 When you take into account these constraints, you reach the conclusion that green growth has many many challenges or many barriers or many problems that typical models don't see. They say they cover all possible futures. That's not true. There are other narratives, but that makes sense. For me, that's crazy. I mean, this, just this, you know, I really have difficulties to to swallow this pill. There is no physical representation of stuff. I mean, when you model, you need to do choices about simplification of reality, and you do your simplifications based on your objectives. You cannot have everything, so you do some choices. There is not a scientific, absolute way to say this is better, no, because it will depend on social decisions, priorities. Hello, everyone, and welcome to the Circular Metabolism Podcast, the bi-weekly meeting where we have in-depth discussions with researchers, policymakers, and practitioners to better understand the metabolism of our societies, or in other words, their resource use and pollution admissions, and how to reduce them in a systemic, socially just, and context-specific way. I'm your host, Aristide, and today we're going to discuss about the urgent need to develop robust models to guide our policies and actions towards a more safe and just future for humans and the living world. While there is a consensus that we have crossed a multitude of planetary boundaries, and we must reduce our emissions to net zero by 2050, the current models and their associated policies can neglect local and global social and biophysical externalities. This leads today to a rather techno-positive policies which underestimate the impacts and overestimate their efficiencies. So how can we build robust models that capture the complex interaction between socio-economics sphere and the biophysical sphere to test the impacts of bold policies such as universal basic income, 100% renewable energy, and even confront degrowth and green-growth worldviews? To help us better understand the importance of these models and how to use them, I have the pleasure to welcome Inigo Capella and Perez, a researcher from the group of Energy, Economy, and System Dynamics at the University of Valladolid. Inigo's research focuses on the modeling of Energy, Economy, and Environment systems. He is the scientific coordinator of the Horizon 2020 project called Local Motion. In fact, this episode is done through a collaboration with the EEB, the European Environmental Bureau, which is also a partner of the project. More on the project in the description of the episode down below. So with all that being said, welcome Inigo to the podcast. Thank you very much. It's great to discuss about models. I think they're kind of the cornerstone today of policymaking, because we live in such a complex world. But perhaps before we start with all of these complex stuff, I want to understand how you got interested into this. Yes, so as a kid I was always very interested by nature, not only interested but really loving it and spending a lot of time. My parents were bringing us to make walks in the forest, in the mountain, so every time. Also when I grow up, then I start to be involved more in activist movement in my local town, in an organization which is called Ecologistas en Acción, which by the way is member of EEB. I also were doing in the cities different dissemination activities and so on. And one time, after one meeting, we went to drink a coffee, I mean, to drink some beer. And there was one member of the group, which at the moment I didn't know her too much, but she was a researcher at the university. She's Margarita Media Villa, and she proposed, she asked me, Inigo, are you doing something now? I was not doing anything because I finished my studies some months ago, because there is some possibility for an internship with us. We are working in energy sustainability, so it will fit with my ideas for my interests and started. It was very bad pay, internship, I was living with parents and so on. And I really loved it because I had the opportunity to enter this group of energy economy assistant dynamics, which was just created. I was the first member non-professor, which they managed to hire somehow. So it was like three, four, five professors with a lot of knowledge and me. So it was like I was being supervised by four, five big brains. So in this sense, I was privileged person. So this is my story. I mean, I started with them, then I continued, then, well, there was the economic crisis in Spain was terrible. And I had to leave by Adolith because there were no opportunities there. But I got a PhD opportunity in the Basque country, in the University of Basque country. And after finishing the PhD, I came back to by Adolith. And this is where I am still. I mean, it must feel great as well to to grow up with a team and then still maintaining, you know, and I guess they work at the same topics for 10 years or so. So they accumulate a bunch of knowledge based on projects and applied in different contexts as well. Yeah, the thing is that now that I have been, well, I know their research field from both academia and research institutes, I can see a difference in the way science is made in general, of course, you cannot always generalize, but I feel like in the university, because the professors have their position and they are teaching. Usually the research done at the university was done without a hurry, taking the time to do the things properly. And then they publish at the time at the beginning, very few papers, maybe one or two per year, not more, but very, very novel with great ideas. So the quality of each paper was very high. While in research centers, many times you are more driven by projects. And then you don't have this time, no? Of course, now our group when I entered was like that. But now we have grown when we are between 20 and 30 people and we are driven by projects. Yeah, of course. We are again in this dynamic, which is not very good, or is not the optimum for science, but the people which we had the privilege to start without a hurry, we got the time to accumulate a lot of knowledge, which you can use. It's like a stock. Yeah, I think that's very interesting, because when you deal with complex topics, you can't be in a hurry, right? I mean, there is so many interactions to consider and the choice is dramatic in terms of what the solutions are at the end. You know, you enable or disable word views or pathways and stuff like that just by including them or not. Sometimes it's not by voluntary, right? A number of people don't do it because they think they should exclude well-being from the equation. It's just because they don't know or it's a step by step. And I think that might be also one interesting way to enter into modeling. Modeling is quite difficult. So many times people just use a model which already exists. That's because it's easier. Because you take a model, you press the button, you get some numbers and you change one number, you obtain different numbers. If you do this four times, you have four scenarios, you can publish. So what I want to say is that there is a lot of inertia in the sense that there are some models which have been developed for many years, which are very well known, especially in big institutes. Big institutes have invested a lot of money and time in some key models and they are not going to get rid of these models so easily. Just because it's their tool. So developing new models of course is possible and it happens. But what I want to say is that it's difficult and that there is inertia against that. And also another thing about models is that many times there are many models which share a lot of central assumptions. You have a false plurality of models. There are several reasons for this. One reason I think very important is that each institution wants to have their own tool. You want to be dependent on the tool of others. So you have your tool, you do, you give a different name. Maybe there are of course always some differences. Yeah, so there is like the JCAM, the Mesa Jax, the DICE and all of these. Now for example I was thinking also in specific models. We talk for example in energy models, you have a plethora of models. Of course there are always differences but many assumptions are the same. Economic models, because then integrated system models is the putting together these specific models. So they replicate many times what it is already in the field. So just before, what is a model when they're used in terms of integrated assessment model? What does that mean? When you have one model, of course it's one model, it's code, it's a lot of lines of code and everything is interacting with everything. But when you are dealing with large models we need to conceptually organize a bit the content. And what we do is we divide by modules what we mean that are sub-parts of the model, which represent a bigger system. Yes, for example in EAMS we talk about energy module, economy module, demographic module and then you work with the interlinkages between modules. For all the models you more or less have the same modules every time or what? I mean every developer of an inter-designment model can have the modules that he or she wants. It depends on the dimensions you want to represent. For example I can just to be concrete, in William we have modules for demography, society, economy, finance, energy, materials, land use, water and climate. For example we don't have biodiversity explicitly model. If we will incorporate this, this will be probably its own module. With interactions of course with the other modules but conceptually this is I mean needed because to develop an internet assessment model requires a lot of work for many people. And you need experts of each module but you also need people able to work on the in the whole picture. And this is what's more my role I mean in locomotion. I am co-coordinator of materials and energy modules and also I am the general let's say not alone with another colleague in Valladolid Ignacio de Blas we were doing the scientific coordination and trying to make all the pieces match but it's not us alone but with the all the coordinators of modules. So in order to have a full integrated model you need to have a lot of conversations, you need to have a lot of workshops to try things, how to connect, how to make everything consistent. So from what I understand we do these models mainly is for a couple of things more or less to understand how everything is linked and how if you push there what is the impact there. But also more and more this is used for policy making right? So what is the the output once you have this model what happens with the model? Well the first thing to say is that a model do not produce results you need to test a narrative and quantify a scenario. This is important and I need to put inputs in so it gives you outputs. Yeah because a model is just a program it's nothing it's not going to tell you anything by itself and I will I like to give the example of the of the limits to growth reports where many people in the critiques to the model there were some critiques which could be valid but many most of them were just unscientific they they called the the world three model a model of doom. Yeah and this is very interesting because in the original report there are 12 scenarios and there are some of them at the end which are not collapsed scenarios. They are stabilization steady state but it became a famous sentence model of doom like always it collapsed and it's not true but it was a way to dismiss the model of course unscientifically. So when you want to do I mean you have a model which is designed for some purpose you cannot do whatever with the even if we try to represent many dimensions you cannot do whatever you need to be sure that it is ready to answer your questions. It's for that that this type of models you never finish developing them because okay you develop something you answer some questions and then come follow up questions but then maybe you need to add some functionality and then you need to continue developing so yes so then when you have your research question you set up a methodology to analyze it properly you get and then you try to get some objectives some goals and usually there is an iteration as you said in the sense that you have a narrative you parameterize your inputs parameterize inputs means put numbers to the inputs and then you run and you see if your goals are reached it may happen that imagine you have five goals and you reach three but then you need to come back maybe to the parametrization to make some policy more stringent or soften one policy which has unintended effects maybe you need to add policies and if this does not work what you need to do is to revise your narrative and what I mean with narrative with narrative I mean for example we have the green growth narrative you have post growth narrative and this is for all the model but if you are talking about transport more specific for example you can think about light vehicles narrative public transport so the narrative is the story because I mean a model of these characteristics has dozens to hundreds of inputs and each of them can have many dimensions because for example okay you say okay I have 100 policies for each policy maybe you need to parameterize hundreds also of numbers so you need to have a system that your what you are simulating is consistent so for example it does if you are simulating you are simulating I don't know in agriculture this meat to be grown you know I don't know hydro hydroponics yeah thank you hydroponic which is a very let's say intensive way and modernist which is aligned with green growth and then in another part of the model you implement I don't know in transport some degrowth policies this is nonsense because the society is one the society is not going to do different things it's going to work to to evolve in the future with some coherence so the narrative is very very important and here we can also come back to your question about central assumptions and one of the central assumptions especially in IPCC in IPCC models and this is a requirement from IPCC because even if IPCC is always in the press and in the street it identifies as a scientific body but the eye of IPCC is intergovernmental we don't have never to forget this the governments have a role and in the IPCC process for the models they are required that GDP per capita and population are exogenous so can you explain that that means that there are some pathways of GDP per capita growth which are not negotiable so the models answer the question IPCC models which technology we call deployment in many many sectors I need to do too much with this GDP per capita growth and a population pathway which is usually aligned with United Nations this is one of the constraints that they need to keep and that the intergovernmental panel is actually their given guidelines the scenarios from IPCC of GDP per capita and population until now this was like that some other some central assumptions will be for example in general these models function not all of them but most of them work under the logic of optimization and typically they have a carbon tax and they try to find the carbon tax which minimize or reach a target in temperature or in emissions well in our models we do policy simulation not policy optimization because we think that the world doesn't work optimizing anything and also because it's a very strong assumption that a carbon tax or a price can drive everything so we know that prices are very crazy that there are a lot of influences and also there is a lot of behavior of humans which do not correspond with prices so it's not that it is not related but it is not all driven by that so this policy this optimization procedures is one difference also which is very very important and then other limitations of models is that for example that they are blind to some dimensions which for us for example are very important and for example until very very recently no integrated assessment model had a representation of materials for me that's crazy i mean this just this is a you know i really have difficulties to to swallow this pill there is no physical representation of stuff well they have energy and they have technology mostly and then they have also land use land use became a very hot topic because of bioenergy no but still i mean without materials you don't have energy without materials you it's a stock and therefore the stock needs to maintain itself which will require new materials i don't know if you don't have a stock within your model what is happening as well another central assumption is the assumption of very very abundant energy yeah both non-renewable and renewable so um also they work on the material thing is nothing super important for them in general because oh more materials will be discovered and they will be mine if there is a price because they already have a demand the price will increase the amount of reserve so they are totally is the the paradigm of conventional economics that economic is a bit isolated from environment because they can't do whatever yeah and also there another critique from the IPCC models which by the way yesterday was highlighted by the policy officer is the lack of climate change impacts so by definition the IPCC models focus mitigation but it's not real we are already one point i don't remember two degrees we already have impacts so if you want to represent the reality you need to have this impact so and i think there is also when i was talking with the lorenz geyser he was also mentioning there's no convergence between let's say the the rich industrial countries and the emerging countries in terms of they should meet at the middle you know in terms of material requirements between the income the GDP or should somehow converge at a certain point right if we want to be just yes but the IPCC scenarios don't go in this direction yeah so it's funny how this is the narrative so it's like the IPCC does an effort to cover all they say they cover all possible futures it's not true there are other narratives but that makes sense but they are outside of the IPCC but there are other research groups which with our capabilities we modestly try to to test them so from what i understand the the IPCC gives a brief to all of the modellers we would like to test this this this and that run your models and give us what the scenarios are well it's not so direct or so like command no but the IPCC is already 30 years a lot of meetings to try to harmonize inputs outputs intermodal comparison projects these are relatively small community they know each other a lot they collaborate a lot so it's more the result of a community than the result of heterogeneous groups doing different things i don't know what is the the margin for being more pluralistic or or but at the end this is a political process if the post growth scenario is politically not a topic in the IPCC then maybe they can do some researcher can do something with these models but it's not going to be at the core of the IPCC but i don't know how much this can change because i have the feeling that for example in the european union many years ago these topics of the growth post growth were very outside and now more and more you see research grants or projects which are explicitly focusing on these post growth ideas at the european level i i don't know also how sustainable is this over time i don't know if in some years there will be some elections then everything will change but what i see that at the european level at least there is a acknowledgement this is maybe important we have to take a look we have to take care of what they are saying even if still we have our methods our inertia and our modeling teams doing yeah i know you're right i mean the amount of projects that are now funded in the post growth sphere there is five you are seeing post growth as well or more so i think it's relatively exciting i'm just afraid this is not just a buzzword that's gonna you know then die off for a new buzzword as we have seen in the past but i think there is also a more fundamental science of post growth starting to be established meaning that we now have the arguments we now have the models we now have the data we now have the narratives and i think there is a i don't think we are so mature no but i mean but we're starting right i mean we have the okay we have now all of the pieces because before it was a lot of narratives not necessarily the models or before we had models and not necessarily the narratives and i think there is a convergence slowly well my opinion is a bit different in the sense that we were i was in the last degree of conference this year i mean 80% or 90% of the presentations are qualitative social science and i'm i have of course nothing against social science but i have something against the disbalance because if you if we don't have quantitative methods we will never become concrete and then when you need to give a policy recommendation specific that you are not able you can only give some general ideas and these general ideas are great because are the basis for any quantitative analysis and we need them and without them we are lost but without quantitative analysis we will not manage to our progress further and there is an enormous disbalance but it's true that i see a growing interest in quantitative methods in this post-growth idea and we get a lot of people interested in the models we develop especially young researchers so yes i see a change but it's going to take time yeah yeah i know you're right and i think it's also hard for me because i come from the engineer and we are to to deconstruct my mind of an engineer thinking that with the right numbers with the right policy just things are going to change this is not true as well so we really need somehow to to bring forces together and also well every every field has its own biases we need to remove as much as possible all these biases and put together all the pieces we have not talked yet about uh william which is the name of the i am i mean the integrated assessment model that is if i understand correctly like the improvement of medea a previous integrated assessment model but is the one that you have within the local motion project what was a bit the rationale within the local motion project what were your okay something was not there yet and we wanted to firstly address some of the central assumptions that we just mentioned before but also that you wanted to improve methodologically from the previous i am yes so well in the medea's project which was a h2020 project with many tasks and objectives one of the important tasks was to build the medea's models which my group was fully in charge um and then when we finished medea's it's like we have the feeling that like instead of finishing something we were like not starting but opening a lot of door and new doors and also we wanted to have help with modeling uh to share the burden and also because we in the group in my group we don't have all the expertise for everything we wanted to model you know so we look for some potential partners specializing some topics so the locomotion project is designed as a follow-up of medea's with this idea of incorporating modelers at the core of the project it's true that the idea was to improve the original idea was to improve medea's but the reality is that we have made practically a new model in all senses so the remaining parts for medea's i will say it's not more than 15 percent was that because of because of new narratives that you changed so much or was it also scientifically some things did not make sense anymore or also programmatically we we just found new routines and we thought that was more important i think i mean when you prepare a project you have a plan uh you have a proposal and you have to be sufficiently specific uh so the reviewers can assess that you are doing what you the you are requested and so on but it's true that you don't write all the details so when you get the project um we are coordinators of the project but we are not directors so we cannot give orders to anybody this is not the idea the idea is that we coordinate and each partner has a say and we have to try to agree uh as much as possible of course when you are so many people it's impossible to agree 100 percent on everything we did our best some people are more happy with some parts others less happy with these parts um and then we have each partner has its opinion you know how to improve or so i think uh why so different is because there are some modeling things which are very capable and very ambitious and then they have great ideas that including ourselves that you try to do maybe more things that you plan at the beginning or also because i mean this is research so it's impossible to anticipate everything of course yeah so for example you think okay i will do this task and then when you start to do it then you realize there are some hidden tasks behind required tasks for this preparatory task if you want data or whatever that takes a lot of time and you didn't anticipate and this has happened a lot to us because especially for the linkages between modules because we had anticipated one full work package for the linkages but it has taken much more work than we thought yeah yeah because we are doing things that in many many things we are doing things that no one to our knowledge has done it so you have to try things maybe work no then also we have the computational problems because the model is very big we have a lot of dimensions so there are some things that maybe do know how to do them but they take too much time and then you need to simplify in order to have a model which runs in a reasonable time you mentioned the modules like we have land and water climate energy materials demography in society economy and finance you also had like a temporal scope and also a geographical scope i think what's something which for me was very interesting or very promising is that your narratives opened up to some less explored paths let's say so these are two in my opinion first is of course all of the justice element and this can be like a fair transition this can be a post-cross transition this can be an equality transition that type of sphere and then the other type of sphere for me which was extremely interesting but i have a huge bias is the material interaction between all of that right so when you're talking about materials of course it's it has an impact to everything to to land it has an impact to water when we see lithium what is the impact of that it has an impact of energy and it's also bidirectional right i mean you need materials for energy you need energy for materials and so it complexifies this so much more a comment to what you have said many of the feedbacks with materials and other modules are not represented in the model some of them we are working on them and some of them are outside of the scope not to give an idea that we have so much integration of course no but you mentioned i think like a what was it materials scarcity or mineral scarcity yes already this seemed to be like a essential you know yes no i was thinking you mentioned i think links between materials and land use well if in materials we include wood yes but if you are thinking about mining impact no because one thing i should say is that this is a model which does not have geographical detail we have regions yeah you have rest of the world india we have china so we are not able to see local issues you have what 40 we have nine big regions of the world which are let's say russia china india southeast asia we have north america we have many countries of latino america and then we have a rest of the world and then in the europe for the european union in some modules we have everything aggregated and in other modules we have the detail of each member state okay yeah i mean this regionalization is in our case driven by the availability of economic data when we report results we have to take a lot of care about this yeah well i guess some policies are good for everything and some others are very context-specific right or let's say economic specific to some of them but i mean when you model you need to do choices about simplification of reality and you do your simplifications based on your objectives in our sense our objectives were of course european union because we are here because we are funded by the european union project and we have a lot of detail of many countries but it's you cannot have everything so you do some choices maybe if you are more focused on land use for example then you will need to have another disaggregation another type of model whatever but in our case because for post growth i mean what really want to compare green growth and for growth you need to have a lot of social policies economic policies so you need to have you cannot forget this dimension which is very very often forgotten in green growth maybe narratives because the assumption is no society i mean we will do technical fixes no and it will take care of society yeah so basically well it's the assumption yes if you go to the street and you ask in most countries in my opinion you this will be a majority opinion yes but i think that's also it's also difficult to to say that because they're not given the entire truth you know or the entire yes of course but it's a situation yeah yeah but it's normal if you if you give some people just a tiny element of choice within a tiny agency as well they're gonna say okay let's do that which seems feasible but if you tell them do you want to also live well do you prefer living well or do you prefer uh you know having uh i don't know exploiting nature and society elsewhere you know people are never gonna say yes to that as well so there are hidden assumptions in in all of that i saw in the material module you had an example i think for one base metal which was copper and over there there was one node which i found very interesting which was called the demands from economy which kind of drives a number of things right and i'm wondering whether this also exists like in in energy and in some other elements because for me demand seems to or you know latest the latest report of ipcc kind of shifts a bit towards demand driven policies rather than supply driven policies and over there we it's kind of a trojan horse for sufficiency right i'm wondering how does that fit within the narrative was that already existing there or do you see the demand as uh an important lever to to tackle or question sufficiency stuff so your question is about behavioral change no in general because the i mean you have the demand you have the supply yeah if they don't match there will be some adjustment and in what you are mentioning about copper this is what we have there is a economic demand to produce to produce some goods and services to fulfill the demand the final demand in monetary terms but then you need raw material demand you need energy so this comes to arrives if we are talking about copper to the material module and the material module has a representation of the upgrades of copper and of course when the upgrade becomes degraded even if you have some technological improvement in general the price will cost because it's non-linear and then you will keep give back to the economy this price of copper and then there will be some adjustment but this is a mechanistic point of view right it's more of a biophysical this is a standard economics and well there are also some problems with this approach from a theoretical point of view but this is the way we have deal with in in the in the model because even if this approach can be criticized to my knowledge i i'm not sure to have a better option right now what we did in medias for example and we still are doing for some metals here not for copper is that because i mean the demand of copper we know very well from economy because we have a sector we have input output structure and we have a sector of production of copper so we have four material sector individuals which are nickel copper aluminium and maybe iron then we know for the input output that's for the rest of metals we have some aggregated sectors based on the rest of metals so this is all aggregated you don't see anything so we have some bottom-up calculations which come from the medias approach and what we do is to try to compute the demand of these materials and then compare with the available reserves and resources from usgs or whatever external source and then you track how much you are consuming of this stock and for some materials in some scenarios you you don't deplete the reserves in others you deplete the reserves and this is what we call scarcity indicator but we don't feedback the scarcity in the model in the sense that okay you don't have silver then you stop installing pv it will be extremely easy to do if then else reserves zero below zero um no more pv this is five minutes but the question if it really makes sense no because they are the data about the reserves of many metals is very bad it also fluctuates you know I mean if you do exploration and the reserves get bigger and how do you basically we didn't have sufficient time to analyze this issue of materials, reserves and resources so we took some data that we know are not robust so we think it's more fair not to feedback but give the information that okay if you believe that this source of information of data of reserves of silver you delete then you have uh consume all of it the message is the same if you feedback in the model or not you know it's like um the models are tools to make us think yes many times models are very misuse and it's interesting that there are many people I don't know if you have interviewed these people with this view in those podcasts but I would recommend because they have a point they criticize a lot models and they think that models are not useful and I think well there are always some theoretical points which can okay you can agree but I think in for many people they have seen so much misuse of models of course that they identify modeling as a bad scientific practice it's because as I said it's very easy to get numbers from a model but and the central assumptions can be that bad that can be hidden and if you are a user maybe you are not aware of them and you can misuse the model and you can not understand very well what you are doing you will always get numbers so it's very it's dangerous I think it's a guy a guy called box or something like that who said like all models are wrong some are useful but it's a famous sentence modeling in modeling field yes I think I didn't ask what your previous question about this the novelties or the strong points of william no maybe this is quite important so for us maybe the central point and that's the reason we call the model william william means within limits yeah so we are very interested about the introducing in the socioeconomic sphere biophysical potential limitations because when you take into account these constraints with different dimensions then it could reach the conclusion that green growth has many many challenges or many barriers or many problems that typical models don't see so if you don't your model is not able to represent these problems you don't see them and then it's feasible no so it's everything is interrelated so I would say these two points are important and then in from the point of view of methodology this is this alternative approach is represented in many things we have a lot for example we have an economic model very detailed then this facilitates the linkage with other parts of the model because if you don't have an economic model then you cannot do these interactions then we have dimensions that are missing as I mentioned materials society finance in terms of potentials we take into account available reason what we think are reason available availability of fossil fuels that this for example many times is not well understood because the idea is if we need to get rid of them why it's so important to model them let's say a peak oil thing but now we are using 80 percent of fossil fuels if you want to represent a transition from a state A to state B seriously you need to have a good representation of the starting point the fossil fuels are going to be very important in the next years because it's the base energy to do the transition with what are you going to do so it's important to have a representation to spend a lot of time modeling fossil fuels doesn't mean that we want them to be but many times it's not well understood I just want to stress in typical energy models the energy potentials are exogenous you do a literature review for a country and you take some numbers okay they say that you can install so much wind the bio energy but sustainable potential is whatever but the bio energy potential depends on land uses because you can if you plant a lot of forests and you decide to or you plant a lot of biofuels you decide to do these policies it will have consequences on food blah blah blah but the fact of having everything hard linked makes every story consistent is the same for solar of course there is a lot of room for solar panels but they will take room from other uses so it's important to have this interaction and to be able of the consequences and then one also maybe novelty which is interesting we're still working on it is that the fact that the model I mean the model another limitation is that the frequency is one year what does it mean that we don't see a summer winter variation day night in order to design an energy system based on renewables the state of the art say that you need at least early resolution one hour hour by hour 8,760 of course I already mentioned the computational problems of our model if you will multiply by almost 9,000 it's unfeasible so what we have done is to develop an emulator from an early model called energy plan an emulator we basically derive some regressions that we incorporate into William and that emulates the behavior of this early model in William so it's very novel and we have some preliminary results which are promising but still we need to finish and another thing that I think we are novel is the modeling of hydrogen when we started the project I mean when we started the project there was no covid no green deal package so you were thinking about it before the you kind of know we decided on the way okay so we this is one example of more work because the reality changes and also I would like to stress that the model is still not finished and we will release of course at the end of the project in December end of November what we have we have to disclose everything but it's not going to be possible to release a model as we would like evaluated and robust and I think here I should say that the covid hit us very hardly because the only way to make this thing work is to have workshops physical workshops where everybody meet and problems and we test things and with online meetings it was not so functional when it was again possible to travel we started to do every two three months one workshop with 10 12 modelers and then we progress a lot but we already had a delay yeah so yes we will release the model and we will explain what is the quality of that model and we will continue working in the next month to have a proper version so when you talk about biophysical constraints what do you mean I mean is it kind of downscaling of planetary boundaries or what is the constraint there there are different types one type is for example materials and even if it's a constraint you don't need to hard program that constraint and I gave the example of the scarcity indicator no so the idea is that economy even if it is the quantity is usually represented in euros behind that there are physical things materials energy kilograms of food kilograms of whatever so the idea is that the model is able to represent as a mirror or as two sides of the coin the monetary dimension and the physical dimension and then we are tracking we track things some things are hard linked others not dependent on the difficulty or so if you say you are producing x electricity we have the terabyte hours of the different sources and if you want to be 100 renewables and you have enormous demand and your country does not have sufficient potential then it needs to import or whatever but we account in this way in this sense all models account for that but the difference is that we are with our integration and some assumptions that we do that we think that are more realistic the potentials are for some of them especially for wind are lower than for other models so we also had this behavioral aspect you call it behavioral I am not sure it's 100 percent or what is the extent of behavioral element to it and how much is other so like sufficiency in the IPCC is defined as a set of political measures and individual actions or something like that that reduce for different elements within planetary boundaries and for well-being etc etc so you have different pieces but at the very top it says it's a set of measures first political and then individual right I think there is no consensus in the literature about the definition of behavioral change some people also talk about lifestyle change yes so in my group I think the understanding I mean we do a dichotomy between technological changes and behavioral changes in the sense that behavioral change is what you individual with your own behavior you can change so for example the typical examples are diet change not having a car not taking the plane of course of course but I mean you know the infrastructure needs to be there if there is a political wheel and there is cycling paths the your behavioral change or is much more yes there are enablers exactly the agency is completely different there yes I agree totally with you and we also take this of course into account but I can say that modeling this is a bit challenging we are exploring a bit for transport these enablers because of course if you don't want to take the train sorry the plane but the combination by train is three days that's the problem but also this is very specific and our models are not so able or targeted for these things so many times at the end a model you can capture everything so you need to have boundaries yeah and in our case this type of things many times are boundary so it is part of the narrative if we I mean if there is an investment in railway for example so then when you produce your report results you need to provide the narrative to be consistent so it goes together if you only provide numbers no it's not useful yeah no because I'm thinking for instance insulation right in buildings yes that's technological change for your opinion but in reality we can also capture it as behavioral change because you as a person reduce your heating right yes there is not so clear I agree but in the model is true that sometimes we need to be simplified to put some order yes yeah but that will be for example efficiency you are with less energy you are heating the same what will be sufficiency you put a sweater yes or you reduce the I'm not suggesting to do this but that will be behavioral change and then if you do both things then it would be a combination of two policies but in the model is like two factors that sum up okay I think it's important that people understand now that we have the model what do we do with it as well like what is the interface with outside of the modeling world right small break before the next part if you enjoy this independent podcast you can support me on tp the link is in the description below you can also help me reach more people by subscribing and leaving a comment many thanks and on to the next part so we've talked about the model and the modules of the model and a bit of the interlinkages as you said a model is not there to to tell you any truth or to tell you what to do you need interactions with stakeholders to do something with it in order to get valuable insights and of course we can think of policymakers that's the first type of stakeholder we can we can think about other type of stakeholders you mentioned at the very beginning that you were with civil society movements as well so I think it's important as well how people that are not experts deal with such interesting models with the media model we develop a game that we have been using a lot after the end of the project in our classes in our courses in university but also with civil society call it a game is basically a way that non-experts can run the model and then you can get a discussion as well but what do they do with this so do they have sliders and they say okay we're gonna do more of this more of that and yes basically yes and then they see the results most of the times they are surprised because the effects of their choices they do not match with their intuitions yeah because this is the idea of game you can I can do a presentation and explain what you get and this is very passive but if you interact with people in this way they get more involved and they have you know someone had one idea and then you are telling them he's wrong so usually people are curious why I was wrong or maybe you are wrong so then you give explanations and then I think it's the process of transmitting information is much better I would like a lot of this way of doing but it requires a lot of time a lot of time and a lot of people from experts to be able to moderate and lately we have stopped doing it just because we were so overwhelmed with William that this type of things then in science the first thing that suffers if you are stressed is dissemination because it's not your duty and Spain is not almost no valued in CVs or for positions so you need to cut somewhere it's first it's unfortunately dissemination I know doing this type of podcast while being in academia was a tricky situation yeah and then the idea in the project was to develop a set of applications to allow non-experts to run the model and here there are different layers of complexity we also had a game which was based on the idea of the medias game but because the model is not going to be finished then the results that this application are showing currently are not let's say fully reliable and we will like in the future to improve I mean to update this application with the model but this remains open because the project will be finished we can then talk about some of the policies that you have explored or you wanted to explore the impacts on well I would like to say something like I always say that a model is like a puzzle and if the puzzle has 100 pieces and you are missing five pieces your puzzle is not finished and this is the situation we have with the model that we have some important pieces missing and then we cannot use the whole model because of that but still there are very large parts of the model an interaction between modules that are working more or less properly and then we are able to extract what we call partial results so of course with less integration that we would like and this is an assumption you are able to produce some results and this is what we have in this delivery about policy recommendation no but just to make a disclaimer that where for example we test the universal basic income there are linkages to economy that are missing there in terms of climate change impacts or from an energy transformation and so on so it's more an economic model to test this universal basic income and these are partial results because you are touching select as you are touching some parts of the model not everything in order to build a consistent transition scenario you need to touch all the modules simultaneously in a coherent way this until the model is not fully validated and robust it doesn't make sense to do it perhaps the ones that you chose were because the modules were ready and all of this rather than maturity of modules and interest of the topic also with eb we we selected hot topic let's say for you that's the reason we have critical materials we have a hydrogen biofuels universal basic income degrowth green growth yeah okay so perhaps we can just explore some of them just to see what it means right so when I read the the degrowth versus green growth I see that some of the policies that come out of it is redefine the meaning of just transition redistributive policies minimum wage guarantee sufficient financial resources so like quite understood mechanisms right or not well accepted but unfortunately yet but you know if someone is doing that is already in the word of post growth these are some of the policies that they already have in mind but the power of the model is to test what happens once you do it right I mean this is the chapter where it is used I think out of greens because one thing in this deliverable is that we are using depending on the chapter William and there's also there is one using GCAM and now I don't remember if we have more but and this paper is very interesting because it shows that all each narrative has advantages and disadvantages there is no perfect solution I remember if I'm not wrong in the degrowth scenario the the depth I think increase a lot yeah so of course until we don't until we are not able to run the full William I cannot answer your question properly but my intuition even before seeing these results from pizza is that there is no perfect solution because the problems that we face are very complex and when you solve one problem in one part you can create problems in others unintentionally so I think it's quite interesting though you mentioned that when running this model we also see for instance the greenhouse gas emissions compared to today the degrowth is by far the one that achieves the most decrease so over there I think it's quite interesting to to see some biophysical some social but then you also introduce some economic or financial one which are the debt elements of course this comes out of our discussion of scientists to well should we just eliminate debts and therefore at the end of the day you have gene coefficients which is going in the right direction greenhouse gas emissions that are going in the right direction I can imagine it's very hard to to negotiate once you have these results right I mean at the end when you learn applying these type of models is the transition even if for many years it has been explained as a technological thing it's a social and political thing there are choices and different choices imply different outcomes and there is not a scientific absolute abstract way to find to say this is better no because it will depend on social decisions priorities if we for example talk about behavioral change and we need to shrink consumption to reduce for example emissions or in general environmental footprint I don't know maybe some people will become vegan but will take a plane I don't know once per five years because they really like to travel far but maybe some people love meat you know but they will never take the plane I don't want to say that all the solutions are behavioral change but what I want to say is that there are there are social individual and also social decisions about where the budget of the country goes this is totally political and social and you can favorize some sectors or others how much so these you're absolutely right how much they're entrenched in social political challenges I think there are some others that also go into completely different directions if I take let's say batteries right which well Europe is really going towards the decarbonization well of the energy system let's say then you add geopolitical issues you also add this flow versus stock dynamic like are we going to have enough at the right time of lithium so it's not even a matter of stock it's more of a matter of flow both yes I think it brings up biophysical geopolitical and political elements that are fascinating this is very good example also with choices no in Spain there are several proposals for lithium mines are people ready to have some lithium mines and having cars electric cars or they would prefer not private cars and no lithium mine of course it's a simplification but we cannot get everything yeah how do we navigate with these models what what how can they help us explicit right choices right because I think that's also one of the difficult element is that today we we think that the only way forward is by doing this a techno fix let's say and this implies well a new form of neocolonialism if you will from central from Latin America right we we accept that we need to decarbonize we accept we need a lot of lithium and therefore we we don't accept or we we don't hear about this will impact the the land use over there this will impact the society over there etc etc so how can these type of models help us better explicit the real choices that are in front of us yeah I think for me there are like two main types of applications first application will be to show that the green growth paradigm has some important barriers and this you do by integrating ideas or information that is not in conventional models I'm going to give you an example some examples now at the European Union level it's very clear that biofuels can be more carbon intensive over life cycle than fossil fuels and there's some circumstances but how how are you going to see this if you have an energy model not coupled to land use it's impossible that this model sees it it's impossible you can have is you are very very smart and you have a lot of knowledge and intuition about that but it's not obvious it's not obvious there is research which shows that the solar pv is already affecting global silver prices so you have a demand from energy affecting materials and prices of silver will go to economy no so how can you capture this and understand the the importance of the premise you don't have these three dimensions and then a third example which I like to also to give is the it's very famous is the rebound effect so engineers will tell you know I will isolate your house you will save 50% of your energy and then if everybody in the country will do this then there are a lot of evidence not only evidence with data analysis that shows that there are rebound effects and even backfire effects and there's some circumstances imagine that some person which saved this money for heating decides to take cheap flights to whatever for 30 euros every weekend at the end what you get is a rebound effect because you can only see that if you have economy and social behavior model so the issue that the renewable energy return on energy invested is going to decrease at the beginning at least of the transition they don't see all this and the second possibility with these models is okay with these problems but this representation of reality what could be the solutions and in this sense you can explore alternative narratives or you can change things I don't know so it's more like critique and the other is more propulsive it's arrives at the right time as well right you know I mean we also had a lack of imagination for a very long time and I think now we we need to propose stuff not only I mean we know that or we know a number of people know that green growth is materially not doable or you know the the coupling happens not at a sufficient pace to to make it worth in the 2050 window so we know all that but what do we do and I think there is a sterile ground that we need to to make much more fertile and also say that not only it's possible but there is some interesting policies that can also redistribute some interesting policies that can also have much more interesting benefits from the green growth which fixes the technical issue but doesn't touch the societal issue right there is a unspoken truth of the green growth that you know trickle down economics and somehow magically everything is gonna solve itself so the the hard connection I think helps to well it's always more is he always easier to criticize than to propose I think in every aspect of life you can see something that you don't like you will criticize what do we do I think this is this is natural well I'm not sure if there is enough evidence piling I mean evidence not in a scientific sense but in a volume terms because then when you talk with people or you go to conferences or most people are very surprised when you show this type of thing and then as we were saying before there are different profiles of people working in institutions no there are people who maybe are more let's say scientific approach looking for the truth but if you are looking at an institution which is politically driven you have to fit there or this is not your position so we are careerists which want just to please up to be up so even if we are talking about science well there are a lot of non-scientific things which are very important if we have to start wrapping this up I think I'm wondering what is okay so this has been an important next step the William model of course there are still some elements to to be finished all of that what's exciting for you in the in the pipeline like what do you think is relevant that we still need to do I think the urgent thing is to make it work and then get results we have already a lot in the pipeline and next year I'm sure we will have a functioning model in this project also this is the normal dynamic you spend four years developing something it's difficult to publish while you are working because you are developing of course you can always publish preliminary work this for example creates different problems in different institutions because in my case okay I already it's like I don't need so much more publications but we are students and psd need this publication so we are reserved we are in a university we are in my group and we are not a research center which has commissions something so we need to at the same time reach project goals academic goals group goals so it's complex yeah no but let's let's go back to yourself right you myself okay you were uh you went into so it's kind of a activist reunion that kind of sparked your interest to become a researcher etc etc let's imagine you are now the person that's kind of inspired you back in the day you have this model or how would you try to inspire your your past self somehow you know or well maybe an answer will be that alone I can't do anything I mean this is a group a teamwork of many people and each of us we can have some ideas but it's a group effort so it should we will have to agree on on what are the strategies and where do we put time but I think it will be nice to to spread the message and to I think in my experience to allow people playing with these tools it's very very instructive in terms of let's say impact at higher levels this we should think about the strategy because of course we are a small group we are not we don't have connections let's say up so we should build them this I I cannot say much yeah yeah no but I was thinking I mean I'm putting myself in your shoes like what do I do now that I have this you know this this big tool do you try to propose like even more aggressive policies and try to see like try to make a set before getting results I mean these tools are tools for planning the transition but we don't have to be naive to think that now in 2023 I can't plan anybody can plan the transition until 2050 you know like you write everything and then you go on holidays whatever no it is a narrative process every time there will be more information better data feedbacks that you didn't realize that are important and are missing I think this should be a interactive process with policymakers in order to adapt the plans to the most updated information this is also why it is socio-political and this tool is helping yeah it's not telling you in an optimization way what for me is nonsense in general I always try to ask for a recommendation or an inspiration in terms of a book or in terms of a movie or in terms of an article or something that perhaps you in your career it kind of helped you to something positive well no inspiring in in sense of okay I learned something from it or it helps you grow in in in the in the direction right of course you know the word 3 is is said to be a doom scenario but at the end of the day it helped us grow very much and it's not a doom scenario that was one of the key turning points in your career reading the limits to growth or certainly I would recommend to anybody interested in modeling I think the best version to read will be the 2004 written mainly by donel amido sub then she died and this is a book about modeling in which there is no one equation and it's going it's suitable for anybody interested in the topics and really helps you understand what is a model how to use properly a model how to play with a model forgetting about this obsession for numbers and quantification quantification is a means to achieve your objective of understanding the the system yeah so this book is very recommendable and maybe another book which for me was very interesting is a collapse from gyre diamond which is a classical anybody who likes history anthropology is worried about the environmental crisis we'll find examples of bad developments and societies which manage to avoid worst cases scenarios okay yeah great I mean of course from the nela meadows I would also highly recommend the thinking in systems book which you know at the end of the day we're just and also maybe if I can recommend third book yes cro potkin the mutual help how do you translate in English I don't know yeah I think this book has over 100 years but it's amazing and also because it was published to respond to a scientific paradigm which was dominant you know about the misinterpreted Darwinism to show how a societal dominant idea can be completely wrong yeah I hope we're at that level right now but in any case many thanks in you go further this conversation and also thanks to you all for you know sticking with us until the end I think these are complex topics but if you like this conversation what I would suggest is to listen to one or two others that that we have made on the podcast the post-growth project from Julia Steinberger Jason Heiko and Yorgo Scales we talked about degrowth models as well with Lawrence Kaiser so yeah just go explore more and and if you want to to have a further discussion or comments thank you for the invitation it was a pleasure