All Episodes

[Episode #49] – Climate Science Part 5 – Business As Usual


When we hear about the emissions scenarios used in the Intergovernmental Panel on Climate Change (IPCC) reports, do we really understand what they’re assuming about future fossil fuel combustion? And what do these emissions scenarios imply about the steps needed to achieve climate policy goals and decarbonize our energy system? For example, when you hear about the worst-case warming scenario known as RCP8.5, do you know that it is based on projections for a 10-fold increase in global coal consumption through the end of this century? Or that many of the estimates of future fossil fuel combustion in these scenarios are based on very old assumptions about how the energy system could develop in the future? And how can we square scenarios like these with our contemporary reality, in which coal is in decline and the world is turning to renewables because they have become the cheapest options for generating power? How should we actually think about the influence that the global energy system will have on the climate over the next century? In this fifth part of our mini-series on climate science, researcher (and Energy Transition Show producer) Justin Ritchie helps us understand what the IPCC scenarios really mean, and how they can be improved to offer better policy guidance.

Guest: Justin Ritchie is a PhD candidate at the University of British Columbia’s Institute for Resources, Environment and Sustainability as well as a producer for the Energy Transition Show. His academic work focuses on the economics of decarbonization, scenarios of transitions to future technologies and cognitive approaches to model-based science.

On Twitter: @jritch

On the Web:

Recording date: July 21, 2017

Air date: August 9, 2017

Geek rating: 5

Chris Nelder: So let's bring him into the conversation now. Welcome Justin to the Energy Transition Show.

Justin Ritchie: Thanks so much Chris. Great to be on.

Chris Nelder: It's really a pleasure to have my own producer on the show.

Justin Ritchie: Yeah, but you know I'm really only there from a logistics perspective. You are choosing the guests and directing the conversation on The Energy Transition Show, wherever you want to take it. So, I'm just there to help make it happen.

Chris Nelder: All right. So yeah I really want to just dive right in and start talking about your thesis. So in your recent paper in the journal Energy Economics, which is part of your Ph.D. thesis, you took a deep dive into the future of coal in the global energy system and how it has played a big role in framing our worst case scenarios for climate change since the first IPCC assessment in the 1990s. And by tracing back through the chain of accumulated research on which those scenarios are built, which I guess it shouldn't surprise me but not too many people have done that and when I did it I found it just incredibly frustrating and difficult to even figure out where the data was coming from, but by tracing back through that chain you found some pretty surprising assumptions embedded in them, assumptions that have been accepted for so long that nobody even questions them anymore. But in fact you found that the world has changed a great deal since some of that original research was done. And although the assumptions might have been reasonable at the time they are really questionable today. So why don't we start with that. What are the representative concentration pathways or RCPs in the IPCC emissions scenarios, and what are they based on?

Justin Ritchie: Yeah, well Chris you say it's somewhat surprising the assumptions and I think maybe they are when you or I talk about then or we look at them from an energy system perspective. But I guess to other researchers it's not so surprising. And we'll get into that in a minute. So it's really a big question and we can start out by talking about what the RCP scenarios are. So climate change is a global problem. And every researcher and discipline has to look at different aspects of this challenge, and to do that we use these standard scenarios. And so the main scenarios we use today are called the representative concentration pathways, or RCPs for short. And these are scenarios of future climate change used to look at physical climate, ecosystem changes, adaptation, vulnerabilities, and they benchmark the policy pathways for say the social costs of carbon, and really shape the economics of climate change as well like a very recent study that looked at county by county level data on the United States and climate change impacts to the year 2100, they used the RCPs. And you'll hear these sometimes referred to indirectly as kind of high or low emission scenarios, but in summary there are these four RCPs that are these outlooks or different levels of climate change this century. So they're named after levels of radiative forcing that we reach this century by the year 2100. Radiative forcing you've touched on it a little bit in some of the other conversations on climate science, but it's this change in the energy balance of the planet. And so we're at around 2-2.3 watts per meter squared of radiative forcing now and the RCPs are named after those levels of radiative forcing like 8.5 watts per meter squared or 6 or 4.5 or 2.6.

Chris Nelder: I always thought they were like version numbers or something. I never realized they actually refer to radiative forcing.

Justin Ritchie: Yeah and that gets into the surprise element of different assumptions because in the climate change research community you can say RCP 8.5 and you know what that means and it's very clear as a climate modeler or physical systems modeler, but then when you cover a study as a journalist or you talk about it to the general public and you say RCP 8.5, it's not exactly clear what that means.

Chris Nelder: Right. So what are these pathways based on? Like how are these RCPs constructed?

Justin Ritchie: Yeah, so we're at around 410 parts per million of CO2 right now, and all GHGs it's basically about 420 parts per million of CO2 equivalent, and RCP 8.5 is this high emission scenario that leads to 1200 parts per million CO2 equivalent of all GHGs. So that's methane and CO2 and everything else. And then RCP 6 is this lower scenario that leads to about 700 parts per million. And then RCP 4.5 is lower than that, which is about 600 parts per million. And then RCP 2.6 it peaks in the middle of the century around 460 parts per million and then declines to about 430. So that's a lot of numbers and I hope the audience didn't glaze their eyes over at all those numbers, but basically you can think of it this way of just comparing it to the level of CO2 we're at today, and so you know are we tripling or doubling or only increasing by roughly 50 percent from todays levels of CO2 to reach 4.5 or 6 or 8.5.

Chris Nelder: Yeah, but how did they arrive at these projections? Where do these numbers come from?

Justin Ritchie: Yeah, so in the debate that was going on about how to create these standard sets of scenarios for use in climate change research, which is really difficult to do because every single discipline has different practices for models and the kinds of models they want to use and they're using these scenarios in different ways. And so the research community about 10 - 15 years ago, they were looking into this question and saying well we'll take all the scenarios that exist in the literature right now and published on future outlooks for greenhouse gas emissions and pick these four different levels based on various specifications for framing the uncertainty range for future climate change. And so the RCP scenarios are not exactly fully detailed scenarios in themselves of how much oil or coal or gas we're going to burn in 50 or 60 years in the greenhouse gas emissions that arises to, they're just benchmarks that represent a wider range of scenarios that lead to those levels.

Chris Nelder: All right. Hang on I got to get this clear. So if I were going to try to project what I thought future carbon emissions would be, I would start with a model of fuel combustion. I would say we're going to burn this much coal, oil, natural gas, and it's going to produce this much CO2, and my scenario for fuel combustion would be based on some assumptions about economic growth and the mix of energy supply in different countries and that kind of thing, and you're saying that's not how they did this.

Justin Ritchie: Well in some ways it is because there were scenarios that approached the question that way, and they previously existed. But then to create these representative benchmark scenarios, it was just the full database of these scenarios that did it the way you talked about were looked at, and then the four levels were chosen from those. So what that does is it kind of decouples the modeling scenarios that different groups use from those specific outlooks for GDP growth or for energy use or whatever it might be. And the reason that happened is because it takes a long time to develop these what are called fully integrated scenarios of society and the energy system and climate change in order to expedite the process of scenarios to prepare for the IPCC's fifth assessment back then, it was just these representatives scenarios that were used. So the range between them is even more important than the actual scenarios because the conclusion is that we'll end up somewhere within that range. And so the highest scenario 8.5, it kind of sets the higher upper boundary for what's considered plausible climate change and then RCP 2.6 sets the plausible lower boundary, because it's very unlikely even if we had the biggest most ambitious scale up of CCS, of carbon capture sequestration, that we can imagine could we get below that.

Chris Nelder: But if the RCPs are decoupled as you say from an actual economic activity and fuel combustion scenario, doesn't it become difficult for especially a lay person to understand the probability of whether we're headed toward the lower or the upper bound?

Justin Ritchie: Absolutely. And I think that comes into the problem with communicating the RCPs and what the results mean, and then also how this what you would call the downstream research community that actually uses the RCPs interprets what they mean as well. So part of the challenge is that even though it's very clear in the original paper describing the RCP process, how they were developed and how the scenarios were built from these benchmarks, that really RCP 4.5 or 6 or 8.5 can represent a business as usual world. In other words a world that doesn't have climate policy. But then in the papers that describe the individual scenarios it shows the RCP 6 and the RCP 4.5, those lower scenarios resulting from mitigation steps that start from baseline scenarios that look like RCP 8.5. And so that's part of where the confusion comes in is that the only scenario described as business as usual in those original papers is RCP 8.5, so then when it's used in a wider range of studies different journalists or the general public or even researchers themselves describe RCP 8.5 as the business as usual pathway, when technically it is a business as usual pathway, but business as usual pathways could fall anywhere within that range.

Chris Nelder: But 8.5 is you're saying the worst case business as usual pathway.

Justin Ritchie: Right. And it always was designed that way. And so if you look at the underlying assumptions it really isn't a business as usual pathway, it's the worst case business as usual pathway. It's basically what would happen in society in our demand for agriculture, in the energy system that would lead to the highest possible levels of greenhouse gas emissions this century which reaches that 1200 parts per million of CO2 equivalent gases in the atmosphere.

Chris Nelder: Ok, so why is it that that seems to be the only scenario or the most often cited scenario that journalists talk about? Is it just a desire for sensationalism or what? I mean do we have any reason to think that that's more probable?

Justin Ritchie: It's partly that, it's partly that it does show the biggest effect. But when you look at studies that a lot of researchers do they use the other RCP scenarios, and so kind of by accident I think a lot of times the RCP 8.5 scenario results from a study are covered in the highest profile way, whether it's sea level rise or whether it's economic impacts in different areas. When if you look at the underlying papers the researchers have used RCP 4.5 most of the time and a lot of the time they use the RCP 6, then again the importance comes back to interpreting that range between the results. So it's unlikely for a lot of reasons that will have the impacts that are associated with RCP 8.5 and it may be more likely that we'll get the impacts that result from our 4.5 as the lower bound of impacts.

Chris Nelder: All right, well I think it would help me to understand the real probability here if I just kind of knew what was in it. So how is RCP 8.5 constructed? Like what is it assuming about the future of fossil fuel use?

Justin Ritchie: Yeah, so these are RCPs were selected from this benchmark of all these scenarios, and RCP 8.5 represents this energy system that relies heavily on coal. And if you look at the original paper on RCP 8.5 in the text it says it's an outlook for a 10 times expansion in annual coal production in the 21st century. And a lot of this coal is used for transportation and coal to liquid synfuels and it's also a lot of oil and gas. And the original paper describes it as essentially all theoretically extractable fossil fuels. It's about three times as many primary energy resources as we use today and it's an outlook on the coal side for twice as much coal use in the 21st century than all fossil fuels used previously up to this point. So you can kind of think of it as the ultimate all of the above energy strategy but with this emphasis on coal. But I think the really interesting part of that is not that the scenario uses a lot of coal, it's that a lot of the coal acceleration starts happening around 2040 to 2050 after a peak in oil production. So oil production in the original RCP 8.5 scenario it peaks around 150 million barrels a day in 2040 - 2050, and then that's when the coal acceleration starts happening.

Chris Nelder: When in reality our oil consumption globally is not yet at a 100 million barrels a day. I think we're at more or less like 94 or something like that right?

Justin Ritchie: Yeah. 93- 94 with about 12 million barrels a day from natural gas liquids, 7 from unconventional and about 70 from conventional oil.

Chris Nelder: And global oil demand growth really significantly slowed down right around the time of the financial crash and the point at which we had the first big oil price spike in 2008. So this notion that we're suddenly going to go back to a steep trajectory of oil demand growth and get to 150 million barrels a day by 2040 seems really unrealistic to me let alone the fact that it's suddenly going to then crash and be replaced by coal when coal use itself is declining in most of the developed world. And there are some very serious questions being raised about the long standing assumptions that the developing world is going to use a lot of coal as well just because renewables have become so much cheaper than coal especially in places like China and India. So why is no one questioning this scenario?

Justin Ritchie: Yeah, well the reason I say it's interesting that the coal acceleration kicks in after peak oil at this really high level around mid-century gets back to the question raised in this Energy Economics article on the 1,000 gigaton call question, and that concept is one of a coal backstop energy supply and it really dates itself back to the 1970s when there was tremendous anxiety during the 1973- 1979 oil crises, there were a lot of energy studies that were taking place at the time, and a lot of the conventions that we still use using energy modeling to this day started with the energy modeling that grew out of collaboration of engineers and economists during the modeling work that tried to address hey are we going to run out of oil and if so what do we do. And the question always looked at well how are we going to support this level of demand, and there are these massive numbers of coal resources in the United States and the Soviet Union, in China, and the answer was well if we run out of oil we'll do coal based synfuels. And that's the idea of the coal backstop that is shown explicitly in the RCP 8.5 scenario, but it's actually embedded in a lot of the scenarios that are produced by the models that create these integrated outlooks for economies and society and greenhouse gas emissions and the energy system of having this vast limitless 3000 year supply of coal that we rely on to support high levels of demand for energy regardless of the scenario. In fact even part of the problem is not that our RCP 8.5 uses a lot of coal but even the mitigation scenarios published in the IPCC assessment and Working Group III used a lot of coal. If you look at the original RCP 2.6 the ultimate mitigation pathway to keep warming to only 1.5 degrees this century, it still uses all coal reserves. And that's because of assumptions around carbon capture sequestration and its availability, and the general concept that coal is this cheap limitless supply. And so if we can get CCS to be economic then we can rely on this coal resource base still.

Chris Nelder: I just got to say this all sounds totally absurd to me. I mean look the world has changed so radically since the 1970s. We're now talking about massively decreased oil demand going forward with the advent of electric cars. Nobody is interested in building a new coal plant anymore. We're definitely not going to burn all that coal just because it's there. We now have renewables that are cheaper than anything else in most of the major markets of the world. What are we thinking to be so interested in these scenarios that seem to be based on an ancient history that no longer applies, especially if you're talking about an economic scenario that should be driving our fuel combustion scenario?

Justin Ritchie: Yeah, well the narrative of society and technology underlying 8.5 is that demand for energy is so high we have to tap into this vast resource base of geologic coal because technological progress and all other energy technologies essentially fails. So solar, nuclear, gas, conventional oil, unconventional oil, everything else is much more limited than coal. And so this only available economic option is to transition to coal. And what you're saying about how you see that as implausible, that gets in the way that the community of researchers that create the emission scenarios think about energy resources over the long term. And so if you were to say well you know look at what's going on with solar like what's going on with potentially EV deployment, oil's at $45-50 a barrel and we're still talking about bullish EV forecasts, you know that's unheard of 20 years ago.

Chris Nelder: And look at the total collapse of CCS technologies that we've tried on our various coal plants. I mean that's clearly not working.

Justin Ritchie: Right, but what you would have to face is an argument that well that is one hypothesis of what's going to happen in energy system and it's shaped by what's going on this decade, and when you look out over 80 years to the year 2100 you really have to consider all possible eventualities. And if that coal exists in the ground in the form of identified resources as geologic occurrences then are there ways technologically that we could possibly figure out how to access it in an economic way, and if that answer is maybe yes then under the socio economics of extremely high demand, extremely high resource intensity and low decarbonization in terms of high energy intensity economy then you end up seeing coal as the best option. So it's this kind of way of contrasting the two hypotheses for technology and society, and how they're going to play out in the long run.

Chris Nelder: All right I mean I can understand that from sort of an absurdly conservative risk averse point of view in modeling, but where is a scenario in all of this that would in any way map to what I would consider a probable future? Right, like in which there's no CCS, in which oil demand starts to flatten out and decline starting now, in which coal demand flattens out and declines starting now, in which we do not have a massive increase in natural gas, and we do have a huge explosion in renewables because that has now become the cheapest thing worldwide or you know will be in all markets within a couple years. Like where is the RCP that maps to that future?

Justin Ritchie: Yeah. Well Chris you're in Boulder, and nearby is the National Center for Atmospheric Research, NCAR, and there's several people there who are actually working on that kind of framing for future scenarios. And so like I was saying earlier these RCPs were just the starting point of creating these kind of abstracted emission pathways, and the team at NCAR is part of this international collaboration to create what are called shared social economic pathways or SSPs. And so those SSPs are the fully detailed scenarios of an energy system that map to the different levels of radiative forcing at the end of the century. And so what you're talking about is a concept that's called SSP 1 which is basically a green growth world is what it's called, and it still leads to climate change that's above 2 degrees. And so by no means does taking the coal out or doubting RCP 8.5 mean that we're going to just do nothing and avoid two degrees. There's a lot of different studies that are taking more hard line geologist approach to this question, so we can link to a few of those in the show notes. But Steve Moore in Australia, he published in 2015 with his co-authors that looked at projections of fossil fuels by country, and really considered hey what's the ultimate bullish scenario for unconventional oil for coal, and he basically calculates an outlook for cumulative emissions in the ultimate bullish case that only reaches RCP 6. And he says his best guess scenario ends up around RCP 4.5. So something more consistent with the idea that you're putting forward. And there have been a few other studies that take although ultimately recoverable resource estimates and kind of do it's called a Monte Carlo analysis where you randomly sample all the different numbers and then run that through the models that create these emission pathways, and what they found is that somewhere between RCP 4.5 and RCP 6 is where they end up with the low probability of ending up below RCP 4.5, but we're still in that range where we go above 2 degrees. But the interesting thing that that study found on the likelihood of climate change pathways given uncertainty in fossil fuel resources is that the dominant factor that explains how many emissions we're going to have in kind of a baseline world is coal. More than 70 percent of the uncertainty in emissions could be explained by the uncertainty in the coal estimates. And so that's why my co-author and I thought it was really important to dig in deep and try to figure out what these coal reserve reports and resource supports really meant in the context of reserve resource definitions compared to oil and gas definitions.

Chris Nelder: All right well let's dig into that uncertainty question a little bit because you know even though climate modelers may understand how much uncertainty is sort of embedded in these assumptions and in these emission scenarios, I certainly have not seen the uncertainty question addressed hardly at all in journalistic coverage about climate change, and in a paper that you sent me written by the late Stephen Schneider of Stanford University, there is a figure showing how uncertainty multiplies from the underlying emission scenarios on up through the impact assessments, like at each stage of the chain between an emissions scenario and its impact assessment the uncertainty grows a little more each time. And given this uncertainty explosion it was called, how much confidence can we actually place on the impact assessments on the climate projections that we based on these emissions pathways?

Justin Ritchie: Yeah, so the challenge is we don't know how much coal or oil or gas are going to burn in 2070 or 2080, or even what the technology is necessarily going to look like. And then given a certain level of emissions we don't know exactly how the carbon cycle is going to respond. Climate change is this dynamic interplay between positive and negative feedbacks that lead to more and less warming given different Earth system responses. So we can look at the past using paleoclimate data and we can develop these different hypotheses about how it's all going to respond in the future, but each level that we multiply it through because we're using these RCP scenarios and other emission scenarios is the inputs that go into the more complex system models that really understand the physical and ecological responses to warming, that level of uncertainty grows. And so that's why it's really important to consider whether these high scenarios should be ruled out or whether there's still a plausible worst case scenario. Because if you can constrain that initial series of emission uncertainties to some degree then it helps to focus in on the things that are more likely to happen. And then you can rule out certain prospects. And even if RCP 8.5 shouldn't be used anymore, even if the most bullish possible outlook for unconventional oil and oil shale and whatever you can imagine in terms of methane hydrate harvesting does not lead the RCP 8.5, then we still have something like RCP 6 or RCP 7 as the worst case scenario, which is still a really terrible world to live in. No one wants to get there. And so there might be uncertainties at every stage of the process but it's hard to express that in a nuanced way because it comes into the challenge of developing any forecast or any projection, which is you sit down and talk with a journalist who has a short timeline who wants to get to the core of the story, and if you say well you know the impacts might be between bad and really horrible, you're going to slowly get distilled down into trying to produce kind of a best guessed answer. And that dynamic plays out in so many ways in communicating the uncertainty. And so Stephen Schneider was one of those researchers who focused as very experienced climate change researcher on how to communicate that uncertainty in a constructive way. Because when you reduce the uncertainty you're really not improving the broader understanding of climate change science in the general public, and that leads to a lot of other downstream effects that are very difficult to get across in regards to the basics of climate change science.

Chris Nelder: Well, I mean based on what you've told me so far, I'm really looking for a better way to understand the probabilities because I'm just not getting that at all from the climate change coverage out there. In fact there was this recent article by David Wallace-Wells in New York Magazine titled "The Uninhabitable Earth", which you know scare the shit out of everybody who read it I think, and that story you said that the effects of climate change could be much worse than people think, including famine and economic collapse and unbearable heat or even the extinction of humanity and other species. And that set off a whole interesting discussion and debate amongst various people in the climate change research community and their journalists about how we should be talking about this or how we should be communicating these scenarios, and some people have said well it was really irresponsible of this guy to put out such a scary scenario when it's not really that probable, and then we've had other people saying yeah but we've really under-represented this worst case scenario in our coverage and people do need to understand that this is a possibility, but then the conversation devolved into this whole should you be sensationalist or not, whereas what I wanted to hear was how probable is it?

Justin Ritchie: Yeah, and that's a really good point. And I think it comes back to actually an argument that Steve Schneider was involved in when the previous set of emission scenarios were published back in the year 2000 that were called the "Special Report on Emission Scenarios" and they were structured in a very similar way and a lot of those scenarios actually were the basis for the RCP scenarios. And there is this argument on hey should we assign probabilities to these. And there's this real contention on representing them as equally likely storylines that lead to different levels of climate change versus trying to assess probabilities. The argument on the storyline side is that really when you think about the long term future so many different things can happen and you can't really test the probability distributions. And so if you think that you're going to try and do this what's called a frequentist definition of statistics and apply that to the future, then it's just going to be proven wrong. So why put probabilities on them. People like Steve Schneider argue the counterpoint of that people are going to interpret them subjectively. And so we have a better means of guiding those subjective probability distributions than the people who might interpret the emission scenarios, so we really do need to make it very explicit, very clear what our assumptions are, and then try to give it a probabilistic weighting in terms of how likely certain levels of warming may or may not be. And that's been a debate that's gone on for decades now.

Chris Nelder: Well, ok fair enough, but I mean that's my personal predilection is to give the audience credit for being able to understand a little bit of nuance, to say there is X percent probability of this and there's a Y percent probability of that, and I'm just not getting that at all. Even from the fallout from this David Wallace-Wells article.

Justin Ritchie: Yeah, and I find pieces like David Wallace-Wells' story on "The Uninhabitable Earth" so fascinating because on one hand I completely agree that we have to address this worst case scenario for a 21st century climate change head on in a realistic way. We really need to know what are the worst impacts because when you do a risk assessment you need to know what these impacts are. But getting back to your point Chris, risk is really this probability multiplied by impacts. And so you can't do a real risk assessment unless you have some way to wait. The probability of these impacts happening and part of the debate in climate change is how bad are those tail distribution impacts, how likely are those? And really when it comes to the sensitivity of the climate to different emissions you hear these ranges of 1.5 to 4.5 degrees of warming for hitting double CO2 emissions around 600 parts per million, well that's really only part of the uncertainty range. The real uncertainty range for the Earth system response is more like 1 degree to 11 degrees. And it's very hard to know exactly where we're going to lie in that distribution of probabilities given complexities in modeling and complexities in interpreting the historical record and how the Earth is going to respond to CO2. And so if that tail really does go all the way up to 11 degrees for only 600 parts per million then we need to figure out how likely that could be and put research effort into it.

Chris Nelder: We really don't want to turn it up to 11.

Justin Ritchie: Haha, yeah exactly. And that can be really scary to do. But when I read David Wells' article, he took the worst case outcome of RCP 8.5 independent of its underlying assumptions and he went around to different scientists and talked to them about what that could lead to. And in some ways he's really not to blame because so many other researchers have been doing that for so long accidentally because of what we just talking about regarding this nuanced concept of well the RCPs are scenarios but they're not, and they're abstracted from developments in society in some way or at least they have been until now, so I think that now that we have these shared socioeconomic pathways and there's so much reaction to pieces like "The Uninhabitable Earth" when climate scientists send out their research to peer review hopefully the peer reviewers are more stringent and challenge the way the RCP 8.5 is framed or used, and that researchers explain it in a better way, and when journalists sit down to cover RCP 8.5 or research that uses it they place a little bit more scrutiny on it. But there's not really a lot of that now.

Chris Nelder: All right, well I'm just going to put a stake in the ground and say people I think we need to be a little more transparent about the probabilities here and we need to start communicating what the real range of risk assessment is, right. It's not enough to just say oh everyone has no attention span so we have to scare the shit out of them or they're going to think everything's going to be fine, or in some other way we need to hack people's attention span. I don't think that's what this should be about. I think we ought to be taking a very realistic, sober, nuanced approach to understanding what are the real probabilities because otherwise we're just going to misinform policymakers, and you know I can already hear it in the back of my mind, Dave Roberts screaming but that's not how this works!

Justin Ritchie: Yeah, and I think the challenge is when you're communicating with the policy community and there's so many concerns that they have in their mind, it's very hard to rise above the noise for them and get them to care about your issue, and that is always a challenge. And so, you know I think about that a lot when I look at RCP 8.5 and I try to figure out should we still be using this anymore, but the bigger issue is not how much coal RCP 8.5 uses or whether we should use it, it's how coal influences all these outlooks for mitigation technologies and the solutions that we're thinking about for the future. So if you go to these policy models, these integrative assessment models that have this coal backstopped baked in like I was just saying a minute ago even the deep mitigation scenarios use a lot of coal with CCS. But I think the kind of debate around deployment versus R&D for renewables gets heavily distorted because of the underlying socioeconomics and technology outlooks that these policy models use.

Chris Nelder: I think that particular one was distorted from the get go because it was really promoted by the nuclear crowd. And they were again trying to sort of promote a nuclear future but under a different guise.

Justin Ritchie: Yeah, well if you think about the idea of the backstop technology, it's this technology that is cheap and abundant, and that's why we switched to it when we run out of more expensive resources. And so the coal or unconventional oil has always been this transportation backstop in these models. But if you think of the backstop as potentially a low carbon backstop, like solar with EVs or wind or renewables in some way, the way that these economics of the scenarios are constructed is well we're running out of the resource, we've got to switch to another resource, and if we want policy to make up the gap between the cheap high carbon resource and the expensive low carbon resource then we need to have all of these expensive policies to get us to this lower carbon more expensive one. But if the gap is actually a lot smaller than we thought, if actually the low carbon resource is actually already cheaper than the extreme coal, then suddenly the amount of cost that going to a low carbon future could have is much lower than we previously thought. In fact, I think, and I'll show this when my Ph.D. thesis is published and hopefully some of the papers that are under peer review will make it out after a while, but I can show that getting to 2 degrees or 1.5 degrees is actually a lot easier than we thought when you take these coal backstops out.

Chris Nelder: Yeah, if you don't assume, again absurdly, that the future of transportation is going to be powered by coal turned into a liquid fuel.

Justin Ritchie: Yeah, or some of the models look at hydrogen and large scale hydrogen scale up, but then the amount of electricity that's required for the hydrogen comes from electrolysis from coal fired electricity, right.

Chris Nelder: And that's all completely absurd also because no one's going to build a big fleet of hydrogen vehicles.

Justin Ritchie: Yeah. And then I think the bigger question comes back to are we really in a business as usual world or baseline world?

Chris Nelder: And what is usual?

Justin Ritchie: Yeah and what is business as usual? And so I think that's another really big confused point when this conversation comes up because the business as usual that was used by the IPCC in 1990 led to what you would call RCP 10 now. It was even higher than RCP 8.5, and it used even more coal than RCP 8.5, and quite frankly I was surprised when I first saw RCP 8.5 at how much coal used, but then I realized it was very conservative compared to a lot of the other mega coal pathways that are used in the research community and have been used in the past. And that original business as usual scenario led to even more warming. And just in the 30 years since then we've downgraded the kind of upper bound that we're using by one and a half watts per meter squared, and it's reasonable to expect that if we draw this line at 2100 and we're producing these scenarios that lead to that point, the closer we get to that point in time some of the ideas and concepts that we had about how the world might look then are going to be invalidated.

Chris Nelder: Well I clearly need to get somebody from NCAR on this show here to explain themselves, and also to kind of dive into some of the other, what did you call it? The SSRs?

Justin Ritchie: The SSPs.

Chris Nelder: SSPs, yeah. We got to get into that one. But you know now you've raised an interesting question that I wanted to explore a little bit, which is how old are these underlying models of future energy use, and how old is the data? I mean something that's kind of become a recurring theme on this show is that a lot of our assumptions and a lot of our policy recommendations about future energy supply or about energy transition turns out are based on very old musty presumptions about the future that just really no longer apply. And that was something we really discovered in our conversation with Rembrandt Koppelaar recently where he explained that EROI of solar for example was far better now than it was 10 years ago but everybody's still looking at that old data and not even realizing that they're no longer looking at a picture of the present let alone the future. So what's the oldest data, let's put it that way, that's involved in these RCPs?

Justin Ritchie: Well when it comes to coal data the uncertainties are very large. And do you think it makes sense to explain a little bit about the uncertainties in the data, because that's really the core of this energy economics article.

Chris Nelder: You're the expert, man.

Justin Ritchie: All right, well we'll dive into that real quick. And basically this concept is we've got to look at the full potential. What's the full potential of total coal resources that exist, and then, could we ever imagine ways to access that. And so this 1,000 gigaton carbon paper looks into these uncertainties in the data, and generally these emissions scenarios are based on the idea that today's reserves, and reserves generally indicate the amount of any mineral resource that we think we can extract within reasonable assumptions of price and technology, and then resources are this larger base that are the total amount in the earth's crust that maybe we could get to in the future. So if we have prices that are high enough or we have technology that makes it accessible then those resources become reserves. So when you say something like an economically recoverable resource that implies that with changes in market price or technology then the quantity will increase. Although I think one often missed point is that it also implies that with developments in other technologies that are competitive, then the quantity actually decreases. And so in oil and gas there is these different categories of reserves and resources and they're assigned different probabilities of recovery. So international oil companies are very strictly regulated on what they define as approved reserve. And so you have these proved possible and probable reserves which have different probabilities of recovery. But when it comes to coal those kinds of categories don't really exist. And so it looks like there's this virtually unlimited supply of coal that can support high levels of demand, something like 3000 years of supply at todays production levels. And you'll hear politicians say stuff like this, and this was really popular in the 1970s to say oh we have thousands of years of coal so we could always just return to coal as this energy source if we wanted to. But what we've seen is that China and its scale up of coal has really tested this concept, because they've reoriented and changed coal markets around the world. And so if we think about what China has done in almost tripling its coal production in a very short time, that's only the start of getting to an 8.5 world because it's got to continue at that rate for the next 80 years to reach the level of coal that's in RCP 8.5. And so by China scaling up coal it's required us to test the underlying ideas behind all the coal use. So what I was just talking about with oil and gas as reserves are used up or demand increases, this reserves-to-production ratio, the amount we think we can get out versus amount we can produce, it changes. And generally since the 70s, this reserve-to-production ratio has maintained what's called an equilibrium level, and that means that information regarding oil and gas is very expensive to invest in. It's very intensive effort to go out and explore and find it, and then develop it into reserves. And so oil and gas companies they don't do that until there's actually an outlook that they might need it. And so they go out and they explore for more. And so that's why a lot of energy resource economists view reserves for oil and gas is this dynamic working inventory where resources are replenishing reserves and the reserves are being produced. That very same concept has been applied to coal. And that's what creates these extreme coal outlooks. And so the idea was the reserves are the lower bound on production for coal because when we use them up of course we'll convert more resources to reserves. But it turns out that coal, since is geologically very different from oil and gas, that concept doesn't apply. And so we've had this huge increase in price and this tremendous increase in demand, and what happened to coal reserves actually declined even faster. And the reason my co-author and I explain this is because it's the very same reason why oil and gas maintains this equilibrium reserve-to-production ratio. It's because you have this incentive now to improve your information. And in coal the incentive to improve the information actually led to a decline in reserves because in the coal assessment process it's much easier to identify where coal is at and to explore for it, and then if you look at the assessment methodology that coal geologists use to find where coal is, it talks about this constant subtraction process where they identify coal in a region and then over time subtract from it to figure out the actual recoverable amount. And the difficulty is that coal trading contracts are signed in much less frequent terms than international oil markets, and so the information gets out dated. So some of it actually goes back to 1913, a big U.S. Geological Survey data source is a study that was done back in the 1970s by geologist Paul Averett, and then after those initial availability studies there are more specific studies of regions that when they go back in at a recent U.S. geological survey study that was done in the 90s and then another one in the early 2000s, it found that only 10 percent of reserves and a lot of regions could actually be recoverable, because over time there is actually a stream there and you didn't know it, because you mapped out the whole area of the coal seam or maybe a city built up or a road was there, and because they're updated so infrequently a lot of things happen that mean you can't get to the coal you originally thought you could get to. And so reserves have declined by two thirds since the late 1980s when this original IPCC business as usual was created.

Chris Nelder: Wow. And there I thought you were going to just lose the thread of this question entirely. But, ok so I think the answer is 1913.

Justin Ritchie: Some of it, yeah. So the policy models that do this energy modeling, they look at this data and it leads them to have systematic overconfidence in the amount of coal that exists.

Chris Nelder: Right. OK, I got that. Now I suppose, you know, that we can forgive a scientist in 1913 or the 1920s for having sort of incorrect assumptions about how much coal we're going to produce, let's say by 2000. They would have no way of knowing where a city was going to be built up or where the roads or houses would be or what have you. But you know that was just shortly before the dawn of the oil age. That was long before the advent of serious natural gas production, long before the invention of nuclear power, many decades before wind or solar existed, let alone became commercially viable. So if we thought it was hard to predict what might happen over those 80 years, let's say from 1920 to 2000, can we be any more confident about what the next 80 years of the energy system evolution will bring? And how do we accommodate all that uncertainty in our projections for climate change?

Justin Ritchie: Yeah well that very argument is the reason why the mega coal pathways exist, because we have to look at this full uncertainty span as it's argued. But I think that the more I look into the socioeconomics and the way that these energy models calculate supply and demand, I think it's created this kind of classic anchoring bias in terms of how we view emissions scenarios and how they're used in broader research because all of these energy models are carrying on in a tradition that started in the 1960s and 70s with the Limits to Growth World3 model and Jay Forrester at MIT, and so the Limits to Growth study, it gained a lot of attention, and then a number of engineers and economists started working together to try and improve on this tradition of global modeling. And at the same time as the energy crisis was happening in the 70s that became an explicit focus on how to adapt these to look at energy, and then into the 80s and 90s as climate change became more of a research focus these very same kind of core concepts were adapted to look at climate change and created a series of climate change integrated assessment models, or IAMs, and basically those IAMs are the tools that we use to look at the uncertainties in the future energy system. But I think a lot of the different concepts of substitutability, of elasticity, of could we have the solar option, could we have the unconventional oil or coal backs up option, it goes back to the way that we thought about it back then. And so it creates these kinds of what you would call needless constraints on the way that these questions are asked and posed. So that means these massive biomass energy carbon capture storage outlooks for mitigation have all of these assumptions and model structures built into them that make certain future technology pathways look more likely than others. And that's a real challenge to try and get beyond that because there's another concept for the outlook for oil resources that has to do with the economic idea of learning by doing, where basically the more we pull out of the ground the more productive we get at pulling more out of the ground in the future. And there's a paper that my co-author and I we have out in the Resources for the Future discussion paper series which looks at this critically, and the biggest issue is that it doesn't factor in the productivity gains and declines that come from geology and from geography for specific producing regions. It just frames it entirely as well technology is getting better. And that really distorts the economic outlook for what's going to happen in the future energy system.

Chris Nelder: So what does it all mean? Like, how do these scenarios influence our understanding of policy and technology that we might use to combat climate change, or maybe put another way given all the uncertainty, what kinds of policy recommendations can we confidently make today?

Justin Ritchie: So, even if all of these RCP scenarios show baseline use of coal that's too extreme, even if we rule out RCP 8.5, there's still no good argument against decarbonizing in the energy system, against reducing fossil energy use, against climate policy. But I think it can make a lot of our climate policy goals look more achievable. And that's the really important thing to focus on as we move forward in thinking about decarbonization, is how do we feasibly decarbonize the energy system? And if our outlooks are hypotheses for what's going to happen in energy resources or demand in society after 2050, may need to change dramatically based on some of the concepts we've talked about today, that still doesn't negate the need for climate policy, and I think it's going to make the way we think about decolonization between now and 2050 even easier.

Chris Nelder: Well I hope so but I mean, if all these assumptions are flawed or at least based on highly questionable assumptions about future fossil fuel consumption, and as you pointed out this most often cited scenario of RCP 8.5 is, and i'll just say in my personal opinion totally absurd, are you concerned at all that this research that you've done will be used by climate change deniers to show that energy transition and achieving climate change isn't nearly as urgent as it's been cracked up to be?

Justin Ritchie: Yeah, it's definitely a concern, and I think unfortunately climate change has become this deeply politicized science. And so everything that's said about it has to fall into this camp that either supports the kind of basic believer or denier storyline, and I don't want to create a straw man argument and talking about this, but the simplest framing is basically there's a believer side generally aligned against fossil fuels, and there's a denier side that's pro fossil fuels, and everything that we've been talking about today doesn't neatly fit in either of those camps. But it's that area outside the kind of scientific meta narrative that the most basic important questions about what climatic process are and global change research really needs to address. And those questions just do not fit into that will climate change be worse or will it be as bad as expected. That's just part of the aspect of doing science in a politicized realm.

Chris Nelder: So I guess we should probably wrap this up but I want to kind of give listeners a preview of the rest of your work. So the first part of your thesis was published in May this year in the journal Energy Economics and that got a very nice bit of coverage in Bloomberg by their climate journalist Eric Roston, and we'll link to both of those pieces in the show notes, but what's next? Can you give us a preview of what the rest of your thesis will say when it's finally published.

Justin Ritchie: Yeah, so I really never set out to do a Ph.D. work on coal. I set out to do it on decarbonization, but I kept running into in the economics of decarbonization all of these different ways that the economics of carbon are assessed and produced. And so a lot of the thesis chapters are about these concepts that energy policy models and the climate policy models use to think about the future of carbon resources. And I think that it has a potential to reframe the way we approach these questions of decarbonization because not only our emissions extremely likely to be much lower than RCP 8.5, but the kinds of technology pathways that we produce to illustrate ways to achieve 2 degrees can become more feasible as we've touched on this conversation. I think it's completely possible that you can believe that humans are the dominant factor in 20th century warming, that we're causing the climate to change and that we really need to do something about it, but that it also RCP 8.5 is not achievable. And a lot of the people who I talk to who do a lot of great work on climate policy, there is an increasing pushback against these RCP 8.5 scenarios because we're so focused on creating realistic scenarios at the 2 degree level, but on the upper end it just goes completely into levels of unreality and that kind of research dialogue is just going to keep ramping up the longer we resist basically invalidating these outdated concepts.

Chris Nelder: Well OK. And your personal ideal world let's say, if it were up to you personally to figure out how to give the world guidance on what to do about climate change, how would you approach it? Would you use this method that's been put out there to just start with these RCPs that are sort of notionally connected to a scenario fossil fuel combustion but not entirely? Or would you start with here's my goal, 2 degrees, how do I engineer back from that to figure out what my fuel consumption needs to be? Or would you take some other approach?

Justin Ritchie: I really think that we need to put a lot more effort into addressing these questions like you raised regarding probabilities of future outcomes and addressing unexpected surprises, because the biggest issue with a lot of these scenarios is that they're kind of set and then they run and then they create this paradox of a world that is very stable and runs out to the end of the century, but then one intervention, one action can change the course. And that kind of problem has played out since the 70s, in the 1970s energy studies, and when you look at the kind of futures they expected for now they highly diverged from what we were looking at. Some of the same models that were used to produce the RCP scenarios were those models used back then. And we can see how poorly they performed. And so the issue has to do with an intelligent use of these scenarios to imagine creative surprises that happen not only in a world where we're not doing anything about decarbonization but really anticipating barriers and issues to deep decarbonization and trying to say hey what's going to happen to society and what push backs are we going to face organizationally as we aim to produce a much lower fossil fuel world. And I think those two aspects of imagining these creative surprises and doing an assessment of that and looking at these probabilistic pathways really can go a long way toward guiding the next few decades of the energy transition.

Chris Nelder: Interesting. Well I'm with you on that. You know let's try to figure out what's realistic and what's possible and go with that. Well thanks Justin. This has really been an illuminating conversation. I mean even though you and I have talked about this stuff for quite a while now I still seem to learn something new every time I talk to you about it, so it's great to also to finally have you on the show.

Justin Ritchie: Yeah. Thanks so much for having me on Chris.