450 parts per million or less.  But, as I have said before, we have to combine geoengineering and mitigation because at least the effect of increasing CO2 on ocean acidity is something we want to avoid.  Okay, so geoengineering should be looked at as a way not to solve the climate problem, but to give us time to solve the climate problem through mitigation. 

I dreamt up some different mitigation -- some different geoengineering scenarios, and if you look at the science paper you will see that this is the same material.  The three cases I considered were where I allowed 30 years to ramp up to some maximum loading of aerosols in the stratosphere.  And I assume that that maximum loading would reduce the radiation balance by three watts per square meter to give you that -- give a context that if we doubled the amount of carbon dioxide, then that is about four watts per square meter. 

So, three watts per square meter is roughly equivalent to offsetting the effect of a doubling of carbon dioxide.  And then, I consider these three different pathways: a rapid ramp up to three, and then a decline back to a lower but continuous level of radioactive forcing.  Another case where I cut off the geoengineering not instantly, but fairly rapidly, and went back to no geoengineering at all, and that is the upper Low Geo case.  And then, I consider one where we ramp up to three watts per square meter and keep that level of geoengineering continuously, and I combine these three geoengineering cases with an overshoot scenario. 

Okay, so here are the carbon dioxide concentration scenarios.  Firstly, the blue curve is a monotonic continuously increasing case that stabilizes at 450 parts per million.  The green curve is an overshoot case where for some reason -- lack of technology in particular -- we cannot go along the blue curve.  But, then given time we develop the technology to get back to 450 parts per million.  And then the red curve is a baseline case where there is no-climate-policy.

And you can look at the emissions consequences of those concentration pathways in the lower panel, and you can see that in this idealized overshoot case then we could imagine technology developing very slowly, so the departure from the "business as usual" case is very slow for the first 15 or 20 years, and then becomes more rapid when the technology becomes available.  So, that is the green idealized curve for the emissions trajectory following the overshoot case.  And if you look at the economics of these things in some crude way, forgetting about the technology development issue, the green curve is economically less expensive than the blue curve.  It is definitely less costly to give oneself time to reduce emissions than to try to suddenly change the whole global energy system overnight to something that does not reduce -- does not emit carbon dioxide.

I am going to consider a comparison of the mitigation case and the overshoot case here.  I will not say any more about the baseline case.  So, what I have done -- I am sorry this is a rather messy diagram, and I will give something simpler in the short while -- but I have a simple climate model where I can put these scenarios in and run them on my laptop and figure out what would happen to global-mean temperature and sea-level. 

Of course, global-mean temperature is only part of the issue and we need to use more sophisticated models to look at the patterns of rain, food change and temperature change and so on.  And other people are doing this, but this is kind of a first pass at the global-mean level.  And let me just pick out -- yes okay, the purple curve is following the WRE 450 pathway.  Okay, that is the case where we gradually approach a 450 stabilization level, but do not go over the top.

The green curve here is where I -- sorry -- no.  The blue curve is where I combine the low geoengineering scenario with the overshoot pathway, right?  So, once we get out beyond about 100 years, you can see that one can very easily get back to the standard 450 parts per million pathway.  This -- the warming is eventually about two degrees, which may or may not be optimum.

You can also see that I have actually got far too much geoengineering in here because I am cooling things -- the globe -- way below the 450 pathway.  For sea-level, sea-level is a very difficult problem to cope with because even with stabilization -- okay, so let us look at the purple curve there, you can see that even stabilizing the level of CO2 in the atmosphere at 450 parts per million causes sea-level to continue to rise for many, many centuries.

I will skip over some of these risks here and give you another simplified scenario.  Now, I just said that those ad hoc scenarios I considered were too intense in terms of the magnitude of geoengineering in the early decades out to about 100 years into the future. 

So, is there some other pathway that we could choose, and what I can do is ask a different question.  I can say, "How much geoengineering do we have to do just to remove this overshoot area here?"  Okay.  And if I do that and then compare that radiative forcing scenario with the ad hoc ones that I considered before, you can see that what is needed just to allow us to do the overshoot is really very small compared with these other scenarios that I considered.

The maximum amount of radiative forcing reduction is only about one watt per square meter, so that is about -- that is less than half a percent of a change in incoming solar radiation.  Sam mentioned that people have been doing studies where they consider a two percent reduction in solar radiation.  So, I am saying that we can actually implement a very useful amount of geoengineering at a quarter that amount of reduction in incoming solar radiation. 

And, furthermore, the peak solar radiation does not occur until the second half of the century and at that level -- at that level there, the maximum -- that radiation deficit is equivalent to the average reduction in radiation that would occur if Pinatubo, a very big volcano that erupted in 1990/91, if Pinatubo occurred every seven years.  Okay? 

So, it is hard to imagine that that sort of scenario would have any adverse consequences on the ozone layer, deposition of sulfur dioxide into the troposphere, climate change or anything.  So, if we do look just at geoengineering as a way of gaining time to develop and implement the technology, if that is our primary focus, then the risks I think are negligible.  But, the positive gains I think are enormous. 

Okay, so I will just summarize now with these few conclusions and a final statement at the end.  That scenario is where the amount of sulfur dioxide or sulfur compounds emitted into the stratosphere has a maximum of about one teragram per year.  And that maximum is a slow ramp up over a 70-year period, and then a very slow decline back to zero over another couple of centuries.  The total amount of sulfur compounds emitted over -- cumulative over three or four centuries is only 100 teragrams of sulfur.  You know, we are emitting 70 teragrams of sulfur per year into the troposphere now.  So, we are talking about less than two years worth of sulfur injected into the stratosphere over a period of four centuries. 

One of the things I really did not dwell on was that the continuous geoengineering case where I ramped up to three watts per square meter negative forcing and continued that for many centuries, that has the effect of stabilizing -- almost stabilizing sea-level.  And I just want to, as an aside, tell you that that is an incredibly difficult problem. 

In order to stabilize sea-level at -- using mitigation alone, we have to go down to 250 parts per million CO2.  We have to go below the pre-industrial level.  And that is even accounting for stabilizing the other greenhouse gases, so not just the CO2 alone, that is a multi-gas strategy.  And the catch-22 of that is that if we go back to 220 -- 250 parts per million CO2 concentration, it turns out that sea-level stabilizes at about 20 centimeters above where we are now, but temperature drops to one degree below where we are now.

So, that is -- I mean that is the perfect catch-22 situation.  We cannot stabilize sea-level -- and this is also, for the geoengineering case -- we cannot stabilize sea-level without going into a climate that was the norm in the 17th century, the time of the little ice age, when the world was about half to one degree cooler than now.  So, there are some really interesting issues there dealing -- that relate to what we have already done -- to be a little coarse -- to screw up the planet in a way that there are -- aspects of the environment that we cannot recover in any way that I can think of.

I will just conclude with these summary points.  And they are very similar to the points that Sam made and it is very good to be on the same page here.  So, I really do not think that we can stabilize carbon dioxide levels at an appropriate point that would avoid dangerous interference of the climate system without either some radical breakthroughs in carbon-neutral energy technology -- and of course, those things might happen -- you know the future of technology is highly unpredictable and we do not know exactly what might happen there -- but I think that it is unlikely that we would develop the technology in a timeframe that would allow us, through mitigation alone, to stabilize the climate of the atmosphere at some sensible level. 

So, that is why I think that we should be considering the science, the politics, the costs, the technology, and all aspects of geoengineering very seriously.  Sam mentioned that -- depending on how you define climate research -- this U.S. government is putting $2 to $3 billion a year in that area.  If we get -- were just to put in another one percent -- an additional one percent -- I do not want to take money away from other important areas of climate research.  But, one percent is another $20 million a year and I think that $20 million a year would make tremendous inroads into removing our uncertainties both in the climate science, in the technology, in the ethical aspects, in the political aspects, and the policy aspects. 

So, I just hope that this project that AEI is developing now is opening the door towards a more serious consideration of additional funds to the community, in all aspects of the community, to investigate this problem in a serious way.  Well, thank you very much.

[Clapping]

Samuel Thernstrom:  Thank you, Tom.  That was terrific.  We can -- I guess I will turn to our discussants in order.  Kerry, do you want to start?

Kerry Emanuel:  Sure.  Tom, before I get into some questions, I wanted to ask you if you might give this gathering a kind of flavor of what sort of technologies have been discussed for -- for example, doing this stratospheric sulfate injection, because I think that helps people actually envision what might happen in a more concrete way.

Tom Wigley:  Yes, I am not --

Male Voice:  Tom, could you just turn the mike on please?

Tom Wigley:  -- oh -- okay.  Thanks.  I am certainly not very knowledgeable about these issues.  However, I do know a number of possibilities are being put forward.  So, one idea in the Academy Report from a number of years ago was firing cannons and shooting the stuff up into the stratosphere.  A bit like H.G. Wells' Journey to the Moon or whatever the name of that movie was.  [Laughter]

Another possibility is very high flying balloons and burning sulfur compounds up there in the stratosphere.  Flying planes high enough and either directly injecting material from tanks of sulfur compounds or less likely putting sulfur compounds into the fuel mix, and then the combustion material would lead to aerosol generation. 

I mean that is not such a sensible idea because the sulfur I think just burns the engines out very, very quickly.  So, you would have these planes flying up and falling out of the sky.  So, all of these ideas are highly speculative and -- but I think that they are amenable to serious investigation.

Kerry Emanuel:  Okay.  What I would like to do is explore with you a little bit about what I considered to be the main scientific issue here -- and that is one that you raised in your talk -- is that we are in a sense fighting apples with oranges here.  That is [audio glitch] that the CO2 induced warming, the greenhouse gas warming, is an effect in the earth's long-wave radiative budget infrared radiation.  We are combating it, potentially at least with the techniques that we have discussed here, by reducing incoming short-wave radiation by putting reflective particles in the stratosphere.

And you can maybe cleverly enough engineer your way to getting a desirable rollback, if you will, in the earth's global-mean and maybe even zonal mean temperature to where you "want it to be," which will be an interesting discussion I think later for the other panels, "Where do we want it to be?"  But, we will not get into that now. 

But, it is not a zero sum game and other quantities, and you mentioned precipitation.  I tried to actually verify the result that you mentioned that global precipitation would be reduced and I could not.  I came up with the idea that it might go either direction depending upon exactly [inaudible].  What I wanted to ask you is, have there been concrete modeling studies yet done to look at geoengineering scenarios?  And if so, what have they suggested about some of these non-zero sum properties of the change? 

Tom Wigley:  Yes -- that is better.  Yes, there have been a few studies and, in fact, a number of years ago my -- one of my colleagues at NCAR, Jerry Meehl and I, and a number of other people, looked at the effects of changes in solar radiation versus the effects of changing carbon dioxide concentration.  And we found that there was a very interesting differential effect in the monsoon area in particular.  And the same amount of forcing -- of solar forcing versus greenhouse gas forcing, did not lead to the same amount of change in precipitation in that particular part of the world. 

Now, more recently in the geoengineering context, Alan Roebuck has a paper which has just recently been accepted for publication where he has used a GODAD [sounds like] model -- climate model -- coupled ocean atmosphere general circulation model -- and tried to look at the effects of adding sufficient geoengineering to offset all of the warming that might occur under some "business as usual" scenario in the future.  So, that is a rather extreme geoengineering case, and he points out in his paper that there is this same differential effect between changing short-wave radiation and changing long-wave radiation on precipitation.  So, he has confirmed that issue.

And there is another very interesting paper by some people at PCMDI, but I do not even know whether it has been submitted, where they -- and I am sure you would be interested in this actually -- where they try to explain why there is this differential effect.  Now, Alan Roebuck also tries to explain that and it is simply because the short-wave radiation effect has a different vertical profile in terms of radiative forcing compared with long-wave radiation effect.  So, somehow it is related to the vertical structure of the radiative balances and imbalances.  There is another -- there is at least one other paper where they have done similar analysis to show again in the model world that there is this differential effect.

Empirically, I think this would be a very difficult thing to demonstrate.  You may know of some work that Kevin Trenberth did where he looked at the eruption of Mount Pinatubo as an analogy for possible geoengineering response and showed that in some parts of the world there were very large decreases in precipitation.  He did not compare those decreases with what the equivalent effect of long-wave radiation forcing might be, but they are sufficiently large that -- I think that that is a reasonable empirical demonstration that there is this differential effect.

Kerry Emanuel:  Okay.  In that connection, you mentioned in your talk some of the risks of undertaking this.  You did not have a side that said benefits I think because we assume that the benefits are self-evident --

Tom Wigley:  [Laughter]

Kerry Emanuel:  -- some of them no doubt are.  But, I wonder if these same studies we have just been talking about actually also show something on the other side of the ledger.  And my own work is in the field of hurricanes, and we know with increasing certainty that the Atlantic hurricane deficit in the '70s and '80s was almost certainly driven in part by manmade sulfate aerosol pollution of the [sounds like] copusphere, so that when we do this, one should be conscious of other perhaps side benefits to the -- do you know of any others?

Tom Wigley:  Well, I -- Yes, I am sure there are benefits like that.  That is a very interesting one.  You know, my perspective is basically that I do not think we can get there from here unless we do some modicum of air -- of geoengineering.  So, that is an enormous benefit.

Kerry Emanuel:  Well, I have one last question, which is more -- it is a question for a scientist, but somewhat philosophical.  As a scientist as opposed to a policy person, what do you think the next steps are in the field of science to try to flush out the viability of geoengineering aside from obviously wanting to take steps to promote research?  Should we move, for example, toward a geoengineering experiment? 

We do not necessarily have to just switch it on and leave it on or conform to the scenarios -- one of the scenarios that you mentioned.  We might actually design a field experiment to do this for a period of years enough to have a detectable effect.  Make enough measurements to confidently quantify what effect particular geoengineering has, and then use that as a basis for designing an intelligent strategy.  Is that a way forward or how would you go about doing that?

Tom Wigley:  Well, I think there is a lot more work that has to be done on the desktop through computer modeling, first of all.  We have only just begun to do experiments using coupled ocean atmosphere general circulation models and one of the problems even with those experiments is the signal-to-noise problem.  And that is, we usually have to force the model with a big signal in order to get something that is identifiable in a clear statistical sense.

We are doing experiments like that at NCAR and have done experiments like that at NCAR.  And as I said, Alan Roebuck has done some experiments.  Other people have too.  But, none of the experiments that have been carried out in the modeling framework have been what I would consider realistic.  And I think the -- a realistic experiment is that that last case that I showed where we do just enough geoengineering to allow us to overshoot the stabilization pathway to something like 450 parts per million.

So, that actually raises a significant challenge in identifying the details of the climate signal and also the stratospheric chemistry and loading signal.  We would have to run a number of experiments, maybe dozens of experiments, and then average the results to get rid of the noise and identify the signal. 

Now, of course, that problem occurs in spades when we try to do some sort of field experiment.  I think, first of all, we need to figure out just what sort of experiment we could perform, where the signal was going to be identifiable above the noise, and where we were equipped to measure all of the right things in order to identify this multifaceted signal that is not just climate, but also stratospheric chemistry and vertical temperature profile signal.  So, we have got a lot of background work to do to decide on a field experiment. 

And a number of people would be very nervous about performing field experiments, but I cannot see any alternative.  I think that that eventually would have to be the case.  In that last scenario that I showed you, I was ramping up the loading of sulfur compounds in the stratosphere very slowly to a relatively moderate level after 75 years or so. 

So, you could consider the first part of that as the field experiment because it would be a very small injection of sulfur dioxide or an equivalent compound into the stratosphere.  So, it may be possible to consider the start of a real geoengineering strategy as the field experiment to determine whether or not we should continue with that.

Vaughan Turekian:  Thank you -- and the science issues related to this, and sitting next to Kerry and Tom, I feel like I can just turn to them and say, "Yeah, what they said." 

One question I would actually have is this question, which either of you started off saying that you do not really have the expertise on.  So, of course, that is where I will start, which is the engineering issue.  And I guess the question that I would have started is, where do you think we will be -- what will we get to first?  Understanding and being able to undertake the engineering or understanding the climate change feedbacks and the implications of geoengineering?  Because it is going to lead to my next question, which is going to get hopefully bridging towards the policy questions a little bit.

Tom Wigley:  Okay.  That is -- well, thank you for reminding all of the audience here that I am not an expert --

Vaughan Turekian:  [Laughter]

Tom Wigley:  -- so you can take what I say with a grain of salt. 

Climate modeling experiments -- chemistry modeling experiments are pretty cheap really.  I think that developing the technology to put sulfur compounds into the stratosphere would be much more costly.  I think both of those things have to go on at the same time, but I can see that there might be much greater hurdles in developing the implementation technology simply because my guess is that that would be more expensive.

Now, if I go back to that number from Paul Crutzen of maintaining some reasonable level of geoengineering -- would only be -- only be $50 billion a year, then I mean maybe we might need to spend $50 billion over a number of years in order to develop the technology.  And then, once we have got the right sort of -- or just suppose we are going to use high flying planes to put the aerosol precursor material in the tropics above the -- stratos -- above the tropopause, then yes.  Okay, how much does it cost to develop one new airplane and -- you know many billions of dollars.  You probably know that better than I do. 

And most of those things are done not to save the planet.  I mean none of those developments have been done with an environmental focus at all.  I mean it is either -- the military -- the Air Force or commercial planes, and commercial planes do not fly up there so there -- maybe the military might be interested in very, very high flying planes and one could spin-off some of the costs towards the military.  But, the bottom line here is that the computer base modeling experiments are really cheap compared with the development of technology.

Vaughan Turekian:  And again the question is though, would we be able to have a better understanding -- how long will it take us to better understand -- and I know climate modelings are the other dismal science in many ways -- but, it is to understand the feedbacks -- the secondary issue is the unintended consequences.  And are there many sort of the -- sort of the -- sort of known unknowns, unknown knowns, and the various things that we will not be able to necessarily understand? 

And I ask this question because one of the issues that comes up when thinking about this issue is that we are now at a time when this kind of activity where sort of climate mitigation often requires the joint activity of many nations and is a global issue.  The question of whether individual actors or individual states that could develop the capability and the capacity due to the engineering without understanding -- to first approximation know that it does cool the climate, but not know really any of the unintended consequences.  Will we -- could we be out in front of ourselves in an engineering pathway, if not necessarily in the United States, in other places to think about?

Tom Wigley:  Yes.  People have suggested that this geoengineering strategy is something that Bill Gates could do by himself, for example, you know, or India could decide to do by themselves.  I think that is an oversimplification really.  I mean I think that the infrastructure for developing the technology is huge, and it probably exists only in a few parts of the world and the United States is clearly one of those.

The computer modeling side -- and I should not just restrict it to computer modeling because you know it requires the availability of observational data, and the interface observations, and model results, and so on in order to test.  So, I mean, I am just using that as a rather simple way of putting this kind of theoretical approach to -- into words.  But, the development of climate models, as you know, has been a slow process.  I do not think it is dismal.  I think it has actually been very exciting and stimulating science, and I am sure that a lot of economists would say that about economic models too.  But, it has --

Male Voice:  You may be wrong.

Tom Wigley:  [Laughter] -- and I am right?

Male Voice:  [Inaudible]

Tom Wigley:  Yes.  If you look at developments over the last 20 or so years, one of the key climate model parameters -- or key climate parameters is the climate sensitivity, you know, how much the world would warm if we double the amount of carbon dioxide?  And the uncertainty there is not much less than it was 20 years go.  So, you might say that climate science has not progressed very far. 

But, then when you look at, for example, how well models are able to simulate present day climate -- the current generation of general circulation models, is very good at simulating present day climate, even precipitation which is extremely difficult to simulate.  And that is a fantastic advance, but nevertheless it has taken decades.

And throwing money at that problem is not the answer because there is a limited number of people out there in this field who can run the models, develop the models, validate the models, interpret the results, and so on.  I mean these models produce massive data sets, and we are talking about even more sophisticated models than exist now -- higher resolution, including stratosphere chemistry, and things like that.  So, there is a people problem, as well.  It is not just money. 

I think we have to have a strategy that is long-term where we try to develop the intellectual infrastructure in order to run these experiments.  And that is something that may take a decade or so to do.  So, I do not see answers -- there will be advances all along the way, but I do not see satisfactory answers coming overnight or even within -- maybe within the next 10 to 15 years.

So, the technology development issue I think is a longer timescale thing.  But, fortunately, I do think that provided we put a big effort into mitigation to the extent that we are capable -- and which is really epitomized by that overshoot pathway -- I mean I think we are capable of doing that.  I think that the technology development challenge there is something that can be met.  Provided we do do that then the amount of geoengineering required is relatively small, and the timescale for implementation is a decade or so.  So, I think all these things are possible. 

Vaughan Turekian:  And I guess I will just ask one more question related to something that I think you both are experts on and definitely, Kerry, you are.  The question of the balance between -- and this is going to get into a world which people do not want to necessarily talk about too much -- the balance between climate geoengineering and weather modification. 

And weather -- and you know people feel weather or are impacted by weather though they live in an umbrella of climate that impacts it.  Is there any room for whether modification in any of these studies, in any of these scientific studies, or is that something that should just be left and moved towards understanding the climate system and geoengineering?  I [Laughter] -- you are the expert on this environment [sounds like].  [Laughter]

Kerry Emanuel:  Well, it is the boundary [inaudible] -- the boundary between weather modification and climate modification can be a fuzzy one.  So, for example, if one of the consequences of climate change is change in the incidents of some kind of violent storm, there will be an impetus to try to do something about those storms. 

And so, weather modification research has had a very rocky history.  In the 1950s and '60s, it was a very respectable endeavor and there were very good scientists, physicists involved in such things as rainmaking and so forth.  Fog disbursal was considered a very respectable thing to do.  There were serious papers in scientific journals about modifying hurricanes.  Some of those papers -- and these are scientific papers, these are not things that occur in UFO magazines -- talked about modifying hurricanes with nuclear weapons. 

And I mean it is crazy, but it shows you that the culture was very different then, then it is now, and particularly in the '80s and '90s, weather modification was not at all respectable -- a scientifically respectable line of research partly because when you do experiments in nature they are uncontrolled.  You do not know whether a particular cloud you seeded would have rained anyway.  It is a very, very tough problem and there was not much success with it.

But what we see, for example, with this rash of hurricanes that occurred in the United States in 2003, 2004, and 2005 was a resurgence of interest, even in the scientific community, of weather modification research and some very interesting proposals had been fielded.  So, I do think that that will become part of it.

And let me take the opportunity to say that one of the problems that we face -- all of us face in even contemplating these issues is that the human race has essentially no experience contemplating problems that affect it on a timescale of a hundred years.  I cannot think of a case in history where such planning on that timescale for some kind of environmental hazard ever took place -- or for any other reason that I can think of, actually.  We are just not genetically programmed to worry about those timescales.

When you look at human history, human beings as an organism go back a few million years, and you look at the timescale of the great ice ages, which really were dramatic climate perturbations, we are talking about timescales of tens of thousands of years.  So, we were well-established as a race during the whole period of the last glacial cycles.

It occurs to me and to my colleagues that on long timescales -- you know if you really want to stretch your mind and think beyond 100 years, if we are going to survive, we are going to have to learn how to engineer climate.  If for no other reason than to prevent us from going back into another ice age, which would occur naturally in a few tens of thousands of years for sure.

Vaughan Turekian:  I am just reminded of the --

Kerry Emanuel:  Yeah.

Vaughan Turekian:  -- of an abrupt climate change report that the Academy did which was co-chaired by Bill Nordhaus, who is an economist, and Richard Alley, who is a paleoclimatologist, and the two of them were trying to define the word ‘abrupt’.  And watching an economist and a paleoclimatologist [Laughter] come to terms with what was ‘abrupt’ and what timescales that was is fairly interesting.  So, it speaks to your point exactly, that 100 years is probably -- actually was the compromised candidate there.

Kerry Emanuel:  I have a number of questions of my own, but we have been talking up here a long time and I am sure the audience has a great deal of questions.  If -- is someone from conferences here with a microphone for us to start taking those questions?  Well, perhaps not, but we will proceed nevertheless.  Maybe Laura can try to run down a microphone. 

Why do not we start right over here, sir, while we run down a microphone?  Please just, for all of you, could you just identify yourself as you ask your question?

Martin Apple:  My name is Martin Apple [phonetic]. I am from the Council of Scientific Society Presidents.  Ultimately the population will grow over the next 50 to 100 years [inaudible] and an [inaudible].  Right now we have gathered a lot of that food that we cannot grow on land any more from the ocean.  If we geoengineer a partial solution to what we are doing, we will be simultaneously pumping more CO2 into the ocean.  That will cause the pH to drop and we will begin to end life in the ocean.  So, essentially we are going to create a conundrum here of being able to temporarily worry about the temperature at the expense of things that otherwise are much more important to us.

Kerry Emanuel:  Tom, do you want to speak to the ocean acidification question?

Tom Wigley:  Yes.  From what I know of the literature on the subject is that the ocean acidification problem becomes really serious once we pass about 700 parts per million.  Okay, I may be wrong there, but I think that if we were to keep the level of CO2 in the atmosphere below 550 parts per million -- and that overshoot case it goes up to 530 -- then, yes, the pH would drop by another one or two-tenths.  But -- and there is a delay too in that response -- in the ocean too.

But, I do not think that will be a serious problem for the ocean biosphere as a whole, but there may be some coral area, some places like that, that would even be damaged by two-tenths of -- a further two-tenths of a drop in pH.  And coral ecosystems are very lively and productive and we cannot afford to lose any of those either.  But, of course, they are affected by temperature too and there is going to be a bit of a temperature overshoot if we follow that pathway.  So, we are jeopardizing some of those ocean ecosystems in two different ways at the same time.

So, these are very serious issues that there have been some reports on -- Academy Reports and so on.  It is just not really my field of expertise to be able to give you more than that kind of general impression.

Kerry Emanuel:  If I can just make one general point in response to that myself before I take the next question.  You know ocean acidification certainly does seem to be an important question, but it is not an issue, as I understand it, that geoengineering would exacerbate in any way.  It is simply an environmental effect of high CO2 concentrations that geoengineering does not resolve.  And so, that would certainly be one of the reasons why a combined geoengineering and mitigation strategy is important.  But, it is not an issue that -- as I said, that geoengineering exacerbates as I understand.  So, maybe over here, sir, in the yellow shirt.  Okay.

Michael McCracken:  Mike McCracken [phonetic] from the Climate Institute -- excuse me.  First, Tom, I think 750 is way too high for the coral.  The [inaudible] that the coral community puts out say that the CO2 concentration in 2050 changes the chemistry.  So, it is not basically acceptable for coral given the experience with distribution.  I mean you are already seeing changes in the compensation depth and people in northern waters are already worried about marine kinds of things.

I wanted to ask though a separate question and that is, why the focus so much on stratospheric aerosols?  I mean, all the model simulation models to date are assuming uniform change in radiation sort of over the planet -- almost all the planets, which is going to be very hard to get injecting SO2. 

SO2 and sulfate, which also turn a lot of the direct radiation to diffuse radiation are going to make solar technologies more problematic by tenths of percent.  I guess -- why is there not more thinking about Angel's [phonetic] proposal of trying to put aerosol up in amounts not to balance two times CO2, but in looking at that overshoot that you were talking about Tom?

Tom Wigley:  Yes, that is interesting.  I think that issue of diffuse radiation of solar energy is still slightly up for grabs.  But, I mean qualitatively, it is an issue that needs to be resolved.  And I -- the overshoot scenario that I have -- that I consider at the end of my talk really just has negligible -- I mean the issues you raise are just not important in that particular case.

You stated at the start that you thought the experiments that have been done with models uniformly changed the radiation balance at the top of the atmosphere, and it is true that the earlier experiments did do that.  But, Philip Rasch's experiments with an NCAR model were not like that at all.  I mean he injected sulfur dioxide into the 10 north to 10 south band in the stratosphere, and then allowed it to disburse, and the same applies to Alan Roebuck's experiments. 

And Alan furthermore did an experiment where he just injected the material in high latitudes, as well, and looked at the response.  And in that model there is chemistry in transporting the stratosphere that changes the nature of the aerosols and moves them around.  So, at least two recent experiments have not been quite so oversimplified as you implied.

Kerry Emanuel:  Fred.

Fred Iklé:  Fred Iklé -- a question to Dr. Wigley.  If I understood correctly you pointed out on the sea-level problem to -- that somewhat under control -- rising sea-levels, we might have to lower the temperature that might almost bring us back to the little ice age.  Is it not possible perhaps to do some of the geoengineering directed for that objective and focus it on the Arctic and Antarctic, that you can do regional geoengineering, which would not bring the ice age for the other parts of our world?

Tom Wigley:  Yes.  I must say, I only did those – that extreme stabilization experiment on my laptop a couple of days ago.   And I mean I could have guessed that something like that was going to happen because you can actually see it if you look at the results in the Nature paper, but -- you know, where sea-level does not stabilize and yet the temperature of the globe goes below the present level.  So, there is an inkling of this catch-22 problem there.

Now, realizing that that is a problem in a very simple model at a global level means that we have to think of -- you know, is there some way to get around this?  And you are suggesting there might be and I do not know the answer to that question.  But, I think it will be fun trying to figure out whether there is some way to get the best of both worlds. 

Kerry Emanuel:  David over here. 

I always like it when questions are asked that we do not actually know the answer to.  It keeps us thinking.

David Schneer:  Thank you, David Schneer [phonetic] from the Center for Environmental Stewardship.  A couple quick comments, the question of papers that are out, there is a new paper that has been submitted to Nature by a long list of people -- Ken Caldera is one of them -- that addresses several issues, including precipitation and temperature modeling that includes the scenario of only doing the polar regions.  It also takes on some of the precipitation issues that Roebuck's argued and I think pretty much put some of those, not to rest, but it does give a different view of them and they are much smaller in the way of effects. 

I would say this about risks -- there is a question at the end of this -- models on risks use the same models that are used to generate estimates of climate change.  They are all the same models.  And so, when you are trying to ask the question, "How soon will we know about risks?"  It is sort of like saying, "How soon will be know what is going to happen to the climate?"  If we think we know enough to take large economic steps because of climate change today, then presumably we have enough confidence in those models to say we can also estimate what the risks would be of modifying the climate.

So, the question that I had -- went back to an early chart that you talked about Professor Wigley, the stabilization target of 450 parts per million CO2.  My reading of the AR4 report and the background work in Working Group I was that it was 450 parts per million of CO2 equivalent, which we have already passed.  And so, I just want to clarify whether my reading of that was in error or whether, in fact, we are already beyond the 450 number.

Tom Wigley:  Okay, yes, that is a very good question.  Now, the concept of CO2 equivalent is basically this, when asked the question, "How much carbon dioxide would there have to be in the atmosphere to give the same radiative forcing as the full mixture of anthropogenic gases?" 

Now, if you put sulfate aerosol effects in then we have not past 450 yet, right?  Because the -- if you just look at the greenhouse gases alone, then we have passed 450.  I do not know what it is -- 470 or something like that, but there is negative forcing due to aerosols and if you put that in then the present equivalent level is pretty similar to where we are for CO2 alone.  The aerosol negative forcing is highly uncertain, so that is a highly uncertain number where we are. 

Now, in my -- although I only showed results or numbers for carbon dioxide, and that was true, that -- project that stabilization scenario for concentration is carbon dioxide alone.  But, the experiments that I have done include other gases and, in fact what I have done -- you know there is a CCSP report where they have considered three different modeling groups. 

They have considered a number of different multigas stabilization scenarios and in one of those scenarios they only go to 2,100.  But in the ones done by Jae Edmonds  and Steve Smith and people, we extended it out to 2,400 -- or 2,300 actually.  And I have used one of those to account for the effects of other gases.  So, although I am only showing CO2, I am not talking about CO2 alone.  I have not interpreted the results in terms of equivalent CO2, but I have not ignored the other gases.

David Schneer:  Thank you.

Kerry Emanuel:  Alan.

Alan Carlin:  I am Alan Carlin [phonetic] from EPA.  I am somewhat -- I am quite a bit less optimistic then you are that we are going to stabilize at a 450, 550 number.  I do not see the forces that are going --

Male Voice:  [inaudible]

Alan Carlin:  -- if that is true I would argue that doing a [inaudible] using geoengineering may be really only part of a larger approach, which would combine both mitigation and geoengineering.  I wonder if you have any comments.

Tom Wigley:  Yes, I can add a little bit to that.  Yes, well, it is nice to have a little bit of pessimism thrown in at the end here.  Andy Revkin has something in the Times or on his blog recently where he was talking to Sherry Roland and Sherry says, I -- and I am misquoting him here, but he says something like, "Well, I think that we are heading to 1,000 parts per million."  Well, boy, I certainly hope not because I think that would be really close to catastrophic, not just for the climate system, but for the ocean as well.

So, yes, I do not see the signs for stabilizing at 450, 550, 650, 750 -- I do not see any of those signs at all at the present.  But, the globe is warming -- well, it has not warmed that much over the last seven or eight years, but there is a long-term global warming trend that is inexorably continuing into the future and as that happens then I think even the most recalcitrant of policymakers will start to realize we have got to do something about this.  So, I am not quite as pessimistic as you.  Even though I agree, I cannot see anything out there happening right now.  So, --

Kerry Emanuel:  Well, I would like to wrap this up reasonably soon, so the next -- we can take a short break between the panels.  But, we could probably take two or three more questions.  I see one over on the side over there and one in the back over there.  Why do not we start here?

Alex Echols:  Alex Echols [phonetic], Philanthropy Roundtable.  I have been in and out, so you may have covered this and I apologize and dismiss me.  One of the concerns I have heard about this has been the impact on photosynthesis and what do we know about -- will it be significant?  And if it will, in particular, will it be significant to agriculture and food production and oceans and microorganisms?

Tom Wigley:  Well, again, totally outside my area of expertise.  And I think Mike McCracken has probably thought a lot more about this than I have.  But, I just want to say again that the final scenario that I considered would have a negligible effect on -- I believe -- on plants.  Whereas some of the more extreme cases where we try to offset with geoengineering, a warming of say three degrees Celsius, you know, the two times CO2 warming, it is quite possible that that would have effects on vegetation.  But, again, I just cannot say one way or the other simply because I do not know.  But, it is something that is an important consideration.

Alex Echols:  Fair enough.

Kerry Emanuel:  In the back here.

Santiago:  Yes, my name is Santiago, so I am with the University of Maryland and NASA.  A question I have is about GCMs and the -- and my concern mostly is about public perception.  Perception of how we scientists deliver information from GCMs.  And I understand the use of them and why they help us to get a big picture and give us the constraints that we are going to be living on in the future. 

But, what -- we see very little about regional impacts of the future geoengineering scenarios and actually [inaudible] warming.  And I think by focusing on more of those aspects of what is happening in the first two kilometers -- bottom two kilometers of the atmosphere, is where people will start to catch what is the real impact of these effects.  And I wonder what the panel thinks about why we do not see much of those regional studies of this geoengineering and [audio glitch] scenarios?

Tom Wigley:  I can give you a quick answer and that is that we are just starting to do global GCM studies.  And to do regional studies, one of the approaches is to embed a high resolution regional model within a global model, so that it is driven by boundary conditions that are generated by the global model.

Now, if you look at GCMs, they differ in their responses.  So, to get down to doing regional impacts of geoengineering, the first thing we need to do is have a number of global models that we can use to drive a number of regional models to get some idea of the inter model uncertainties both at the global and the regional level.  And hopefully those studies will be done over coming years.

You know, for example, I imagine this impinges on the issue of whether geoengineering is going to do anything to tropical cyclones and maybe Kerry Emanuel can say something about that.  I think the details of the monsoon, for example, we really need to understand that better by imbedding a monsoon in Southeast Asia and India’s monsoon region -- a regional model within the results of the global model because the aeorography is very extreme and the land ocean contrasts are really important.  And we are still at the bottom end of the learning curve in that regard.

Kerry Emanuel:  I would just like to reemphasize that I think climate modeling is not today far enough progressed to answer your very important question of the regional effects.  If you look at the differences among models, and even among ensembles of the same model, they are pretty profound when you start getting down to regional levels.

All we can do today -- and even this I think is a bit dicey, is to bracket the range of possible outcomes on the regional scale.  And this is something that I think we all need to emphasize, is that although there has been a lot of progress in climate modeling and climate science, it is still a field that is relatively young.  We have a long way to go.

Samuel Thernstrom:  And I would like to wrap it up, so we can all take a break before the next panel, but I do not want to leave anyone hanging.  Is there anyone who has a specially burning question? 

Male Voice:  [inaudible].

Kerry Emanuel:  We will take one last one over here from Raffe [sounds like] and then we will --

Raffe:  I just like your comments on the issue of masking.  And what you might -- is there anything we can learn from the fact that copuspheric [sounds like] aerosols are masking a large amount of warming at the moment [inaudible] the sulfates at a lower altitude?  And is there anything -- I mean it is kind of a -- we masked a lot of warming it appears over the last several decades and is that -- in a way it is geo -- an inadvertent geoengineering experiment?  And what I am wondering is if it suggests anything about the stratospheric issue itself or is it utterly different?

Tom Wigley:  Oh, boy.  This really impinges on a field of analysis called "detection and attribution" and when we try to understand what has happened to the climate of the globe over the last hundred years, then what we do is we try to get some idea of the expected signal from greenhouse gases, some idea of the expected signal from aerosols, and so on.  And we can do a sophisticated multivariate regression analysis to back out the relative importance of the different contributing factors to changes in tropospheric climate. 

And that is quite a difficult thing to do and just to give you one idea of the difficulties involved, people have included aerosol effects in those analyses, but they have not included indirect aerosol effects.  You know, they have only considered the direct effect of aerosols.  So, what we do is we look for a pattern of direct aerosol climate effects in the observation record when we know that indirect aerosol effects are probably larger.  So, we are hoping that the pattern of indirect aerosol effects are similar to the pattern of direct aerosol effects.

So, it is a very difficult statistical problem partly because even though there has been a global warming of say seven or eight tenths of a degree, at the regional levels the signal-to-noise ratio is still not large because at the regional level, the noise is very high.  And in order to properly identify cause or factors, we have to look at spatial patterns.  We cannot just say, "Well, the world is warm and that has got to be carbon dioxide."  We have to look at the patterns of change and try to quantify the relative effects like the masking effect of aerosols in the troposphere. 

Now, I think one of the big problems is that if we were to perform a field experiment in the stratosphere, then one could apply the same detection and attribution approach.  If we knew what composite signals we were looking for, then we can get the observations and see if we can identify those signals, particularly the geoengineering signal, in the stratosphere or in the consequent changes in the tropospheric climate.

But, if we are considering a slow ramp up or a small field experiment, then the signal we are looking for is incredibly small and the obfuscating effects of other factors will be really difficult to overcome.  So, I think there are some supreme statistical challenges in store for interpreting observations from field studies in the future.  And they relate to this issue of masking where different effects are compensating each other in complicated, spatial pattern ways, so tricky stuff.

Samuel Thernstrom:  I am sure we could keep going for a long time, but I know that Tom at least has a -- and Kerry actually both have planes to catch.  So, why do not we call it quits here for this panel and we can take a break for a few minutes and then Panel Two will start shortly.

[Clapping].

[Start of File:  AEI9473-PanelTwo.mp3]

[Break in audio 1:50:00-2:03:24]

Lee Lane:  Welcome.  Thank all of you for staying around for our second panel, and I am sure we will have a very interesting discussion following up on the interesting discussion that we just