 Okay, good morning everybody. So we're gonna start. I'm a bit surprised that there are so many people missing this morning. This is an important day because you will have case studies today. First Augustin, so we had some changes in the program. You received an email yesterday, probably. Maybe all people have not yet seen that. And I repeat that we start at 8.45 every day, because of the first shuttle. So this morning we'll have again Augustin Collette and Olivier Chanel to see some cases. And then in the afternoon we'll have more. So we'll make some groups and you will have to do some work with a laptop. So later on we will propose the different groups. So Augustin, please, you are going to start with a presentation and then we'll have the practical exercise. Thank you. Good morning everybody. So as a follow-up to my presentation yesterday on air quality modeling, I will present you what we do with these models in terms of assessment of the best practices and the best strategies to mitigate the air quality. So I come back to my suite of tools for integrated assessments when we start from human activities that lead to air pollutant emissions that we convert into air pollutant concentrations to assess the exposure of the population and therefore impact. And as I was telling you at the end of my presentation yesterday, the most important is perhaps to go in that chain of tools backwards, to go backwards, to go from the impact back to human activities and try to have an impact to improve the air pollution. So the way we do that, there are two arrows on this diagram. The upper arrow is where you basically do different trials and errors. So you test successively different scenarios. So you have different hypotheses about human activity and you want to test for example what happens if I ban diesel, what happens if I set such or such threshold for industrial emissions and so on. So here you can just make a list of measures and you run them through all your integrated assessment model to assess impacts and then you will compare those impacts. That is what most people do, that is what I do most often and that is what I will present largely today. The other option is what you have in the lower arrow where you have an optimization algorithm which is what is being used by the GAINS model. The GAINS model is quite important for air quality policy making, especially in Europe. It has been used for 20 or 30 years in Europe and now there are several versions that are being developed also for China, for India and it can be developed also for other parts of the world. And the way GAINS works is it has all these items in one model and the useful part of that model is that it is embedded in an optimization algorithm. So you can basically run all the different options in one go and the model will tell you directly what is the most effective human activity lever to have the most impact. So you know where is the leverage that you should put directly with the optimization algorithm. So it is a mathematical problem. As long as you have simplified all the different parts of your integrated assessment tool you can design an optimization algorithm and to try to find what is the optimal strategy to have lower impact. So GAINS is a very useful tool. It's relatively complex. I just wanted to mention it today but I will not go in the detail of that algorithm but you should know that it is a tool that is widely used in the policy. But again so today I will really focus on the trial and error part of the assessment where we basically test different scenarios and try to assess according to various criteria which which scenarios are the most efficient. So what I will present today will be divided in two parts and then there will be the case study. First I will present a few of these assessment activities either for very long term or more like mean term. And then I will go to very short term looking at forecasts and how we can play with these scenarios with an air quality forecast model. And that will be the case today part when I will ask you to contribute actively. So starting with the long term I'm not sure to be honest that you really had presentation before during the week on interactions between climate change and air quality did you? Not so much. So let me just say a few words. Climate change and air pollution are different problems but they share part of the of the stake. So climate change obviously greenhouse gases, air pollution, air pollutants. The thing is there are some air pollutants that are also that greenhouse gases that they have an impact on climate. So I'm thinking about ozone black carbon. They can have a radiative forcing and therefore they can have an impact on climate. And also climate change has an impact on air pollution because climate change at the end is just the statistics of the weather. And the weather has an impact on air pollution. You can have some weather condition that can be conducive to air pollution events, heat waves for example. Therefore if the climate changes air pollution is likely to change too. So you have this feedback loop that makes everything a little bit complex. And then the last point that makes everything even more complex and more interesting is the fact that atmospheric mitigation policies will have an impact on both. So mitigation policy can be either designed for climate or for air quality. So if you design it for climate you will play on greenhouse gases. If you design it for air quality you will play on air pollutants. But the thing is you don't really play with atmospheric species. You actually play with people and what they do and emission sources. So if you target greenhouse gases you're actually going to target a source and that source will also be emitted air pollutants. Let me explain. If you want to curb greenhouse gases emissions for traffic you will target cars and say okay I put more electric cars. But once you put more electric cars on the road you have less air pollutant too. So your climate policy is going to have an impact on air quality. And it's also the case when you design an air quality policy it's going to have an impact on climate policy. So all this is linked. So in that framework you have interactions between climate and air pollution in the geophysical part of things. Climate as meteorology has an impact on air pollution events. And air pollution as radiative forcing has an impact on climate. That's geophysics. And you also have an impact on policies. When you design a mitigation policy it can have an impact on air pollutant and greenhouse gases at the same time. So all this is quite interlinked. That's why we designed this even more complex integrated assessment framework. When we start at the very top with different scenarios, storylines as we say, from different possible futures either in terms of climate evolution or air pollutant evolution. And we will feed these storylines into a number of models to assess their impact. And those models would start with the IPCC type of coupled climate models. Also chemistry climate models to have really the long-term evolution of the tropospheric chemistry. But then if we're looking at air quality we want to have a finer view. We don't want to look at the whole globe so we look at a regional climate model. Regional being like Europe or one continent. Regional climate model that you use to drive your regional chemistry transport model. I was mentioning yesterday that you use to assess your health impact ecosystem and various type of impact. And once you've done all that you can go back to the source and do what we call this cost-benefit analysis to see how much it costs you to have your climate mitigation and what is the benefit you have for society. So yesterday I spent quite some time to explain to you that one of these bugs was a complex model. I'll let you just imagine how complex it can be to have all of these bugs put together. But I will try to simplify and just show you the result when you do that. So we start with the storylines, the climate policies. We take emission projections for various targets. Here I'm sorry you can't see it but we use the projection of the global energy assessment that was done quite some time ago already in 2012. And that work really a large part of the work that was done for various projections in terms of energy consumption in the world. But we are focusing on Europe and we selected two of these scenarios, one business as usual and one mitigation, climate mitigation scenario. And if we look at the primary energy consumption in Europe, so those two scenarios are quite different for the business as usual it increases energy consumption and for mitigation it decreases. And that's why mitigation can achieve this two-degree target at the end of the century. The most important, not the most important but there is this difference in total consumption but there is also as you can see a large difference in the share, the energy mix. So for the business as usual you still have a lot of fossil that are being used. Whereas for the mitigation fossil is really going to the lowest part and you have a lot more renewable energy sources. So these two scenarios are very different for climate because one of them manages to remain below the two-degree target and not the other one. That is the main scope of these two scenarios and the work was really done to assess climate mitigation. What we did is to look at the side collateral benefit for equality, if there was a collateral benefit for equality. So we plug these two models in our integrated assessment model to see if what would be the impact on our quality and obviously I will not be here to explain you if there were no impact. So we are looking at ozone, that indicator for ozone, summer 35 in the historical situation, so basically the present day and the two scenarios for the future, the reference which is basically the business as usual and the mitigation scenario. As you can see there is virtually no impact if you change with the reference, but there is a clear improvement if you go for the mitigation. So once you've run your air quality model, you have your new maps for the future, you can assess your health impact, you have a similar thing for a particular matter and you put your air pollution into your health impact assessment tool, what you've been shown in the previous day. So you have air pollution, population density that gives you exposure as population weighted pollution. If you combine that with your relative risk functions and baseline mortality, you have the rate of mortality, you can combine that with life tables to have the years of life lost that you can ultimately transfer into economic coal valuation as Olivier was explaining you yesterday. If you do the valuation, you have this analysis, cost-benefit analysis, sorry for the title, and you will see how we can compare what is the cost of mitigation and the benefits that you can expect for the society. So if we start with the mitigation, we have two different scenarios, one with no climate policy, one that has climate mitigation at two degrees. Obviously one is more expensive than the other. So here we have the cost in terms of maintaining and operating the energy system annually in Europe, the cost is high, and it's even higher if we consider the climate mitigation scenario. We have a difference of 100 billion euros of the climate mitigation scenario, 100 billion euros every year to operate an energy system that is more climate friendly. But when you operate that energy system, you have a lot of savings for air pollution mitigation. What I was explaining you before for the electric car, if you produce a diesel car, you have to put a particular filter. That particular filter has a cost. If you design an electric car, you don't need to put the particular filter. So you save on that side. So air pollution mitigation costs are lower under a climate mitigation pathway. This is the same for energy. If you design a coal power plant, a coal power plant, you need to add filters and depollution mechanism. If you design a wind farm, you don't need to have depollution mechanism. So you save. So basically that scenario, the climate mitigation scenario is cheaper than the no mitigation scenario in terms of air pollution mitigation. And then obviously health costs. So the avoided health damage, what I was showing you before, we have our maps of air pollution. We plug them into a health impact assessment. And we have a difference in terms of health damage using the valuation that Olivier was showing you before in terms of years of life lost. And here you have a very large difference between the no climate policy that has a larger cost for society and the climate mitigation policy. And if you compare all of them, you can see that you relatively balance the two. And therefore, you can conclude that it would be quite valuable to mitigate climate because you have a benefit for the society. But there are a few things that need to be considered when you do this type of analysis. And you should remember yesterday that Olivier was telling you that in terms of cost benefit analysis, the time lag, the delay is important. So on that plot, the delay is not taken into account. So the money you actually invest today that will benefit for your society in 50 years from now, that can be a large source of uncertainty. You have also, you're comparing here the cost and the value, the actual cost of putting a new energy system is something you actually do need to pay with real dollars, so to say. And the value for society is a different concept. But nevertheless, when you compare the order of magnitude, you can witness the fact that those numbers are very large. And you should remember what I was telling you in the beginning, these two scenarios that I'm mentioning here, they are not designed at all for air pollution. They are designed for climate mitigation, which is a big deal. And if they manage to mitigate climate, it's already like a huge benefit. I'm sure you're all aware of that. If we manage to stay below the two degrees, it's a huge benefit. What I'm just saying here is that that huge benefit is also compensated by a collateral aspect that was not sought at all when they designed the scenario. It was not expected from their side that you would have such an important collateral benefit for air pollution. We should not claim, we should not overstate what we're saying. We should not say, okay, air pollution is going to solve the climate problem. The benefit you will have for air pollution is going to be so huge that climate mitigation would be a bargain. But we should still remember that there are such important co-benefits between the two types of policies. Now I want to move to another part looking at both intermitigation, but maybe before, since I'm finished with this part on climate and air quality interactions, I'm wondering if there are any questions. Maybe it's too early for questions. So do we have now scenarios for improving air pollution together with climate? Do we? That's what I'm going to show now. So I will show that just for Francois. So now I was looking at, and relatively long term, the middle of the century. But we should also encourage action at the shorter term. We should not wait for 50 years before doing something. And a lot of the air quality policies are actually designed more for 2020 or 2030 type of time frame. It can be considered a really long term, but actually by then the climate penalty is not so important. So when we look at air pollution mitigation for the next 5 to 10 years, we are actually neglecting the climate impact. And I want to mention to you a work we did for the French Ministry of Environment to try to provide a ranking before the different options that you can consider for air quality mitigation and how we do that. So there was a national plan that was ultimately published but one year ago in France. It's available online. I'm sorry, I forgot to put the link, but if you Google what is on that and if you're able to speak French, you can have a look. It's a work that was initiated by the French Ministry of Environment, but the way it works is actually it's a consortium of various bodies that get together. So you have people who are experts in energetic prospective or air pollutant emission assessment, air quality modeling. So that was my team and health impact assessment. And for various hypotheses, what we do is we do other various impacts in terms of legal consideration. So if you take a measure, if you want to take a measure, can you take that measure? Is it legal according to the directive you are bound to to actually take that measure? What is the acceptability in terms of social acceptability, even if the ultimate decision always remains with the policymaker? What is the expected benefit in terms of environment, so air quality indicators? And what is the benefit in terms of economics, especially also taking into account the cost of the measure versus the benefit that you expect? So we had a long list of maybe 40, 50 different measures. I'm sorry, I did not translate all of them, but those are the type of measures that we would look at individually. And I will highlight specifically those that are in the red circles. For example, for the road transportation measure, we will look at what is the benefit of implementing the Euro 6 norms, both for light and heavy duty vehicles. Or we will look at measures that would reinforce the fraction of electric and hybrid cars in the vehicle fleet. Or the third measure here is increasing the taxes on fuels. So you have different type of strategies. Here I just put the transport measures, but we have the same for industry, for residential sectors, agriculture, and so on. So when you plug these measures in your emission model and then in the air quality model, you can produce these maps. Here it was so Schimmer, the chemistry transport model also with data fusion, where we can have a map of the difference attributed to different measures. So here we're looking at the PM10 difference for the high performances wood stove. So you see that you have a larger benefit, especially in the Alps area. And if you look at the benefit here, for example, of reducing the access to the city centre for the road transport, you have a larger benefit, especially in the Paris area. No big surprise. But still you need to have those quantitative maps if you want to assess health impacts on a quantitative manner. So what you do is when you have your air quality, air pollution concentrations, you can actually go to the health impact and then to the economical evaluation. But staying for the moment on air quality, a big problem we have in Europe is the exceedances. I'm not sure you talked about that this week because for most people, air pollution is a problem of health or ecosystem, actually impact that you can see. But the way the European legislation is written is defined with thresholds. And the number of days where you exceed the threshold is a problem because there are also some legal issues between the commission and the state parties, the member states. And those member states that have too many exceedances of those thresholds can have to pay a fine at the end of the day. So meeting these exceedances levels is a big issue for the various countries. And with the air quality model and data fusion, we can actually take the measure we have in the plan and compute the exceedances. And for example here, you have the situation today, the red situation today. We have exceedances basically for everybody, for N02, for PM10, for PM2.5 and for Ozone. And when we go to the preparer plan for 2020, or there were two versions with a more ambitious preparer plan, but even the basic plan for 2020, we realize that the exceedances, we really go down a lot, especially for PM10 and PM2.5, you basically go to zero exceedances by 2020. So this, policymakers like a lot. They say, okay, the plan is well seated and we will go to that target. I said the plan was actually signed in 2017. We are today in 2018. 2020 is very soon. We are not there yet. We hope we'll make it, but this is still a prospective experiment. But that's the first analysis you do. When you play with your measure, you want to look at exceedances to see if you can go below the exceedances level. Then you want to have something a bit more synthetic. So we put all the different pollutants together to derive a synthetic indicator. And then you do also your ranking of your different measures. And for example, the most efficient measure is one of those that I had highlighted before, the Euro 6 dots. This is clearly the most efficient measure that you can take, assuming that Euro 6 actually delivers. And then you have different other norms. Euro 5 is just there. You have also here, for example, flambert is for the residential heating, the use of wood burning. And you will see one of those that I wanted to highlight before, increasing tax of fuel is here. In the top six, it's not bad. It's quite a good ranking. And then you have also the ranking for all the other ones as well. That's just the impact on air quality. But then what you want to look is net benefits. So do your impact on air quality, you translate into else impact, and then you translate cost evaluation, and you take into account the cost of the measure. The cost of the measure can be a benefit if you increase the tax, make money. So you have that measure that was not the best one, increasing tax on fuel here. It was not the best one, but in terms of cost benefit, it can be very good. So that's the first thing you learn. The second one is the Euro 6 norm, the second best. So this is also not that expensive, also because basically we prepared everything and we know it's manageable, we know it's doable. So it's not that expensive. And the last one, the last measure here, very expensive and not so huge benefit, electric and hybrid vehicles. But that's a little bit of a misleading aspect because when you put electric and hybrid vehicles, it's very expensive to support electric and hybrid vehicles. Today, it's still expensive and for the state to give incentive is expensive. But the thing is all this cost should not be paid only for air pollution plan. It should be shared with the climate plan. So here, when we do that exercise, we assume that all the costs of that measure is expected to be compensated with the air pollution, whereas it should also be compensated with the climate part of things. So it's one misleading part of the exercise. Of course, the ministry asked us to do that. We do the way they want us to do. And we always come back to them saying, okay, you should discuss with your climate colleagues. But that's a long story, always the same. But it's just to show you an illustration of the various ranking that we do and how taking into account the costs can change the ranking between your measures. So we have a small problem with the camera here and you cannot see my diagrams. But we also did, I was telling you before, beside this ranking, the quantitative cost-benefit analysis, we also did some ranking considering legal aspects and social acceptability. Here, you have these matrices on the X axis. You show the environmental, economical aspect, if it's one is very good. And then you have the social acceptability and the legal feasibility. If you start with the bottom one, you can see that the electric vehicle, in terms of environmental aspect and economical valuation, it's not so good. It's the last one on that list. Okay. But in terms of acceptability, it's very good. People love electric cars. Here, you have a different one. Euro 6 and Euro 6 now. It's very good in terms of environmental measures and also cost-benefit analysis. And it's also very good for legal and for acceptability also because it is already planned. If you buy a car today, it is already Euro 6. So people accept by fact this matter. Unfortunately, you don't see this one. But I already told you it's very good for environmental and for cost-benefit analysis, increasing tax. I let you guess that for social acceptability, maybe it's not number one. But that's more like a political decision to make, where you make your threshold in terms of acceptability of other people. I'm finished on that part. And that was it about the long-term assessment plans that I wanted to show you, how we take the measures, the type of indicators we can derive, how we do the ranking between the measures. We should highlight that this work we did for this French national pré-plan, it just supports decision-making. A public body like mine cannot be decision-maker. The decision always remains to the policy stakeholder. We do support. We do a ranking of the measures. But then at the end, the one who is going to take that decision about increasing fuel tax, for example, is not going to be a scientist. It's always a political decision that remains. So I will move to the forecast part. And then we have a little bit of more interactive work. But maybe there are more questions now. I don't know if you are going to talk about it later, but I just wanted to ask, in the optimization problem, when you have to assess the air quality, you don't use a CTM model, right? Do you use a surrogate model, something faster? So I just wanted to know, because I know gains, but honestly, I don't know so much. And I don't know which kind of surrogate models it's used to assess this air quality. Surrogate models. Do you have a question also on surrogate models, or it's a different question? Okay. So I will answer about surrogate models. It's a very nice field of study, actually. It's not very new. So gains was originally the Reims model. It was started to be designed in the early 90s. So it's not very new. Still, it's very, very efficient and very useful. In gains, you have several parts. You have, it's integrated assessment. So you have the whole chain. You have the list of measures. You have surrogate model. I will come back to that. And at the end, you have the health impact and the economical valuation. And all this is embedded into one single model to do the whole optimization. But as you say, you cannot afford to have a full model there. So you have this surrogate. The surrogate of gains is designed with the MEP blame matrices. Those blame matrices is what has been designed in the 90s to assess the contribution of one country to the pollution of the neighbor. It was initiated by the issue of acid rain. That was a big problem, especially for Scandinavian countries, when they realized that a large part of the ecosystem effect they had on their forest was actually due to pollution emitted in the UK, in France, in Germany. That was transported to Scandinavia. So the Scandinavian people in collaboration with the other countries set up a new body that was signed under the Geneva Convention to assess the contribution of these countries on acid rain in Scandinavia. And to do that, they use a CTM. They use the chemistry transport model MEP to try to see what is the contribution of the UK to the position of sulfur dioxide in the Scandinavian forest. You do that with a CTM, but you also do a lot more simulation with your CTM. Hundreds of simulations, literally, to address the contribution of every single European country to every single neighbor. So you do all the combinations. Once you've done all these simulations, it's a huge work. So really a lot of modeling with a CTM. Once you do that, you compute your blame matrix. The blame matrix is just the fraction that one country is polluting the neighbor. Those blame matrices are available on the web. You can download. Everybody has a look. It's quite a sensitive issue when they publish every year. And actually, part of the motivation of the work I was showing you before is that for the French government, we are being asked to challenge those matrices. It's one of the motivations of the government to support the development of a model like Schimmer is to make sure that what is in the MEP model is right, or more or less right. Because if one day the Scandinavian people are telling us the blame matrix for France is so much for Belgium, for Germany, blah, blah, blah. The French government says, no, I don't trust you. Your model is wrong. And I want to be sure that what you are actually accusing me is right. And that's why a lot of European countries support the development of models so that they have the in-house expertise on the modeling to challenge what is being said by the MEP model and by Gains. So that's why you have this quite nice interaction between the different modeling groups in Europe, because we are challenging each other. So you have this one model with the blame matrices inside. It is used for European policy, for the Geneva Convention, that's a UN policy, to assess the blame of the different countries. But if you want that system to work, individual countries need to have their own legitimacy and need to be able to challenge what is being said. So this is why we have bilateral negotiations all the time between either the UN or the European Commission. I think you said you can reduce the pollutant by putting some costs over the fuel. But we can't think in another way. If you reduce the source of pollution, for example, reduce the sulfur and the crude oil or in the fuel oil, there is some units you can add in the refinery company to reduce the sulfur and the crude oil on the fuel oil. I think this is the better way to reduce the source of pollution. I can think of an even better way. Can you think of an even better solution? Yeah, I think if you reduce the source of pollution, better than another cost. You're right. I mean, there are various options. If you want to reduce the emission of fuels, there are a lot of different options you can take. I was just giving an example, and I had like five or ten measures, but in the plan we had 40 measures or more. And the thing is with sulfur content of fuel, we already did remove a lot of them, a lot in what is being used in the road transportation. There is basically virtually no sulfur anymore. And I was talking about acid rain. Acid rain is not a problem anymore in Europe. And we really have decreased that. I mean, I think since the beginning of the 1990s, in 20, 25 years, we reduced acid rain by a factor 30. 30 times less acid rain. It's huge. I mean, people are asking us, you know, you're talking about the environment, you have people are afraid, it's a concern. I mean, there are huge success stories as well. There are things we can solve, problems we can solve. Acid rain is one of the, I think, the best examples. So there is virtually no sulfur being used for car traffic. There is still a lot of sulfur being used for shipping emissions. So the maritime transport, that few use still a lot of sulfur. And we are negotiating to have less sulfur in those. But so there is a measure that is reducing sulfur in shipping emissions. But that's not part of the assessment. Me? Okay. So I was wondering Augustin, whether there are plans now to have an assessment of cost and benefits, combining effects on climate and air pollution? Such plans at the national level. The quick answer is no. There is, we are working on ecosystem impact at the moment, actually, more than climate. Because the timeframe is 2020, 2030, it's too short for climate impacts. And everything I was showing you today had economical valuation for our health. We want to add economical valuation for our ecosystem as well. That's the main part of the development of the tool at the moment is to look at ecosystem, particularly anthropogenic ecosystem, agriculture. Ozone has a very large impact on yields of agriculture. And that can be monetized. We're talking sometimes, you hear numbers like 10, 15% impact on European yields due to ozone pollution. And that can be quite true in terms of agricultural impact. So this is a big part of the work. If we, the priority to work on economical valuation is to look at ecosystem at the moment. But climate is not so much in the target because of the timeframe for European policy. But there is a very nice, I can forward the reference, there is a very nice European report to the European Parliament using gains, using co-benefits for climate. That's the best example I can think of. When they look at the co-benefits in terms of cost benefit analysis of air quality and climate mitigation. Really in one go. And that's a nice story and also a quite complex one because of the role of methane in that context. I did not really discuss in detail, but that's a very important precursor to be taken into account when we talk about air quality and climate because methane is a precursor of ozone. It's also a very efficient greenhouse gas that happens to be relatively mid-sized, mid-life greenhouse gas, lifetime of about 10, 15 years. So it's a quite complex beast to be taken into account. And we don't even know in which legal framework we should put methane at the moment. It's only being taken into account by UNFCCC, the climate convention. And there are some discussions whether it should also be in the air pollution conventions. So that is the work as it stands today. It's being dealt with at the European level, but for national policies we are more focused on relatively short term. Thank you for the presentation. Maybe I missed the one you were talking about, but your modeling was done before or after the diesel gate. So about the emission factors because your model is based on emission factors. And if you go for Euro 6 diesel and now we know that they are not so clean, then your model should be run again. We know that and actually the emission factor that we use are not those of the norms. It's been a long time that we know there is such problems. So when we take the Euro 5 or Euro 6, there are some corrective factors. We're not saying it's perfect. To be very honest, it's very difficult to actually know what is happening in the field. But the emission factors that we are using is not the plain ones that are provided by the car designer. For a long time now, there has been some corrective factors and we already take that into account in the model. But then the point is how you can compare in Europe if anyone is correcting his own emission factors, I mean among European models for example, you are correcting the emission factor for Euro 6 diesel. Germany is correcting his own emission factors. So maybe the differences are not as you were showing yesterday. That is homogenized. The emission factor methodology is defined by European guidebook of the environmental agency and you cannot use whatever you want. And those emission factors, they are not exactly derived from the norm that is the problem here. Thank you. You have reported that air pollution mitigation policies impact to reduce the climate change costs. But I wonder why in the government reports, the climate reports put the emphasis more on other actions than the air pollution. For example, you realize that there are some actions that are more important than air pollution actions. I don't know the answer to that question. I think I see what you mean. When you see the importance of the co-benefits, you really want people to take that into account. I can tell you that we've been promoting that a lot to the ministry. We have really good relationship with them. But still, for them, it's a long shot. All this climate interaction, when you talk to policy and you put the number in the middle of the century, it says, okay, goodbye. I have something. I have my term. It's like five years. I want to do something for the next five years. And it's more difficult to develop support to that. Hi. Good morning. And thank you for your presentation. So I have a few things to say. You have shown us the global model yesterday. And I find I'm from India. So I was very worried about my India's condition or your model. So that is one point. But when you are showing the maps or your actual work, it is coming to Europe. So for me, I would have been happy if the data is coming from Asia as well, from India as well. And second point is that India has very distinct four seasons, climatic seasons. Now it is rainy season is coming. So when it is rainy season, the pollution is very low. So when the data is collected, that is also a big issue maybe in your data model, global model. And second thing, if we think about the climate change, we have some positive benefits, I tell you frankly. Very, very nicely. We are happy with the change. Why? The thing is we are having very good cultivation over the mountains of Himalaya. So all the apple, orange, tea, coffee, it's going up. And we are getting very good economic benefit out of that. Secondly, the landfill sites were the solid waste disposal sites. You won't believe that those landfill sites, some farmers are cultivating vegetables. And those are really good quality in taste, in size, in look. So maybe in Calcutta, we get 70% of our vegetables coming out of those landfill sites. And those are the best in shape, size and quality. Say good economic return is coming from that. And secondly, since 1991, India is monsoon dependent. You may be aware of that. We are very much dependent on monsoon rain. So 70% of our economy coming from agriculture. So since 1991, we have not faced any drought phase. So we are having very good economic return. So how this thing can be calculated in the model? Cost-benefit analysis of climate change and pollution as well. Thank you so much. Thanks to you for the question to answer first to that part. I don't know if I mentioned the term today. We talk a lot about climate penalty, but there is a climate benefit. There are parts of the world, part of the economy that are going to benefit from climate. It is true. And the problem is how you weigh things together. Some of the benefits you mentioned is more short term. It can be assessed. I mean, okay, you're saying your agricultural production can increase. It could be the same, you know, for these maritime ships that go across the North Pole economically. It opens a new route. It's faster. It's very easy to quantify the benefit of that. There are more other aspects of climate that are not so easy to quantify. And the very large penalty that we expect around the world with heat waves and, you know, displacement of population is more difficult to quantify than a short term benefit for agricultural productivity. That's the problem. But I'm not saying that all the climate is only a penalty. There are some sectors where people expect a benefit. You had a question about the seasonality. So this is something in India you are saying, okay, you have to watch when you do your modeling during the rainy season and all. So when we do this type of modeling, we always do annual averages. And sometimes even we take averages over a longer time period. But again, I'm sorry to have shown only examples for Europe. And I will continue in the remainder of the presentation because this is my job. But the good news is that for India, for example, there are also similar activities that are undertaken. And in particular, that gains more of the blame matrices that I was mentioning before. They developed a specific version for India. I don't remember the project, but the tool is now ready. I can try to find a reference if you want to. And there will be some similar analysis also for India that can be done. Okay. Thank you for the presentation. I have one question. Yesterday you presented the assessment pathway. And one of the main point was to assess the human activities, to go to impact on costs of air pollution. And today you presented the hybrid gas, the vitu hybrid, as one of the ways to reduce the pollution. But I know during the construction of this gas, there are too much pollution done before. Is there any, it's not a shift of pollution? And do you consider that on the cost of pollution, that part of pollution was done before? We don't. That's what you would do in a life cycle analysis. If you do the life cycle analysis, you take a product and you go from the production to the use to the disposal. And you do that during the whole life of your product, what are the different environmental impacts? That's a different type of analysis. It's not what we do here. But you're right that it's very important to take that into account as well. Okay. You presented a projected simulation for BM, which is a model you depend on. You depend on the only one model, chemistry climate model or multi-model, and for which driving force scenario used in the next 10 years? So the model that we use here is always the same, the Chimer model, the chemistry transport model. And so that's a French design domain that is operating on either France or Europe. So it is the only model we use. And for the drivers, I was explaining that we have different drivers in terms of emission projections. So we designed that ourselves on the different scenarios that we want to look at. And the other drivers in terms of global boundary conditions, the influx of air pollution can be obtained from a global circulation model. Or you also need to take into account the driver in terms of meteorology. If we look at 2020 or 2030, we consider that the meteorology is not changing. So we use the same as today. But if we look at 2050 or even beyond, we take a climate model to use this altered climate condition that we expect to have by the end of the century. I have a question. I have a follow up to your to your sulfur question. I was asking you, but I'm asking to everybody in terms of measures that I listed. So I had this hybrid cars measures. I had you were suggesting to have this sulfur content measure, your own arms. Can you think of other types of measures that we can take? What can we do just for the example of road transportation? Okay, I will rephrase. I had in my mitigation plan, I had several measures for air quality. If I just focus on road transportation, I had different norms, Euro 6, Euro 5. I had hybrid cars. You were suggesting to have sulfur content to decrease. You could have taxes on fuels. Can you think of another type of measure that we should explore for road transport? I don't know if it is a major as a major man or major as something. An action. An action. If it's an action, I don't know that I think if it is a road transport, when there are much pollution where they are fewer or less velocity of reduce the speed. When the speed is reduced, I think the pollution is most of the time high. It depends on the car. It depends on the speed as well. But this is, you're right. It is one of the measures we are exploring. It depends on the regime where you are, the type of roads and so on. But reducing the velocity, indeed, is one measure. Any other suggestions? I was thinking also about, I mean, we are talking about transport, road transport. Also the more individual measures that people can take. That's like, I don't know, commute by bus instead of take a car or by bike or by walking. But in this case, there is, it's not so, I guess it's not so easy to implement it in the, like in a measure database, right? Because you are not acting on the, it's not like an end of pipe measures. But these are some, this is something that probably, at least at a regional scale, local scale, we should consider. And no, that's all. Okay. Thanks, but it's very good. I'll take a few other answers. Hi. Okay. Thank you. Maybe we can divide the commuting zones. Say we can restrict the vehicle to enter in the residential area. A person should work down from certain point to his own residence. So that may help to reduce the pollution one end. And second, the physical exercise may help the good health. So that may be, and the metromovic electric train or tram, that may go around the city, lane by narrow lane. So that may be another option to reduce the pollution. Thank you so much. Thank you. My suggestion, which, which I apply, and which I find very efficient is working at home. Okay. Okay. Yeah. Thank you for the presentation. Maybe the electric cars would be a very environmental friendly solution. But I have, I remember now a small debate that arise just now. Why do the general motors kill the electric cars like 15 or 18 years ago? Maybe no one is talking about this issue. We would have saved the planet, maybe a lot of pollution since they invented the electric cars. Maybe you can say some word about this. I don't know the issue at all. Just to reply for the electric cars issues, actually they are not going to save no one because the PM coming from the traffic affecting the atmosphere, so causing air pollutions, is due mainly to combustion process, of course. So when the car pass, the PM will be emitted. But now we know that a lot of PM is caused by the tires rolling because the road are not clear, are not flat and clear, there are holes. And so there is a lot of big part of the contribution comes just for the car passing through the street. So the only solution is to cut the emissions, to reduce the number of cars, not to change in other kind of models. For example, in, in Italy, at least in the western country, they shift from small, medium-sized cars to SUV, I don't know what's this, big truck, just one person sitting in. They are big, so the weight on the street is even bigger and worse. And the toxicity of that PM is not safe for our health. We have recently published a paper from PM coming not from combustion process of the car, but coming from other sources. And that PM is very dangerous for your health in terms of cardio or respiratory morbidity as usual, we are talking about particular matters. So there's no solution in changing that just the only solution would be to find alternative ways of mobility. So implement public transportation, implement bicycle, implement other when possible. And if you can, as I'm one of the few persons of a medical background in this room, try to work, because if you work four kilometers a day, it's very good for your health and you can save the planet and yourself. Okay, thank you. I think that's a quick addition to that. And what we are trying to do in Ghana within some localities is to try and pull cars. So like Isabella said, if maybe you know that by the weekend, during the weekends, there are lots of activities, for example. So if during the weekend you know that there's somebody in this family or this locality who will be traveling to maybe a village for an activity, you try and join them so that two people or two different individuals won't have to drive two cars to almost the same place. Then another measure has to do with slapping huge taxes on importing overused or overaged cars, because we know that the older the cars are, the less effective they are in the combustion and as a result, you have a lot of emissions. So what we are trying to do in Ghana now is to try and say if you are getting in a car that is maybe younger, then the tax you pay on it or the duty you pay on it is lower so that we motivate people to get in newer cars than older cars. Thank you. Thank you for, okay, just maybe one last contribution and then we need to move to the next part. Yes, please concerning the idea about the traffic also, replying one final point about the cars, we don't, we have to find another solution. Yes, you are right, but concerning the elimination of the fuel consumption using the electric cars, we are eliminating a highly contributed source of pollution. We can use the electric cars, yes, we have the tires and the asphalt polluting gas, but at the same time, we are getting rid of the pollution coming from the oil or from the combustion of the petrol ingredients. Thank you. I'm sorry, I want to keep some time for the use case, but I say it was the last one, sorry. There will be a general discussion, there will be a general discussion, but before I forget, just follow up to what you just said, it's true for cars you can do some things for an individual, for industry you can also do something as an individual in your vote because those are bear by policy indeed, I mean it's being taken into account in a higher level, but policy is influenced by the opinion of the people, so this is what you can, just a short follow up to that. The reason why I was asking my question about road transport is in my list, I only had technical measures and that in some of your answer, I could feel that you had this idea to have also non-technical measures, those individual measures that you mentioned, and that is very important, but less easy to quantify. So if I got the list that you just suggested, for example, reducing the speed remains a technical measure, you can assess that quantitatively. You can add congestion charging that is similar to taxes, you can handle that as a technical measure. Increasing taxes on imported cars, also you can assess quantitatively, but then the problem of individual measures like reinforcing the use of carpooling, cycling, working, staying at home, what else, reinforcing the public transport is more difficult to fit into this type of model. So this is more the type of measures you know they are going to be beneficial. Many of them are what we call the non-regret type of measures, also because you know that if you work you have benefit for your health and so on. So these measures are supported anyway, but they are not part of this type of very quantitative assessment when you rank the measures, because it's more difficult to be quantitative for this measure. So there is also work involving people from social sciences to know how we can assess the cost of these behavioral changes, but it's still more, I mean it's not at the same level of quantity to the different measures I was listing. I'm not saying there is nothing, I'm saying it's not in the analysis I was showing you. Okay, I move to a different topic, looking at more short time scales and also I want you to engage with a tool that we design and to try to play yourself with these different policies. So I discussed a little bit of what we do, but like looking at the middle of the century or 10 years have now, now we'll see what we can do for tomorrow, literally. So these models that I've been using before, I'm showing you for, they're used for assessment, you do a one-year simulation, they also do to do air quality forecasts. You forecast the next couple of days and you see what happens. A big strong activity in Europe today is what is being done in the Copernicus program of the Commission. So it's, and in particular the Copernicus Atmospheric Monitoring Service. I encourage you to go to that website and you will see that there are very many tools that are available in terms of computing the emission fluxes of air pollutants, doing forecasts and also hind casts for global air quality. You have different products in terms of UV index, impact on ozone layer, but the one that is of most relevant for me here is the European air quality forecasts that are being produced every day on the web. You have the air quality for the next two or three days. You have the maps as pictures that you can download, but you also have a lot of access to the actual data which you can play with and that is available also throughout the world. So this is really a very strong project of the Commission now and it's bringing a lot of strength to the whole assessment process. These air quality forecasts that you have, you can use them for different prospects and maybe the first one that has been used for quite some time is to warn the population. And then when you know that there is an episode you would issue a warning to the people who are sensitive to air pollution, the elderly, the people who have illness conditions or the youth, a different type of population that can be targeted and you just do a warning. Then those people may prevent from having an activity outdoors and not be exposed to air pollution. This is the first type of action that you can take. This warning can have also an impact in terms of raising awareness. You never know what will be the benefit, but it's always good to talk a bit about air quality from the day to day. We know that the main problem of air quality is over the long term, but it's good when there is an air pollution episode to communicate about that so that people remember that air quality is a problem. This is mainly a communication issue, but communication is important. The third type of action that you can take is to have an emergency mitigation measure. You forecast an air pollution episode and you say, okay, I have an air pollution episode, I need to do something now. For example, you would enforce like a restricted circulation for a given area. You say, okay, I have an air pollution episode, I don't want any diesel cars to enter in my city for the next couple of days. So this is emergency measures. Forecast can be used for these different activities, but mitigation measure is very difficult to actually design because you never know. I mean, you never know. It's difficult to know what will be the benefit of this measure and is it really useful to target more residential sector or industry or road traffic when you have an air pollution episode. So in order to inform that, we have been working for Copernicus to design a number of tools that are available on the web and can be used by decision makers to assess the benefit that they can expect from the mitigation measure they shall enforce during an air pollution episode. One of them is an interactive scenario analysis tool that we call the CAMS air control toolbox. You will go online in a minute and you will see that tool as it stands today. You enter a date and a pollutant in the menu and then you have the map pollution for that pollutant for the given day. And you have a number of scrolling bars on the left hand side that you can play with. And with scrolling bars, you can assess what would be the benefit for that day of reducing, for example here, traffic by 70%, 75%. I don't do anything for industry. I cut residential emissions by 50% and I cut agricultural emissions by 25%. And in one go, the web tool will tell you what is the benefit that you can expect for that one day. I think I have to know only telling you that everything is complex and everything is simple with an online tool. So I'm not going to go a lot in the detail, but just want to mention how that works. It works with a CTM basically except, so the shimmer model now should be famous, except that we have been designing a surrogate version of shimmer, simplified kind of shimmer light version. So every day, we run the model for a three day forecast for all the various compounds that are of interest, those on PM and so on. It's a medium resolution, like 25 kilometers. And then we run this model with a number of scenarios, training scenarios, where we do different hypotheses in curbing, for example, agricultural emission by 60% or 100% industrial emission, curbing residential heating emissions and traffic. So you have different combinations of your scenarios that you run, about 10 scenarios in total. And then with these 10 scenarios, you fit your surrogate model with a non-linear combination of all these training scenarios. And this is the formula that we use actually. And this is a polynomial. It's very fast. I mean, computing a polynomial is immediate. And you can upload that to a web tool that will be your surrogate of the shimmer model. So running shimmer is still very long. We do that with a high performance computing. I mean, it's not so easy to run like 10 different scenarios every day for the forecast. We have large performance computing, so we can do that. And afterwards, once you've done all this analysis every day, again, every day you update a surrogate model that you upload on the web. And that surrogate model can be used on the interface. And I will have you play with this tool in a minute so that you see how you can assess the sensitivity to the various air pollutants for the various emission sectors on air pollution. I don't know if I will enter into that for the sake of time. I will skip it. Yes. So now the case to this. I would like to have you to have a look at the contribution during air pollution episode of activity sectors with the tool I just mentioned. And if we have time at the importance of local and non-local sources, not sure we will have time. So for the first one, what you should do is go to that web link. Maybe we set up the groups now. Now we are going to set up a few groups or four or five people per group. We have a list that Karla prepared. So you will see your names and you should go in different parts of the room clusterized in a way. And ideally so you would need to have at least one laptop per group to go on that website. And if you don't have one laptop per group, then you should go to the computer room facility next door. I mean it's easy. If you find yourself in that list, according to the color that we have six groups, I try to mix a little bit. So if your name is not in the list, say it and we can join us all. Stand up and move in different areas. So we spent a lot of time for the discussion. It was very nice, but now we need to speed up a little bit if we want to have time for the case today. Can you see the list that I used? But several people should see their names. No one is in that list? Maurizio, I'm sure that you are. Ah, okay. So those are together. Yeah, sure. Okay, so yellow and green, violet and orange and gray and orange. So let's say the side. Yellow and green, please. Where are the yellow? Where are the yellow, please? Yellow? Are you yellow? Yellow group? Yellow group? Yellow, yes. Oh, these are yellows? Course is finished that you don't know. You buy names. Okay. Okay, are you ready? Can we start the exercise? Yes. Okay, we need to save time. If you have a problem to find your group, you go to the near house group and we start. Okay? No, it's not your fault. Okay, do you want us to decide? So let's go and I class them. Should I do something to put it there? Okay. So now, please go to that website. It's okay. You are on the website already? Is that with the fiction? Okay, no big deal. So now everybody should have a group. Where is this? And try to connect. So I need to have your attention a little bit. So can you please connect to that website and wave when you're ready? No big deal. Forget about the groups, really. No, no, no, it's okay. We forget about the groups we need to move on. So please go on the web. Okay, you are on the web link. Okay. So if you are able to see the same map as what I'm showing here today. So this is the map of PM 10 for the first of December 2016. Okay. This is the actual forecast we have with the chemistry transport model. Then we are going to play with the various sectors. But the first thing we need to do is to withdraw the natural contribution. On that plot, you see that, for example, you have some dust. Dust is not an activity sector. It's a natural contribution. So there is a tab here that says access to the raw forecast excluding natural source. So please go to that one. And you should see the map that I'm displaying here. What you will need to do is to play with the color scale. So here, the value I've used is from zero to 36. It updates the gradient of color that you have here. And then you should see really the same map. If it works. Okay. Once you've here, you go to the next tab here. You go to user scenario. So user scenario will give you the forecast of PM 10 for that day with the level of reductions that you have set. And then if you start to play with the scrolling bars, you will see the map changing. So what I would like you to do is to select different levels of reduction of air pollutants. You can just go between zero and directly to 100 percent for the various sectors. And I would like you to tell me for that day which activity sector is the most efficient to target. So what you do again is you select the date and the pollutants you did before. First of December 2016, PM 10. You play with the color bar. You go to user scenario and the various levels of air pollutants. You go one by one. You reduce traffic. You look at your map. Yes. So if you are in a group, can you take just one PC for the group? Yes. Okay. We'll solve the yeah. I've never heard of that. But basically I gave you the I gave you the direction for the assignment. Here I live on the screen. Just one more minute of attention, please. Just one minute of attention again. I will leave the assignment on the screen and then I will go in the room to help you. But just to summarize, what you need to do for the contribution of activity sector, you do this first part. That would be for the second part of the assignment if we have time. Okay. So you just look at this first part. And then what you should look at is activity sectors for these dates. I started with the first December 2016 for PM 10. And then you can go to this other date, 18 March 2015 for PM 10 and 21 June 2017 for Ozone. So you can play with these three case to these. But for each of them, you should do the same thing always. Go to the web link, user scenario, color bar. You change the date and the pollutant and you play with the reduction. And what you need to tell me is what is the key activity sector to mitigate our pollution for that day. I will go in the room and try to help you now. How many errors in the screen? Just you solved? Okay. So if they maybe take some times to, okay. Great. I know. So may I have your attention now? I will show you a little bit the result that maybe you could see, maybe not. I noticed there was a number of technical problems. So what I will do is to show you the type of analysis that you might have done or you could have done. And I would suggest that if you have time later on, you try again on your own. I think a lot of the problems we had at the moment were technical problems due to the fact that there are too many connections at the same time. So I suggest that you connect later on your own to try to see how it works on your side. But just briefly, if I take the example of the first case today I wanted to look at PM 10 on the 1st of December 2016. The reference simulation is this one. We have an air pollution episode over the big part of Europe, France, the southern UK, the Po Valley in Italy always, and also Eastern part of Europe. Okay. Here, this is the reference simulation. So if you go to the tool and you select the user scenario, but with all the menus set to zero, you don't reduce any of the sectors. This is what you should see. And you also need to play a little bit with the color bar. And then you play again with that map, but reducing the various activity sectors. So what I have on these screenshots that I did yesterday are to remove traffic by 100% and I don't change the other one. Here, I remove industry by 100% and I don't change the other one, remove agriculture, and I don't change the other one, remove residential, and I don't change the other one. Okay. And what you will see if you do mentally the difference between the different, the reference and each of these scenarios, you can see what is the expected impact of your various activity sectors. So for example, if I remove all the agriculture emission, I can remove a large part of the large scale pollution, but I don't remove the hotspots like Paris, the Po Valley, and Eastern Europe. If you play with the industrial sector, it has very little impact on Western Europe, but the hotspot that you had here in Ukraine disappeared. So if you are a stakeholder from that part, you say, okay, maybe it's more efficient for me to play in the industry or the other sector. For that day, you also had a large contribution of the residential sector, and traffic is clearly a main contribution everywhere. Okay. Another case today, when you take a reference for March 18th of March, 2015, here we have a huge episode of a big part of Europe, and clearly it's only the agricultural sector that is efficient. You can do whatever you want with the other ones. It's not helping a lot to curb your pollution. Right. And the last example was for ozone. Ozone, a reference in June this time, 21st of June, 2017. And here, the most efficient sectors are traffic and industry. You have little residential heating emissions in June anyway, and agriculture has a very little impact on ozone. So with this tool, you can assess really what is the main sector if you go from zero to 100%. And if you are a more reasonable policymaker, you will see, okay, if I expect that I can, tomorrow, reduce the traffic emission by maybe 20%, 30%, you can also design your own emission scenario and have an immediate answer, assuming that the tool works. You can also, do I have, yeah, I can take questions at this stage, and then I will, yeah, if you want to play with a scenario, you need to be, there are several tabs in the top. And the first tab was the raw forecast. And then you have the raw forecast without the natural sources. And then you need to be on the tab that says user scenario. This is the, in the top menu, you have to go to the user scenario. So if that was not clear, so you will have these slides afterwards. You will be able to go to the web link and you follow the directions. And I hope that it will be more successful that the technical problems we had so far. Any question? We don't have a question, but a lot of expectation, you know. So basically, from here, we are from Africa and Asia. And this model does not work for African Asia. So it helps us to develop such model for calculating the pollution level in our country. What I can do is the methodology is going to be published in paper. So you can replicate the methodology with your own area. But of course, it would be interesting to support the different countries to develop such tools if you think it can be relevant to your purpose. Thank you so much. I think that the results depends mainly on the inventory emissions we use. On the inventory emissions. So if there are some studies on the sensitivity study, there are sensitivity studies about the inventory emissions, if it's quite good or not. Okay. Of course, the model is a model. It's sensitive to what you put in, the methodology, the emissions. So yeah, that is clearly a point to be taken into account. I have a suggestion. Is it possible we did six groups? Is it possible to divide the world in six parts and do the same work in a real way? I think it is a useful tool, I hope, and I would be very happy to develop that elsewhere, but it's a matter of resources. Okay, I will move. If you have more questions, maybe, yes? What's the range of the resolution we can use? What's the resolution? A range of resolution for all. The model is validating between 0 and 100%. That's what I had that slide before. I didn't introduce that, but we don't fit a linear model. We have a nonlinear polynomial. Here you look at the emission reduction ratio between 0 to 100%, and the model is fitted for the whole range as a polynomial. So we know that it remains valid between 0 and 100%. This is the first view. This is what you have as a user. You can play with these maps, but actually, when you have the tool, the surrogate model, you can also use it as a source apportionment tool. This is what I had here. Basically, you look for different days. Again, the three episodes, March 2015, December 2016, and June 2017, when you have different days going forward. Here we do an attribution of the various sectors. We use the same type of approach, except that here we plot that as bar plots for Paris. When we have the contribution of the various activity sector, so agriculture in green, industry in red, residential heating in blue, and traffic in black, and then there is residual and other fractions. But the important message I want to convey here is that, for example, for this March 2015 episode, you were looking at the 18th, which is that day here. Clearly, you see that agriculture is a big contributor for that day. Also, traffic is not bad. But if you look, for example, at December 1st, 2016, which is here, agriculture is much lower and you have a much larger impact of residential heating and traffic. So this is how you do your attribution. If you go later in the months, even 2016, you see that you have different typologies every day. And for ozone, you have a large part of the background, which is natural. And then, really, agriculture and residential heating have no impact. The only two sectors that you see coming out are traffic and industry. Depending on the day or the location you look at, the importance will be different. I just want to close with another tool that we did not have time to look at. But you have the link here. CAMHS has produced also another tool, which is a source allocation that will give you the relative importance of the city and non-city emissions to an air pollution episode. So if you go to that website, the second one, contribution of local, non-local source, you will see similar plots like this. And this is an extraction of the tool for Paris in December 2016, the same TM episode I was showing before. And here, basically, you have the whole suit from the December 1st to December 8th. And the share of the value sources is given here. So in black, gray, you have the contribution of Paris. And in blue, you have the rest of France and the rest of Europe. Other is also like a natural fraction. And what you can see from this plot is that, depending on the day, you have a very different contribution of the local sources and the non-local sources. And basically, a pattern that we see very often in similar air pollution episodes is that the beginning of the episode, you have an importance of local sources, half more or less half of the episode. And then it goes down. Here you have the beginning of another episode. And again, it goes down. After a few days, when the episode is well set up, you have a contribution of regional pollution. I will take a question at the end. I just want to finish with one thing. So with this analysis, what I would like you to take home as a message is that there are various things that contribute to air pollution episodes. I hope that these kind of tools are useful to understand how episodes behave. But what is important at the end is that there are different activity sectors. Depending on the day, it's not the same activity sectors, but more or less on average, they all contribute. This is maybe the most important message. There is no magic in doing air pollution mitigation. You always have to target all the different activity sectors together. And also, you should not target only point sources. When you look at this plot, you say, okay, at the beginning of the episode, maybe it's efficient to do something in Paris for Paris. But after a few days, the air is being addicted and the air pollution from Paris is not staying in Paris. So you have to act at the global level. And that's why I want to finish with this plot. Again, it's the same episode, December 2016, where we have the situation that did happen on the left-hand side and the situation that we expect will happen in 2030. So we played again the same meteorological situation except that we used the targets in emissions that we set ourselves for 2030. And what we realized is that if we match the targets we have, the air pollution episode would be much smaller. Basically, it would be really localized only in Paris. It's not like a large air pollution episode anymore. But that 2030 target is far from being only for traffic or only for residential. It targets everything. You have to go to all the sectors and not only in one place. You have also to go to like European scale measures. That's the only way it's going to work. That's what we promote and we promote it because we have evidence with the modeling we are doing that it's going to work. So I'm finished. I have again a nice movie. I take the opportunity to add the link to the YouTube video so that you can have a look afterwards. It's in English as well. And happy to take further questions if we have time. Thank you for the presentation. I have just one question about your last tool. If it's possible to have an idea, not to calculate, but have an idea of the European sources, European contribution, I mean the local or the local, but the European to have an idea of the radius of influence. So what you have, what I was showing here is a result for December 2016. The product has evolved since then. So if you go on the website now, you will see that you have slightly different products. If you go like a year and a half ago, it's quite basic. Now if you go to air pollution episodes that are a bit more recent, or if you go to the forecast of today, you will see that you have the contribution of local, rest of the country, and then all the other countries. Basically what they say is the eight or 10 larger contributor in the neighboring countries. So from that, you don't have really a radius of influence, but you will see in, I take Rome for example, and in Rome, you will see the first eight largest contributing countries. So you have that type of information. Just have a question from Simone on the internet. She's asking you if it's just all that valid for Europe. And what's about all the Mediterranean areas that are highly affected by climate change and pollution? I know the focus is Mediterranean areas, I know. But I'm sorry again, this tool is mainly developed for you at the moment. Okay. There is a question here. I have the mic. My colleague was asking about the data resolution. So how can I cover the land with the resolution? I mean I can take small places to original places, the scale that I can use. I can downscale the resolution to, I think, one kilometer. At the moment, the resolution is 25 kilometers. So it's covering the whole Europe at 25 kilometers. So it's not there to design a pollution for a very local city. It's really like for a European scale. Okay, thank you. We use CIMER. You can use CIMER up to one kilometer resolution. One, yes. That's okay. Thank you for the wonderful presentation. I would like to know that. What you showed us, it was very informative. But could you a little bit describe about more limitations about this model? Just like what I'm thinking that I'm not getting anywhere, the options there should be or maybe something about the latitude and longitude. Because this model is, I think, very, I think, confined. Just I want to know, is there some option is given for latitude and longitude so that you can also compare from another places, even in the Europe. So I think this is the limitations of this model. But I'm thinking, but I'm not sure. Just I need to learn. And another question I have, maybe a little bit different from your topic. But yes, what I saw in your some slides, there was a mention of the surface flux. Something was the surface flux. If you are going to describe or going to explain about the model or some climate change, so if you could you a little bit explain a little bit about what is the role of the surfactants in the air pollution and the sources, because surfactants, which also play an important role in the air pollution, even in the formation of the CCN and cloud. So all these things are, I think, need to be added if you are going to describe this type of broad climate change and what I'm thinking that. I am not sure I understand your question. So my main question is the what are the limitations of this tool? Limitation of the tool. So the tool we are drafting a paper on to explain what are the uncertainties. It's actually, I'm not very impartial, but it's actually very good, the uncertainty of the order of one percent. So it matches very well the model. I'm talking about a surrogate model that fits a model, but then the model itself maybe won't. It's another problem. So I am designing a surrogate and the surrogate is actually very efficient, but the model itself has uncertainties as well. So if you want to do a validation, we can do a point-to-point validation. I took note the fact you want to have like the geographical coordinates to do comparison. But on the other point, I'm afraid I did not. Okay, that was a little about the surfactants, role of surfactants in climate change. Just I saw the some... The role of what? Surfactants. Surfactants, that reduces the surface tension. Because you are going to, you showed a little bit about the surface flux. That's why I remembered maybe something, role of the surfactants, that I'm curious to know. The reflectance of the surface? Surfactants. I'm sorry. I don't. Surfactants. I don't know what that is. I'm sorry. Okay. Thank you. Thank you. Thank you. And does this tool include other pollutants such as PM2.5 or? So the model actually exists already for NO2 and for PM2.5. It's just that it's less validated. The priority was to do PM10 and Ozone, because it's the main pollutants we look at in air quality forecast. For a health impact assessment, you look at longer term. But when you do the forecast, the warnings that we issue for the next couple of days in Europe, it's mainly designed for PM10. So that was the priority. But it's, of course, it's part of the development plans to add PM2.5. Yeah, thank you. Well, this question is not maybe directly related, but I've been meaning to ask it. So I'll ask it. You know, with this, a case study we did, we tried to look at the impacts of the various sector activities on the pollution. Now, my other question is, assuming we're trying to do a similar scenario with met conditions, meteorological conditions, like wind speed, direction, temperature, precipitation, which of them, in your experience, really has the most impact on desperation? That's not an easy topic. There is a lot of research on that field, especially when you look at interaction between climate change and air pollution. You look at the future climate and you consider, okay, it will be warmer, for example. It's expected to warm. What is the impact of heat on air pollution? But then climate change is not only heat. When you change the climate, you will change also the weather patterns and precipitation and so on. And that also is going to have an impact. So there is a lot of work that has been done in the climate air quality community to try to isolate the factors that are more important. And it depends a lot on the area of the world. It depends also the type of time scales you look at. It depends also on the pollutants you look at. What I can tell you about Europe, because it's the location we've been looking at. For ozone, clearly, the main factor is heat, heat and surface radiation, which increase in the future and you increase ozone. It's actually quite simple for ozone. For particulate matter, it's much more complex, because particulate matter is made up of different compounds. First one is dust. Dust will increase in the future because of surface dryness. So you dry, you have more desert surfaces, you have more dust. So this increases. For the other aerosol is not so clear. Some people will tell you it increases, some people will tell you it decreases. So you have climate penalty, climate benefit, depending on the aerosol you look at, okay. And the two factors that will play here are temperature and precipitation. Precipitation will increase the scavenging of aerosol. Problem is, projection of climate precipitation are very uncertain. And the impact of precipitation on PN is also very uncertain. It's quite difficult to model. And heat has a two-fold type of impact. For aerosol, if you increase heat, you increase the volatility. You have a large part of your PM that is going to go back to the gas phase. So you reduce PM. But at the same time, when you increase heat and incoming radiation, you can also increase the photochemistry and the creation of organic aerosol, precursors and so on. So you have a lot of competing effects. And at the moment, really it's really difficult to conclude firmly if it's one direction or the other. There is a lot of debate, different studies. It's not conclusive. It's a complex issue. I'm afraid I'm finished, but Konstantinos, if it's a quick question or... I think it's time to have a break, okay. So 10 minutes break and then we come back at 11 sharp to continue with Dr. Olivier Sanel.