 So, welcome everyone to our third press conference of this year's virtual EGU, VEGU 21. This press conference is titled New Geoscience Tools, Novel Applications. And this year's EGU, we have more than 14,000 abstracts and 16,000 people from across the globe participating in the meeting. So I'm Erin Martin-Jones, this year's EGU press conference assistant, and I'll be hosting today's press conference, which will include a question and answer session, following on from presentations by our three speakers. So to allow members of the media to ask your own questions, we're conducting this as a Zoom meeting. And once the last speaker's finished, please write the letter Q in the chat box to ask a question, and I'll call on you in the order that the questions are asked to come back to you. And of course you're welcome to write your question out fully, and I can also read that out from the chat. And hopefully this doesn't happen, but if for some reason Zoom should suddenly quit, we'll just restart the press conference and we'll give everyone a couple of minutes to rejoin the session. And likewise, if you have troubles at home with your own internet, that's fine, come back through and someone will be around to let you back in. So there's lots of information on our online press centre, that's media.edu.eu, so you'll find the abstracts and other documents relating to the press conferences here. So please check there for more information. And I will introduce our three panellists now to make for faster transitions in between them. So as I said, this is new Geoscience Tools Novel Applications, press conference three. And first up we have Dr Stephanie Yipma, who is postdoctoral researcher at Utrecht University in the Netherlands. Hi Stephanie. Next up we have Dr Eunice Issel, who is a doctoral candidate at the Christian Albrecht University in Germany. And last but not least, we have Veronica Tolena, who is doctoral candidate at the University Libre du Brussels in Belgium. So I'd now like to call on each of our panellists to provide us with a brief presentation and we'll start off in that order. So Dr Yipma first, if I could hand over to you and you could unmute yourself. Yes, thanks a lot for the introduction. I'll start sharing my screen. I hope you can see it now. Yeah, got it. Yeah. Well, thanks a lot for inviting me to this press conference to take part. Today I will be talking about the Project Galapagos Plastic Free. It's a huge initiative from the Galapagos Conservation Trust based in the UK and together with many other collaborators among which us at Utrecht University. So we are physical oceanographers, that means that we're really interested in how plastic floats around in the ocean, where it's coming from, and the focus of this talk, how it then arrives at these islands, because these islands are pretty special, one too much. When you think Galapagos, you might think pristine beaches or nice weather or maybe Charles Darwin, and you find these very unique creatures here that you don't find anywhere else. And there are many more reasons why these islands and also the marine reserve surrounding it by their UNESCO heritage site. So it's kind of this natural laboratory to study how species adapt to changing environments, which is quite relevant nowadays. And the reason for this unique biodiversity can be attributed in part to the ocean flow. So you have these three main ocean branches that meet near these islands, and they all bring these properties and together they kind of form this melting pot for life. So that makes these islands very pretty unique. But unfortunately, these ocean currents also transport quite a lot of plastic towards the islands. So they find about eight metric tons of plastic each year on the beaches, and people that organize these cleanups, they estimate that they have visited far less than 1% of the coastline of the Galapagos Islands. So this eight metric tons, it's really a huge underestimation and there will probably be a lot more plastic out there. They think that the main source of the plastic is not from the islands themselves, but mainly from the fisheries and also from the mainland. About 45% of the plastic used along the East Pacific coastline of South and Central America is actually mismanaged. And that number is only increasing. And unfortunately, with the COVID pandemic, it gets worse as well. So it's a pretty huge threat to the Galapagos Islands, both for the wildlife and also for tourism as well. So to tackle this problem, part of the project is focusing on the sources and trying to improve the race management. But a big part of the project is also focusing on the cleanup and how to make that as strategic as possible, because they just don't have the resources also financially to clean up the entire coastline every year or every day. So that's where we are coming in and we are trying to give them a tool that can tell them where to go to have maximum impact. So this is an animation where you can see locations where at that moment in time, a lot of plastic is arriving and our aim is then to predict this. So when and where is the plastic washing ashore and how can we then support really strategic cleanup measures. And the word strategic is quite important here. So we're not only want to make forecasts, but we also want to optimize this model for, for example, how accessible is the specific location? What's the impact there on the wildlife? And also, some of the plastic does not necessarily stay on the beach. So it can also return back into the ocean due to storms or high tides. So what's the chance that from there it will spread to other islands? So also those regions we want to target. The way we do this is kind of by this, this formula here. So we use a tool called ocean parcels. And with that tool, we can simulate how, how plastic floats around in the oceans. And also how it then arise at these islands. And that part, so how it moves from the ocean onto the land, there are still quite some uncertainties in that, that area of study. So we also execute observations to try to better understand these processes. And for that, we use drifters. So these are floating devices at the ocean surface. They have a GPS and they have a satellite communication. So we can really track them live, study how these, these objects move through the ocean and in particular, how they move from the ocean onto the land. So this field campaign is planned for the summer and we're super excited to see how much information we can get from that. And then we hope to combine these observations with the simulations, apply machine learning techniques and then develop forecasts. So this is going to be the tool that will tell the park and all the people there where they can best clean up. So this animation is an example of one of the simulations that we use. You can watch these, or I can at least watch at these for hours, but all these points are, are plastic, virtual pieces that float through the oceans. And there are two colors here. So the dark green color, those are the ones that we're actually interested in. So those are the ones that really get stuck on these islands and beach. And those are the ones that we want to forecast. So we don't want to run these kinds of simulations every time. We want to just get like easy velocity fields from that, hopefully understand what is likely going to happen. So this is kind of our realistic scenario and we want to mimic this using machine learning techniques. So one way of doing that is, is we have all these particles. We know where they arrive. We know what day they arrive. And then we can try to find properties that are really similar. And in this way, we can make clusters, for example. So that's applying a clustering technique. And every color here is, is a different color, a different cluster that's based on, on location. So for example, if we focus on the cluster, the green one, which is in the north of the island Isabella, we can then look at what kind of ocean properties are there in this moment of time that we have a lot of particles beaching. And can we then understand what kind of physical mechanism are responsible for these beaching events? And that's shown in the bottom plot, we show a difference of the mean velocity field with the velocity that we have when we have a lot of particles beaching at this location. So in, in the red color, you can see that means a stronger flow. So in this situation, we have a much stronger flow in the north of this island and a much weaker flow in the south. And from this, we can extract parameters. This is just one example, but you can do this for, for all kinds of different clusters. And in this way, you can get hundreds of parameters, and you can use all these parameters and apply machine learning. So there are lots of different techniques. I just shown two of them here. The left one is a linear regression and the white one is a random force. And what's plotted here is, is how well these models are working. So on the x-axis, it shows the model beaching. So that's, that's how many particles arrive in our simulation, like our reality, so to say. And on the y-axis is what our machine learning is predicting. And the more every dot is located on the y equals x-line, the better our forecast is working. So here you can already see that the right one, sort of random force regression is working a lot better. And it's actually really promising for us. So we're, at the moment, trying to improve this and using this, trying to improve our simulation, hopefully make it more realistic. Also we've, including the observations that we will get this summer. And also improve the parameters that we put into this machine learning tool. And hopefully, at the moment, we're really focusing on the Galapagos Islands. But of course, we hope that this regional scale model then can also support efforts of island nations and archipelagos worldwide to tackle the global challenge of plastic pollution and to hopefully make a difference. So thanks a lot for your attention and I'm happy to answer any questions. Thank you, Stephanie, for that really nice introduction. And so shall we move on next to Eunice, if you're able to share your slides? Of course, soon. Thank you. Yeah, we've got those. So thank you, Terry, for the invitation to present my topic about geophysical investigation of the medieval paintings at the St. Peter Cathedral in Schleswig with Gio Radar and Thermographie. Our team is represented through Erjan Erkwil, scientists who applied geophysics at the Kiel University, Dietle Fotekortner, engineer Christian Leonhardt, the restaurateur of our team and my working group leader, Professor Thomas Meyer. In the next six to eight minutes, I hopefully will show you why is it necessary to involve more geophysical, non-destructive methods in the restoration workflow due to analysed moisture content in walls of historical buildings without distracting them. But what is it? Yeah, I'm so sorry. What is the motivation behind my research? The old German proverb says innovation is only successful when the incoming water becomes less. Or just say it another way, only if you know where the moisture content comes from, you can do something about it. So incoming moisture has a strong effect of major injury due to the damage caused by it. So you can get a softening boil by water absorption or by physical, chemical erosion, micro-organism and mold growth, or as it is in Schleswig, damage due to site and gypsum crusts. So on the left side, you can see an example image of the so-called Joch with the medieval paintings here. I hope you can see the red lines and some damages which are in the dark areas and points in this area here. One of our first aim was to determine an effective solvent for reducing the gypsum crust by using a thermography to look how the solvents react with the crust over a couple of months. And the second aim of our investigation was to test the success of combination of geophysical measurements here, also thermography and georada for determining moisture content. But before I show you some results, let me quick show you where Schleswig is located. Here on the right part, you can see the northern part of Germany and Oritz-Wildt is the federal state Schleswig-Holstein. And if we zoom in, you can see Kiel, Martin Purple on the Kieler Fjord and Northeastern from Kiel is the town Schliessnicht. So if you get an impression how the church looks like here on the right side is the church. And if we go nearby, you can see the tower of the church just for impression. But we measured inside the church in the cloister. And this is how it looks like. It's a three-sided cloister and you now can see two of them left and right. And you see the medieval paintings in the dome. And I will show you some results from this side of the cloister. So what did we use for methods to get near more closely to the damages that they, which are there? So first we use thermography for the purpose of non-destructive moisture detection. We are looking at near-surface temperature variations and record them with the term an imaging camera. And the second method is georada. Electromagnetic waves are the same in the underground. And the analyzers of the travel times of the waves allow statements about deeper structures. So a subsequent analysis is to look at the attenuation. And the attenuation can possibly allow conclusions about moisture content. So let me show you a result of the thermography. On the left side, you can see again the joch. The dome shows clear the paintings, as I said before. And you can see in the joch, the intact areas on the left side and some damaged areas due to moisture on the right side. So during the investigation, the question arose as why the damage occurred in this area? And the answer is seen on the right side. You see an average, you see a picture of an average value of each temperature pixel recorded over time of one and a half hours. And it's just the mean, the temperature average of all one and a half hours. So blue colors indicates cool areas and red colors indicates warmer areas. So what do we see? There's a bit in a horizontal barrier. And you can see the insulating layer in dark blue and this area. So furthermore, we see a bulge in the right area, which comes from here and goes up the wall. And it indicates a defect of the horizontal barrier. Moisturizes mineral reacts and so gypsum crust forms to confirm that moisture is the reason for the gypsum crust. We use georadam. So on the far left side, you can see market profiles where we have measured with a two gigahertz antenna. And the image in the middle shows the profile of this measurement profile nine. And by moving the antenna along the profile, you get a two dimensional representation of the reflections in the ground. It's this picture. The profile is at a height of one meter 50. And what you actually see is that dark areas show strong electromagnetic reflection and light areas show weak electromagnetic reflections. And now you see the orange line, which marks the transition from wet areas to dry areas. So what does it tell us now? Let's look at the attenuation. The last figure here on the right just turns the attenuation of the direct wave and the reflection. So the direct wave is the wave which runs directly along the surface in this time slot here. So what do we see? High beta value means low attenuation and a low beta value means high attenuation. And you can conclude that the low beta values are generated through to moisture content. You actually can see that you actually can see that the area of the horizontal barrier is in this field and the damping decreases appropriately. So you can possibly use it to identify moisture content without damaging the wall by drilling, for example. So let me give a short summary. Thermography and georad are non-destructive methods to analyze damages in historical buildings and they are sensitive to the thermography for shallow sources, while georad is more sensitive to deeper structures. And the combination of the two methods can provide a new insight in the detection of moisture to not to not distract the walls of the historical building. So and if you want to learn more about it, I have a session on Thursday. You can join it and my post is also online. And thank you for your attention. Thank you, Janice. That was really interesting. So shall we head over to our last speaker, Veronica. Yes, I will share my screen. Okay. Here we go. Now start the presentation. Okay. Can you see my screen? Yeah. We've got it. Yes. So I will talk about a data-driven approach in the search for Antarctic meteorites. My name is Franny Katollenaar and I have been performing this research together with a team of scientists from the Netherlands and from Belgium, related to the universities in Brussels, the ULB and the VUB and in the Netherlands, the TU Delft. So with satellite observations, we can predict where to find meteorites in Antarctica. This is very important because meteorites form the understanding of the formation and evolution of our solar system, which in its turn teaches us about the origin and sustainability of life on Earth and also on other planets. So here you see Antarctica. And in 1969, there was a Japanese team who went on fieldwork to this location. And somewhat coincidentally, they found nine meteorites in a rather limited area. Here you see a newspaper article published about this very special find. It's really, it's not normal to find nine meteorites in a limited area. It's something very exceptional. And so from this exceptional find, the belief originated that there is a concentrating mechanism that brings meteorites together in specific areas. So basically what they think that's happening is that meteorites fall on the on the ice sheet in Antarctica and they become embedded in the ice whilst the ice sheet is building up. And then due to the flow of the of the ice, the embedded meteorites or at least a part of the embedded meteorites, they are brought to the surface. And once they are at the surface in specific areas, they can remain there really for thousands of years. Here you see these meteorites. And in the areas where they're brought to the surface, typically blue ice is exposed. This underlying blue ice, it makes it very easy to see the meteorites when you're in the field. And given that there is a high concentration of meteorites in a limit, limited area. It makes Antarctica a very productive place to collect meteorites. This is also why 62% of all meteorites recovered on Earth are recovered in Antarctica. Almost every year there are meteorite meteorite collection missions to Antarctica. And in this missions, the Belgian a keep is very well represented. Here you see a picture of the last field work they did. You see the field camp with the containers where they sleep and also the ski dues that are used while searching in the blue ice area. You also see nicely how big this area actually is, although it's still limited and there are a lot of meteorites there. So the Belgians they have been involved in four joint field work missions and they were very successful because they collected more than 1300 meteorites. However, the conditions in Antarctica they can be very extreme. The average temperature at places where they found meteorites is minus 36. And also the missions are not always successful in some sense you need to be lucky to find meteorites. So, until now, the places where they go to search for meteorites they are selected by experts who study maps and visual imagery, and they have performed a lot of costly reconnaissance missions. Of course, these missions are expensive because you need a lot of support to get there and also there are quite some dangerous risks that are involved in these missions. So, this is basically where we come in because we want to increase the success of these reconnaissance missions, and to increase our success, we will, we want to use satellite observations. To measure from space it's not possible to directly detect meteorites, because the typical resolution of the data that we use is 500 meters, whilst the average size of meteorite finds in Antarctica it's a couple of centimeters. So, we need to use indirect properties of the concentrating mechanism to predict where you find meteorites. So, if two examples here, you see the ice flow indicated with the arrows. So typically in areas where you can find meteorites the ice flow velocity is very low. And also you see the surface temperature is also very low in areas where they find meteorites, it rarely exceeds minus 10. So it's really cold there. So we use these indirect observations to predict where you can find meteorites. How this goes into practice is indicated here. You see all different layers with observations and again the two examples I indicated the ice flow velocity in pink and the temperature in green. And basically we combine these observations to predict to make a prediction. However, combining these observations is not as straightforward as it might seem so it's, it's not possible to make simply an overlay analysis because there is interplay between all the different observations. So we bring in machine learning, because this is a this the machine learning algorithm is able to capture the interplay between all the different observations. So basically what it does the machine learning in it spits up the data into a set of training data and a set of calibration data. So we give the training data to the machine, and we see how good it can predict based on the training training data by using the calibration data. And then of course we go back to the first step because I'm usually in learning there is some iterative process involved. In the end we get a prediction of where to find meteorites as shown here on the right, this is a snapshot of the continent wide product we are making. And the results we obtained are evaluated with unseen test data so this is data of places where meteorites are collected or places that turned out to be absent of meteorite finds. And this data has not been used for training or calibration whatsoever it's just completely unseen by the machine and we use this for the evaluation. So, the results of our evaluation, it indicates that the meteorite map we made is great advancement over the current manual approaches in meteorite searches. And basically we can create a kind of treasure map of Antarctic meteorites using satellite observations. Also, there are indications that there are still a lot of meteorites really many more meteorites to find in Antarctica and we believe that our treasure map will be a great asset in the meteorite hunting missions, allowing a targeted approach to collecting the remaining meteorites. Our research will be peer reviewed and at the meantime, we are making an interactive tool not only for scientists but also for the general public to just explore the continent of Antarctica and the meteorite collection, the meteorite concentrations that are in Antarctica. And I'm very happy to share these two or other results or materials so don't hesitate to contact me on the following email address. That's it. Thank you. Thank you Veronica I really like the idea of this, this treasure map as you've described it as well. That's cool. So we've heard from all three of our speakers there three really cool and diverse topics. And so now open up the floor for any questions. As I said at the start, if you've got questions you can either type a queue, a letter queue in the chat box, and I'll come to you for questions. Or you can write out your question fully and that's absolutely fine. I can also read them out from here as well. Any questions for our speakers. Okay. Can I speak. Yeah. I have a question for you Veronica. So you mentioned a satellite satellite observation and I was wondering if also this method has been also coupled with with surface observation with surface geophysical techniques. Thank you for your question. I think it can definitely be combined with surface observations. However, the areas where meteorites are found they're really scattered over the over East Antarctica. And so the surface observations in in this area they're rather sparse because it's it's really more in in in towards the interior of the continent and the conditions are really extreme and it's hard to have a lot of in situ campaigns there with continuous data. So it is possible for more local analysis but for the continent wide analysis we are aiming to make it. It doesn't. It's not usable because there's no coverage. Thank you. So we have a question for Stephanie coming in from from Sarah Darwin and can you speak a bit more about how your work could be used to help drive waste management efforts on neighboring places. Yes, great question. So I'm mainly focusing on really the physics part of this to really understand how the particles are flowing, but of course as I said in the beginning it's part of a much bigger project. So there's actually a Pacific, well this is Southeast Pacific wide project going on to really include all the neighboring countries so Ecuador Peru, GB, etc. And so we, we in particular want to find case studies where we can apply our tool and see if it works for different locations not just for the Galapagos. But what they are also focusing on is just improving using education, talking to the governments, trying to implement better ways management strategies. So the way we help with that is also look a little bit at the sources so if we find plastic on the Galapagos, can we find out where it's coming from, what is the main source. Can we tackle those companies or should they focus on cleanups on the beaches on the mainland or should they focus on better waste management flows, things like that. So, I hope that helps. Sarah says excellent thank you. And I was wondering as well Stephanie. You showed us, you showed a slide with a map on it and it had some dots showing where you're tracking your your plastics. Was, was that supposed to be an animation. Yes, they did not say. No, but we can, we can, I think we've got time to watch it again. I'll try it. Yeah, yeah, I think that there's a trick, Fabio, is this right for sharing that you might have to. You might have to share screens, you might have to take a couple of options in the bottom left. Okay, yeah, because it's a gift. It's going. Oh, it's going. Okay. Okay, there were more animations so it's a bit sad. This one is the nicest I guess so yeah. Yeah, if you want to talk us through. And if you want to pull out anything from them that's absolutely fine. So this one is just a simulation with that ocean parser's tool that I was talking about. So these are virtual, virtual pieces of plastic so to say. And the dark ones here those are the ones that got stuck on the island so you can really see the islands getting contaminated more or less with with thoughts. So yeah we really like to just sit back and check this out. Many hours. Yeah, the other one was in the beginning I'm going to see if I can go back. So this was more to explain what we hope to develop in the end. So it's kind of showing if the color is darker there are more more pieces arriving at that location in time. So, yeah, so this is all simulation but hopefully and there will be a combination of observations simulations and things like that. Yeah. Excellent. So here's got a follow up question to that let me find it. Will your, will your drifter flights help you to identify the biggest sources sorry how to decode that bit. And maybe indirectly, I can explain a bit more about that. So, so what we are using the drifters for at the moment so we've ordered 50 of them and we're in the process of getting them transported to the Galapagos at the moment. So that field campaign is mainly focusing on on how the plastic moves from the ocean onto the beach so that that's that process that there's still quite a lot of unknowns. So that's what we, we are focusing on with this drifter field campaign. There were more planned also larger skill to get a better indication of the ocean current because that's of course that's all models at the moment. It's really difficult to measure the ocean can talk about that for a really long time. But that was a much bigger field campaign also much further away from the Galapagos really large scale flow patterns. But unfortunately, there are a lot of funding cuts in the project. So that part is for now, at least not possible. Unfortunately, so we have to cross our fingers for the future. Yeah. My fingers crossed. Thank you. So we will hop back to you and this for a question from Philly. And Philly says, can these tools be used to repair damage in the future. I hope it is possible in a long time but first of all it's to detect moisture without drilling in the water. So actually today you are doing some course to get information about moisture content in the striker walls and if you want to have it all over the wall you have to do a lot of course so and what we are wanted to do is to reduce this and just take three or four course on the wall and then just a measure with radar and thermography to get an information about where is water content and where is it not enter how you can see in a faster way in it, especially cheaper way and in a faster way, where water is and what you can do just before you start to restore your object before you're doing the restoration so I hope that's okay. Yeah, that's great. Thank you. Another one for you Eunice from Sarah. How thick do the walls need to be for your technique to work. Can this technique be used on different mediums. It's better if it's thin, because the georada has a penetration depth, but we tried it in non kitchen. It's another church in Schleswig-Holstein, and there were the wall about two meter a one two meter so one meter 90 maybe this thickness so it works great on historical old areas, especially so it doesn't matter how thick your wall is. Of course there's some regulation about the sickness but what we tried worked way. Thank you. And from Philippe and this is back to Stephanie. Can the modeling be also applied to areas like the Mediterranean or Californian coastlines. Yeah, it can be applied anywhere. Let's put it that way I know in our group, there's a lot of work being done on the Mediterranean so yeah you can check out our, I don't know where is her Twitter and you will find some research related to that. There's quite a lot of work and what we're doing at the moment is really confined to the Galapagos so before we can scale up to other locations we have to see what what are the parameters that that will be applicable worldwide and not just for the Galapagos so that's that's one step we need to make in the future to make it sure that we can apply globally. That's exciting. Any more questions for our panel. Any more. If not we can, we can leave it there. And all that all that remains to say is, yeah, thank you for our three panelists for their wonderful talks and taking time out to answer your questions. And you can also have a look out for more of the content and materials related to these talks online at the EU media media center press center and I'm sure all of us speakers will be happy to be contacted by email for follow up queries if you've got questions.