 The main question, the main thing that we are trying to answer is to figure out when the Arctic sea ice might be gone. So Arctic sea ice, that is this ice that is floating on the ocean in winter when the ocean cools down ice crystals form and they then freeze together to form one meter, two meter thick pack ice. And this ice has been disappearing really rapidly in recent decades. So if we just look at the aerial extent, it used to be such that the entire Arctic ocean, this ocean that covers the North Pole, used to be covered in sea ice even in summer, but that area has shrunk by about 50%. And also the thickness has roughly shrunk by 50%. So the ice that used to be maybe three meters thick now only is one meter and 50 centimeters thick. And so if we take these two numbers together, the shrinking in aerial extent and the shrinking in thickness, both have shrunk by about 50%. It turns out that over the past 20 or 30 years, we've lost about three quarters of Arctic sea ice in summer. And so one last quarter is left. And what we are trying to understand is when this last quarter of Arctic sea ice might be gone. The method that we are using to figure out why the Arctic sea ice is disappearing so rapidly at the moment and what the future might bring is based on the combination of satellite observations and model simulations. And the unique idea about our approach is that we are trying to identify the underlying key processes that give rise to the ice loss, which then allows us to use insights from the models and insights from the observation record and combine them because through this understanding we get a much more robust prediction than we could have had before where we simply had to trust the simulations or where we had to extrapolate the satellite record. If we look at the observational record, we do have pretty reliable data going back to 1953. For the first two decades, roughly from 1953 till the late 1970s, we do have reliable ship observations. We do have reliable airplane observations because in the Cold War there was a lot of coverage of how much ice there was in the Arctic from these two sources. And then from the 1970s onwards, we do have the satellite record with satellites flying over the Arctic once a day or every other day and giving us a complete coverage of how much ice there is. And by combining these records, we do have a pretty good observational record going back about 60 years of how the ice was changing from year to year and how it was then going down over the past few decades. In terms of modeling, we use large scale climate models. These are very complex codes, very complex programs that we use to simulate the entire climate system of our planet including trees and waves and clouds and sea ice. And from these simulations, we do get a series of evolution of simulated sea ice starting in 1850 when we start these simulations and then going back all the time until today and then going about 100 or 200 or 300 years into the future. And we've combined these observational records with the models in order to find the key underlying processes that give rise to the loss of Arctic sea ice. In addition to these records of sea ice from the observations and from the simulations we also use data on how the global temperature has evolved and how much CO2 we humans have emitted. And it turns out at the end of the day that the main variable that we need to understand and to set into relation to the sea ice loss really is this emission of CO2 from humans. So the main finding was actually something that occurred pretty early in this research because the first thing of course we did when we tried to figure out why the Arctic sea ice is disappearing and we had these observational and the models was to simply plot how much sea ice is there in the summer and how much CO2 have humans emitted in total since 1850. And if we do that, we get a very, very clear linear line, linear relationship between the amount of CO2 that humans have emitted and the amount of sea ice there is in the Arctic. And we got that in the observation record and we got it for every single model simulation that we looked at. So there must be some underlying reason for why these two are so very clearly linearly related. And the great thing about this linear relationship is that it allows us two things. So first of all, it allows us to figure out how much we individually contribute to the ongoing sea ice loss in the Arctic. Because the slope of this line is a very simple relationship between CO2 emissions and sea ice loss and it gives us about a three square meters of Arctic sea ice loss for every ton of CO2 emissions. And that for the first time really allows us to understand how we as individuals contribute to the ongoing climate change on our planet. Because one ton of CO2 emissions, that's about driving a car for 6,000 kilometers or it's flying from Europe to the east coast of the US for example. It's not a crazy number. It's something that we do in our everyday lives basically. And so for each of these tons of CO2 that we made about three square meters of Arctic sea ice disappear. So that's the one thing we can do. And the second thing that we can do is we can extrapolate that relationship because we see in all the model simulations that this linear relationship holds until all ice is gone. And in extrapolating that relationship we actually figure out that there isn't a magic year when all Arctic sea ice is gone but there is a magic number of CO2 emissions when Arctic sea ice is gone. And so it really depends on the future CO2 emissions when the Arctic sea ice might be gone. And so if we extrapolate that relationship it turns out that we can emit another 1,000 gigatons of CO2 and then Arctic sea ice will be gone in summer. And then it depends on how much CO2 we emit. At the moment we emit about 40 gigatons of CO2 every single year. So for current CO2 emissions we will have another 25 years and then Arctic sea ice is gone in summer. If we were to emit less and less and less in the future it takes longer and longer and longer for the sea ice to disappear. And if we manage to keep our total CO2 emissions below this magic number of 1,000 gigatons of CO2 we might be able to still rescue at least some of Arctic summer sea ice for a very long future. What is the main relevance of this finding? This finding that we do have a very clear relationship between how much CO2 humans are emitting and how much sea ice we have in the Arctic. Now there are three aspects to that. The first one really is that for the first time we are able to visualize the fact that it's us individually that contribute to the ongoing global warming. It's not something that comes out of the blue but it's our individual actions that make the Arctic summer sea ice disappear. We for the first time have been able to come up with numbers that are so intuitive that everybody can understand them. The second important thing is that for the first time we've been able to make a really reliable prediction of how the Arctic sea ice will develop in the future because we simply find that this linear relationship that we get from the observation record we can just extrapolate that into the future. So we don't have to use very very complex climate model simulations which do have uncertainties when it comes to Arctic sea ice but we can make this directly from the observation record and so we get a much more reliable extrapolation of Arctic sea ice evolution in the future compared to what we had before. And the third one is that because of this much more reliable extrapolation of when Arctic sea ice might be gone we can inform society much better on how the individual actions will for example affect Arctic sea ice how for example the Paris climate treaty might affect the future evolution of Arctic sea ice and people can plan much better. So for example if you run a shipping company you might be interested in when the Arctic sea ice might be gone because you would like to use your ships and go over the North Pole in order to make trade between China and the US or China and Europe or wherever. For the people who live in the Arctic it's important to know is there any chance of rescuing this sea ice? For biologists it's important to know what needs to be done in order to have sea ice all year round for a very long future? How will it affect animals? And so all these myriads of aspects that are related to the existence of Arctic sea ice we've now been able to put very concrete numbers on. Our study now for the first time has given us very clear guidance as to what we need to do in order to preserve this very, very beautiful landscape also for future generations. So when it comes to what happens next there are two aspects. The first one is to really better understand our findings and the second one is to possibly apply it to a broader range of questions related to sea ice. And so when it comes to better understanding our findings the fact that there is this clear relationship between sea or two and Arctic sea ice coverage that is clear. This linear relationship we just see everywhere. We see it in the observation record. We see it in the models. So that is something that is robust. Also the slope of this line is quite robust. Three square meters sea ice are lost for every ton of sea or two emissions. What is less robust possibly is the explanation that we give in our study as to why there should be this linear relationship. Because in order to explain that linear relationship we examine the energy balance at this edge of Arctic sea ice. So where the ocean and the sea ice meet. We looked at all the individual fluxes that are there. The incoming fluxes from the clouds, from the atmosphere, from the sun. We suggest that it really is related to the warm, the heat radiation that comes from the atmosphere that gives rise to this linear relationship. That is a hypothesis that we formulated in the paper but we haven't been able to test that hypothesis against observation records because we don't have measurements of these fluxes. We don't have measurements of the changes of these fluxes as the climate gets warmer. And so in order to test why we do have this linear relationship and why this slope is three square meters Arctic sea ice lost per ton of sea or two emissions we do need more observational campaigns and we hope that our study allows researchers now to make these measurements in the Arctic and thus to better understand why we do get the relationship that we find. That is one of the main follow-up things that we hope will happen after our study. The second one relates to the method that we've employed. Namely, simply trying to find robust relationships in the observational record and then using that combined with extrapolations from model simulations to better project the future. We've now only done that for sea ice in September in the Arctic. So September is synonymous with summer sea ice. It's the end of the summer sea ice at its lowest. But what happens another month? What happens in winter where the Arctic sea ice is retreating as well? Can we project when Arctic winter sea ice might be gone? We hope that that will be not even happening within this century but possibly next century depending on future CO2 emissions. Can we make more robust projections about that? Can we make robust projections about how long shipping periods might be in summer, for example? Can we make projections as to what will happen, for example, to mid-latitude weather systems when the sea ice is retreating more rapidly than the models are simulating? And so using our approach to get a more robust understanding of the future evolution of Arctic sea ice and all the research questions related to that is the second main group of follow-up questions that could emanate from our study.