 This is an unusual lecture for me because I'm going to be speaking to you about something which I first of all feel passionately about and secondly, which is moving very rapidly in the work I'm doing on it with now with a group of very unusual partners that I'm going to describe to you about halfway through it. But I think it's important and I'm going to try and leave you imprinted in your minds with my main messages at the end and I'll follow that up with the deficiencies of the kind of approach this is. So with that let me begin and let me see if I can do this right. Apparently not. Yes. So here is my summary of where we are today. After the allegedly dramatic success in 2016 of Paris, we had two advances. One was a change in the target for temperature at 2100 that the world should not overcome, overexceed, that is three degrees, moving that down to two degrees, and hopefully being able to get it as low as 1.5 degrees. You could not exceed that number over industrial, pre-industrial times by the end of the century. 191 countries more or less agreed to take on independent national limits for their contributions, for their reductions. In that period 100 nations agreed to make voluntary commitments. None of them are held firmly. To the year 2020 and 2020 there would be a revaluation that is this year, I think at the end of the year, every five years in order to reduce emissions and they would do this progressively so that hopefully at the end of the century we would have met the target of not exceeding three degrees temperature. Now the results of that people looked at it, both people who were there and others, certainly great advocates of climate reductions and realized that first of all the volunteer contributions are not being met in a very large part of the 191 members, 160 of whom I think actually filed voluntary reductions, and they will not be met according to the annual report put out by the UN about how are we doing it meeting those voluntary reductions. And I'm going to report to you that the United States is of course one of the countries that will not meet its commitment and in part the reason it won't meet its commitment is because we've had a very successful period of years of very good economic growth which is bad from the point of view of emissions. That's one of those awkward truths. So it's realized that even if everybody did it and even if we had gotten rid of all emissions by mid-century which happens to be one of the goals which is being adopted by a large part of the world especially for example the European community, it is not the case that we would avoid we could meet the 2100 target. We would still exceed the 2100 target for warming. So emissions cannot do it alone. Emissions cannot meet the targets that we have set by the end of the century. So there are additional control mechanisms. One of them is in fact very sharply mentioned in the Paris Accord by the quite good summary statement by the chairman, the French chairman, saying we have to do more about adaptation. We have to place greater priority on adaptation that is taking measures, human measures now which will avoid, reduce the possibility of temperature of the consequences, the consequences of temperature increases. The biggest example of that is to build dams, I mean to build dykes around let's say Miami in order to stop the possibility of cities, coastal cities being undated by what rising coastal waters. So adaptation is one very important climate control measure that could help deal with avoiding the impacts of climate change. Now a lot of people talk about adaptation. There are very few programs, there is no program in the government of the United States for adaptation. So it is not a measure, a control measure that is really being pursued vigorously and importantly there is not much research or analysis going on on adaptation. Now it turns out for me adaptations are particularly attractive possibility because you can marry it to infrastructure improvements and put on its, take as a collateral benefit to adaptation not only that you have built walls or built different kinds of support for buildings but also that that provides resiliency for the structures and protection and improved use of the buildings, roads, ports, highways across the country. And then there are two types of control measures which I link together and call geoengineering. One of them is efforts to remove carbon from the atmosphere directly. Let me just use it as an example, direct air capture but there are other agricultural methods like growing, growing forests and a lot of attention is being paid to that negative emission technology since the reference goes by. A very good report was put out by the National Academy on this subject about what it could be done, how much it would cost to do the development for a negative emission technologies and what was the prospects of deploying these over the next few decades. So that is a climate control mechanism which is available with adaptation and of course with emission reduction to try and do more to reduce the emissions by the end of the century. And the third is a very difficult, very controversial geoengineering measure. We heard about it here last year when David Keith, a professor at Harvard who is probably the most well-known advocate of this approach or at least thinking about this approach which is basically a variety of different techniques but I'm going to use one as an example, one which I think is the most likely, most promising technically. You take sulfate particles, you put them at the top of the atmosphere, the atmosphere, what those particles reflect, a bit of the sunlight that changes the balance of energy coming in and out of the atmosphere around the earth and it reduces the temperature from what it would otherwise be if the full solar insulation was hitting the top of the atmosphere. But there are lots and lots of problems with the solar management of that kind. Who's going to decide what you do? What are the other consequences of it besides just allowing the temperature to be slightly warmer, slightly lower radiation forcing? Many, many questions about how the decisions, who's going to do it? How would it be carried out? All of these are questions which are highly controversial in what especially concerns people. If it becomes attractive and cheap, then it will begin to tell people will stop spending money on the difficult and expensive emission reductions and put it all on this inexpensive, allegedly inexpensive and easy to perform technology. But nevertheless, it is a candidate for climate control and if we're serious in a university, and I'm difficult to say it, but Stanford is a serious university, you really have to assess what the costs and benefits are, what the problems are, social, political, as well as technical problems of particular technologies, especially if they come up in critical times in critical areas. And here's a perfect example of it. There is no subject taught on geoengineering at MIT. There's no subject taught on geoengineering at, what is the name of the school? Harvard. There's no subject taught. Because of this difficulty that scientists properly see of managing this technology in a responsible manner. Nevertheless, I say if you want to have a serious appraisal of what needs to be done in order to keep global temperatures below 3 degrees warmer than pre-industrial times, you have to look at a combination of these and try to assess. I'm going on too long. Here's a summary of how they go. The emission reductions tries to lower Q of t as the emission rate. They're in the y-axis. It reduces emissions over time. Removal, when it begins, there are no operational deployed, large-scale CO2 removal technologies. It would actually take emissions down. And eventually, by 2100, hopefully you'd get to a situation where the whole globe was at a net zero balance. Okay. Now here is the joint conceptual model that I wanted to present to you, which tries to say, how do we think about these four climate mechanisms simultaneously? My first and most important point is that there are four, not just one, not just emission reduction, they're four. My second important point is, if you want to be serious as a person who is an analyst here, you have to face what are the four, what are their costs and benefits, and how do they work together? So what I do is I take a very, very normal starting point of a damage function. What is the damage to the whole economy for a temperature increase here, denoted delta t? There's a lot of discussion about what the shape of that damage function is. It's a function of temperature increase. Let me say, as a chemistry professor, it is stunning how many of these equations which are used in climate, and especially the economic analysis of climate patterns, where there is no numerical support for any of the function's parameters. They are all parameters for which there is no field of data to support them. Here's a perfect example at low temperature deviations from industrial times. It's supposed to be a quadratically increasing function, that is, damages done to our economy increases a quadratic function of the temperature increase, but there are many, many quarrels about what its shape is for larger deviations. And then there's this one key equation which you don't have to do anything but look at, which I'm going to describe to you right here. Here's the damage function, but now with the four climate control variables and what their effects are. First of all, in front of the damages is a function 1 minus chi of t, which is the fraction of the damages, which is reduced by adaptation. It lowers all the damages, whatever the temperature increase, because you're really buying insurance against any effects. For example, those fikes which are going to build in front of Miami will have to be, let's say, 60 feet high, because that will be sure that you don't ever have any water get into Miami from rising sea levels. I don't think that about predicting that. I'm saying that's the limit that that variable takes at value 1. At value 1, when chi is 1, there are no damages, but you're talking about very, very massive investments. The intermediate term, delta t squared, the quadratic term for temperature exchange, has in it two of the control variables. One is phi, which defines the fraction of the flux of CO2 into the atmosphere that is reduced. Annually, there's a flux Q of t and phi is the fraction of that that is reduced. Then there's the fraction phi, which tells you every year how much you take of CO2 out of the atmosphere, how much CO2 is taken out of the atmosphere. Finally, we have to do something, which is not in here. Yes. Finally, you have the effect of geoengineering. It is quadratic. Here it is right here at the end. Quadratic because geoengineering covers the whole hemisphere, all the area underneath it. It goes like lambda squared. Those are the four variables that together will decide how much you have reduced damages from what they would have been if there had been no climate control measures at all. That's the model that I'm investigating. There's a star up here. The star up here reminds me to tell you two things. They're very recently, I think, on people's minds. They may be on my mind, but they're not anything that I've been able to make progress on. The first is delta t tells you the change in temperature is supposed to tell you what the damages are. But the mechanism by which the temperature increases causes damage are multiple. They happen in many, many different ways. They influence agriculture. They influence temperature and health. They influence a whole series of things. So the effects of temperature are what cause the damage, but there are many different avenues, not one single road, which is easy to display analytically or to determine its quantitative characteristics if you're trying to anticipate what might happen in the future. The second is a little bit more subtle. Here the reasoning is one directional. If there is a change in emissions, it causes damage in the operation of the economy and the society. But it is also the case that when the damages go up, they may have an effect on how well your economy can run. So to take the example I was just mentioning before, if the temperature goes up, more people have heat strokes, less work gets done, the productivity of the economy goes down, but so does the flux of CO2 that you put into the atmosphere, so that there's an interplay, an interaction. If you like a loop between both the damages and the cause of the damages, which is not included in the models, certainly not included in most of the models that are not included in the models, which I'm familiar, but that interaction, that interplay is something which I think deserves to be pursued. And this is what economists formulas are just for the, I just try and be cordial to the economists. The two measures are the discounted damages that you have, discounted at some discount rate. I met with some students at lunch today, there was quite some sharp discussion about what that discount rate should be and there has been for some years. Or else the social cost of carbon, which I put down here in mathematical terms, which is what the U.S. government does to try and give an indication of what one extra kilogram of CO2 added to the atmosphere, what that cost is for the economy, the social cost of carbon. It is the quantity which is most, based on a few models, most generally used to determine what you might want to have for an emission charge or a tax on carbon. So these are the costs we had. What do we have to do to get some numbers here? We have to first of all think about how the concentration changes, C of T, in the atmosphere over time. We have to think about what the connection is between the concentration, we have to think about the connection between the temperature and the concentration, which is a logarithmic relationship with a parameter called the equilibrium climate sensitivity. And we have to do those integrals knowing prospectively, or estimating prospectively what the emissions are going to be as a function of time, Q of T. And then through a series of not so complicated arithmetic, one comes up with a solution for actually determining what the damages will be with whatever climate control variables, the four of them that you have in mind. Here you have the effects of both emission reduction and CO2 removal. You also have to add the geoengineering and then the adaptation, and that's what you do to get an effect, to get an estimate in the simple model of what the effect of these four variables acting together will be on reducing the damage. I'm going to go and tell you about one calculation, if I can get to it. Oh, you need to know a cost curve. This is the cost curve, and since there are no numbers for costs, I don't understand these integrated assessment models. Is that what they're called? They're called integrated. There are costs in there, but nobody knows what those costs are. There's no empirical field data. Here I take a very simple case. There's the budget that I'm going to spend for reducing, for putting in these climate control measures in order to reduce the damage. As for illustration, I'm going to choose a very simple example. All the costs for the four different variables are a certain amplitude, always by the value of that climate control variable squared. For those of you who want to mean to me, I will tell you I choose a convex form for that cost function. That means it increases as the variable increases. If I don't choose a convex function, it becomes what we would call in my world a mess. So it's going to be convex so I can give you nice and true results. And they're different. Who knows what those costs. So here I choose for the emission reduction, what we know most about, I say the amplitude is going to be 1% of GDP, 1% of GDP in this budget, which is the relative amplitude compared to the other. Then I have the unknown, but certainly very promising and very interesting and necessary carbon removal technologies, these net zero or negative emission technologies. And I say I don't know how much that's going to cost, but I'm just going to say it's double. Now the third thing is what about the cost of geoengineering? I'm sorry, there are two churches here. Church one says it's very cheap, and the other one says, look it's a tremendous danger, they're huge uncertainty, so in fact it's very expensive. So I'm going to look at two cases. It's 1.5% or 1.5% what emission reductions are, and I'm also not on the slide, I should have put it there. Another case which it's 5%, so it's five times larger, and these are all percents of global GDP at that year, think $90 trillion, and here are the results. Now these are results done by this hand, and excuse expression, this mind. I actually did these. This is as complicated a case as I can do, but on the next slide you can see that I have armed myself with vast amounts of technical help. What is done here is, there is the damage which happens for an interesting policy case. In Paris they said it's going to be three degrees and we're going to move it down to two, and the question should be raised, how much is it costing society to have made that decision? And so this calculation says over the period from 2020 to 2100, every year the emissions will be the same, two parts per million by volume of CO2 into the atmosphere. You might say to me, John, aren't the emissions going down? The answer is no, they're not. Worldwide emissions continue to go up, and at best they look like they're steady. So for this case, I chose it's constant for the whole time period. The fact that it's constant is what allowed me to do the calculation by myself. And then what you do is you say now choose those values of the variables which will minimize that budget to accomplish the reduction that was agreed on at Paris. So this is a minimization problem, an optimization problem. Tell me the right combination of variables to achieve the policy result you want moving from three degrees to two degrees max in the year 2100. And the answers are in this little table here. Undiscounted, over the whole 80-year period, there was about 100, undiscounted, about 100 trillion of damages. Now that's 80 years, about 90 trillion a year of GDP. So it's not a hell of a 1% more or less. If you discount it, it's 35 trillion. This is with the high cost of solar radiation management, but that's with the one which is 5%. And here is with the low cost, it's 42%. Not everything is, of course, done by solar management. Some of it, in the low cost case, of course, you use more. It's lower cost. The total cost for that is lower. And then if you discount it, you get slightly, you get much different numbers. And then underneath are the budget required to achieve that, these differences. And you can see that they're now in billions, not in trillions. And over an 80-year period, they're certainly affordable. You can get your political system together. Certainly affordable to actually, if these cost functions were known to be true, and if the cost amounts that you had selected were true, they're perfectly manageable. You say, look, that's not the way the world is. We expect that the flux is going to go down. In fact, we want it, and there's a law, and what's the name of that state? What's the name of the state? New York, that's the name of the state, where they've actually passed a law saying that we will have only 50% CO2 emissions by the year 2050. And they're fiddling with that kind of a law in California, but it's not quite a law yet, right? It's not a law. But in the European community, it's a law for the whole community. So they said, you better go and know how to do this if you have a very variable rate of emissions. And in fact, whether it's through good sense or whether it's through mandatory regulation, you actually drive it down. So then I had to move on and find some talent. It's difficult for a man of my age who's been married, is married, to say that he's fallen in love. Well, I have fallen in love with Julia. It is the most stunning. I mean, it is so stunning everybody here should go home and learn Julia. It is the most stunning programming language I have ever encountered. And with the nicest and most talented people I have ever encountered, including economists. So these guys came to me and said, let us help you. It is unbelievable what they can do. There is no simulation, no optimization problem that I can't, that I can think of, that they don't immediately say to me, quarrel among themselves, who can do it most quickly? Will it be milliseconds or seconds? It's certainly not going to be longer than that. These are great guys. So one of them is Ron Revest, who is a quite distinguished, I mean very distinguished computer scientist, best known for his role in the RSA encryption algorithm. The second is a guy named Alan Edelman, who is a professor of mathematics at MIT who actually has invented this programming language and developed this worldwide army of charming talented people to go around and do it. And the third is an atmospheric PhD, an atmospheric sciences PhD student, Olly Drake, who is a coming most valuable player in the academic world. And these guys are on me every day to push these calculations. So now I'm going to show you the first trosh of how these calculations come out, just as an example. I can't tell you, I mean how they regard this as a elementary trivial example. Here they said we're going to, we're going to, yeah, we are going to look at, they feed me so many examples I frequently forget. So we're now going to take the emissions that we hope will be the case. We say they're fixed now at two, he did four, I guess he did something. And then they bring them down to zero and they remain at zero. So the emissions here are cut off. And he says this is the blue curve, this is his base case, just as I described it too, except my was constant, he's, and he's extended it to, he's got to extend it because they don't leave, CO2 stays in the atmosphere so long, to 2,200. And then he says how did the concentration behave? And he sees here in the uncontrolled case, no policy, what happened is the concentration went up until it stopped. And then it stayed constant, did not change because there were no, there were no control measures in place. Now he says let's go and optimize what we have to do in order to get to two degrees by, what was this, I forgot exactly what happened now, I forgot what its optimization function was. But the result was that he changed the control pattern, much lower controls. The concentration went way down and then stabilized at 450. The optimized control deployments were very different. In this particular case you could see that emission reductions were the most, then negative emissions were the blue line. And just a, just a tad of both adaptation and geoengineering, or I should say solar radiation management. Those are the optimum, optimum results for the lowest cost of meeting whatever the objective function was. So there's a dashed line here which is very important. These four control measures do not have the same technology readiness. Some of them are here, you can use them today, emission controls. Some of them are not here today, negative emission technologies. And some of them we hope will never be here today, geoengineering, solar radiation management. But you have to say here's how quickly you might bring them on board. And that's put here as a constraint on all four and you see the result so you can't go up too quickly. Here are the costs in the bottom frames. These are the costs of each one of these, the cost of the climate control in billions of dollars. And this is the result in warming that happens if you have no controls you go to four degrees. But with the controls you come out between two and 1.5 degrees. But notice, and keep deeply in your heart, all four were necessary. All four contributed. The amount that they contributed depended upon, of course, their relative costs, the sea hats. But you can't do this with, and in fact, I should have kept it. If you say how much would it cost to do it with emission reduction, alone, you can't do it. So the main point here is four is better than one and one is not enough. But remember, I whispered to you, whispered to you, this is where concave cost functions. So we know that these common minimums exist. If you don't have that, then you're going to have quarter solutions, people bumping into each other, it'll be a mess. So this is more and more on my mind. Because now with my friends from Julia, I really feel I can do anything. But then I have to come back and say, well, what haven't I done? And I give you here a list to remember from any speaker, not only me, the things that these climate models really don't have in them. And this is what I want to underline very much is that it is certainly the conceptual model that I've given you has no quantitative merit. But the more complicated models are not much better. And they do not address many of these points that I want to draw to your attention is being on the alert when we try and talk about the complexity of real world global climate systems. I must say I'm very much down here. All of this to me is principle, theory, important. But something I'm not good at dealing with. But down here, that's what really, when I really, who the hell in our government is good enough to do this? Who has even got the general intention or the tools to get a program like this done? And I tell you, even if we had good numbers, even if we had reliable models, this is not something that you could just let the US Congress with its many committees and many different appropriation committees, appropriating different amounts of money for different purposes every year. And meanwhile, we have to remember, we're not only trying to deal with climate in the United States or in Westchester County, we're trying to deal with it across the globe. And we've got places like Africa, we've got places like India, China. So there has to be some, no central decision making authority. And then the last point, which is again important for people at universities like Stanford, no real coordination between the policy formulation and thinking about the policy and what the R&D program will be. So if you go to the Department of Energy's Research and Development Program, so ably put together by Arun Madhrundar in the good old days, you will see that there's not, I think I'm right about this, a penny being spent on many of the technologies that were mentioned here on adaptation. You could argue a few little corners, net zero technologies, except for carbon capture sequestration, which has not been one of their great successes, very little going on, nothing in geoengineering. So the R&D program, which has to really in some way proceed sensible climate policy, my worry is it isn't as tuned to these major issues as I wish. Here are my major points, I'm not going to repeat them again. Thank you very much for your attention. I appreciate it very much. Thanks very much, Don. I think that was both thought provoking and provocative in the very best way for an academic institution, but I'm sure we have some questions. So let's start as usual with the students in the back. Ivana, do you have anybody excited? Thank you for that. I had a question regarding, so you talked a lot about climate control mechanisms and the cost, and recently Microsoft pledged to be carbon negative by 2030, and it seems like the private sector is taking strides towards also taking charge in some of these costs. What do you think needs to happen to encourage more private as well as public corporations to take a similar approach to change some of the progress of climate change? Well, I applaud Microsoft's decision or actions. There are several other large corporations which are understanding both the actuality of the dangers of climate change. They see the way the political wind is moving towards dealing with climate change in a serious way, but it hasn't happened yet, and so I applaud these moves. I'm going to say something. My wife, Pat, will be angry at this later on. At MIT, we have students come all the time saying we should be doing more. The companies that we deal with aren't doing enough and we should be pushing them more, and I keep on saying to them, it's not think global act local, it's think global act global, and so I'm concerned that when we deal this one grape at a time, we never get to the whole vine. So I'm delighted with what I hear about, I don't know anything about Microsoft. It's going to take more than big OECD companies taking the line to really make a difference. So I applaud it, but not enough for me. Yes. Yes, sir? Yes. I didn't quite follow that because you also said that the temperature increase will be exacerbated by economic growth. So I would have thought it would have been a negative feedback, where as the damages get worse, economic growth goes down. So what could you say about that? I didn't quite follow. So if what we do is we really let things get very hot, you're really going to have an effect on worker productivity. And that natural expectation that economic growth is always going to occur, it's going to be a point where higher temperature is just going to have a depressing, lessening. I don't say it'll completely reverse it, but the interaction between the effects on workforces through health, through work days present, through water availability, through the wars they may have to go and fight, will decrease the expected growth that you would expect under a lower temperature scenario. Is that good? But I thought I also said, I'm just beginning to think about this interaction between the effects of climate change and what their effects are on changing the way the economy is moving and how quickly it's moving and at what expense. It's something to think about in my mind. That's where I'm at it. Yes? Yes? Thanks for the presentation. So you mentioned the gap between climate policy and climate research and development. I was just wondering how you think we can better link policymakers with research on climate solutions and if there's been any good examples of collaboration recently. Well there certainly was great collaboration in 1976, John Wyatt mentioned, great collaboration then. Since then it's been sort of downhill. But I can't point to Congress is a very big competition here. And some Obama, the Obama administration was very, very vocal in the White House with, what is the name of that place? Economic Council for Economic Advisers. Council for Economic Advisers, OSTP and my student John Holdren was very active. You read some of that stuff now, it's pretty hollow. Pretty self-congratulatory. Yes, sir? And so do you take that into account in the effect on GDP of a higher energy cost? Well presumably it's a very good question but I'm not here to increase the hostility that I've expressed about economics. That's not my purpose. But I'm sure that in the integrated assessment models they do have some feedback on the effect of more regulations, for example, on what it drives does to two things which have some genuine political force. One is wages and the other is employment. And I would think that when people say, oh, I'm really going to get in trouble about clean energy, technology advance means more wages and more employment. I think that would be a relatively hard proposition. You want to say something about that? Say it again. If you raise the price of conventional energy, fossil energy, you cause substitutions toward green energy. It boosts employment in green energy and there's a reduction in employment in fossil energy. The net effects are a little bit hard to determine but there's some work at resources for the future that suggests that the net effects, though small, are positive. I know I'm sorry, I've got to say I'm ashamed. I was a director of the sources for the future for 30 years, I don't know, a long time. But there's also trade. Trade. Is it going to be good or bad for the American worker? Trade. No, I'm not. Go ahead, get another question. Hi. That there's a debate about what the discount rate of climate change should be. So as a climate change model expert, what are the discount rates that you have seen in the models and do you think they are justified? That's another very complicated question and it's been a lot of really intelligent work on the trade off between economic discount rates and utility discount rates for future generations, the present generations. It's not easily determined but some of the best work, I can say proudly, comes from economists at Stanford. From my point of view, it's a very difficult question. I don't like putting future generations at risk, although it may be logical. Yes, sir. Yes, ma'am. Before about thinking globally and acting globally. Could you speak to the mic? Okay. About what you said about thinking globally and acting globally. Yes. And that kind of made me think at this point in time, do we as individuals have no agency in having a partner? I don't believe that. I don't believe that. I have to, you have to find the agency, but I can tell you a group of thoughtful, well presented students from around the country arriving in Washington would scare them to death. Scare them to death. I believe there are many different things you can think about doing, which are more than just show. But it's better than thinking about should you feed up on your trustees. I mean, that's too easy. So in your formulation of your costs and or like discounted damages, did you take into consideration like the dynamics of like right now, economic, the fastest growing economies in the world are in like Africa and India. And they might, you know, the rate of carbon generation is only going to accelerate. And so I was just wondering if like different technology readiness levels plays into the cost that you posed in your convex truly optimization once I don't know. I think technology readiness levels is very important here in the big but also in what you say. So if I'm looking at the place which I'm most in touch with in internationally Bangladesh, what you recommend to them for technologies is not what you recommend for Palo Alto. And frankly, I think that we've, I'm going to say that again, the World Bank seems to think that every place is the same. And I do think it's very important that you look at the country, its economy, its politics, and you tailor your technology proposals to them in your systems to things which will work and make sense in their countries. I mean, India, for example, has a strategy whether it's going to work or not, I don't know, of doing everything by through a grid expansion. That's one way. So then you have to look at India different than, for example, other countries. But it's a very good question. Not the same deal. And they worry a lot about the U.S. or the OECD countries or the pushing stuff down their throats. Well, I think we're out of time. So let's thank John one last time for everything.