 I think this is a good time to start. So good morning everybody, or depending on where you are in the world. Good afternoon or good evening. Welcome to the CSDMass full webinar series. I'm very excited today because we have two speakers who are runners up for the CSDMass students modeler award. Derek Neuhardt and Dangang Si. Derek is from ETA Zürich in Switzerland and Dangang is from Boston University here in the US. The actual award went to Kyle Rai from the University of Texas and he presented his work on a grain stew plastic sorting of transported materials in deltaic areas. He presented that during the CSDMass meeting when I said in May I think it's a little bit of a blur, quite busy here. So, but before we're going to start with presentations, I would like to introduce or give the floor to Dr. Sam Harrison. Sam is an environmental modeler at the UK Center of Ecology and Hydrology and Sam spent a summer here at CSDMass in Boulder and he actually suggested to have a Euro CSDMass webinar series. So, Sam, can you tell a little bit more about that? Yeah, absolutely. Thank you. Yeah, so this was my my second one over in CU Boulder was as a part of a project where the bigger effort really is to bring together communities that do integrated environmental modelling, specifically communities that write software to help with with integrated environmental modelling, which obviously includes CSDMass and all the amazing work that you guys do. Plus, some software communities, some work that we do at our institute over in the over in Europe over in the UK. Part of this as I say is very much building a community of practice and we saw a bit of a gap here or an opportunity here rather to broaden, broaden the geographical reach potentially of the CSDMass webinars, the CSDMass communities over to Europe. So with that in mind, we decided this time around to run a couple of the four webinars as Euro systems webinars. They're obviously not just for European people, anyone from around the world can join them but we've kind of focused them a little bit in terms of where the speakers come from and made the times a little bit friendlier for the European people. So we've got to two webinars happening the first that have this Euro systems systems theme the first ones on the 5th of October so that's actually the next webinar in the series. That will be Gordon Blair from the UK CH my Institute talking to us about digital twins of the natural environment. And then the final webinar of the series on December the 6th will be Julian Jovey from the University of Lausanne talking to us about glacial evolution modeling. And you're all very welcome. Thank you Sam. That's great. And I just put in the chat the link also to the webinars that Sam just mentioned so you can register for them if you want. So that's there for you. All right. In no particular order. I think Derek you're first listed on the on the webinar in the announcement so I'm going to give the floor first to you. So this will present today on how in America models provide insight in our service processes influence tectonics on divergent and strike slip boundaries. So with that, the floor is yours. Okay, let's see. Perfect. I can see your screen looks good. Okay, thank you for the introduction. Thank you everyone. So today I'm going to be talking about the evolution divergent and strikes the boundaries in response to surface processes, though for the most part this is going to focus on rift systems. So to start I want to go over a little bit how, or I guess how risk work in the different types we see and there's three main types of rifts that we see nature. So if you have deformation that's localized in a very narrow region, you'll get a narrow rift, and then whether this deformation is symmetric along the rift central axis as you seen this one can have a narrow symmetric rift or if it's asymmetric you get an asymmetric rift. And then in some cases though defamation isn't as localized this and you actually get fault in current over hundreds of kilometers. And in this case you'll get a wide ripped. The first kind of work is you have extension being applied some extensional force that is thinning the lithosphere. And so as this happens, you begin to thin the lithosphere and some region. And so this occurs, both at the top and beneath the lithosphere and at the top we have brittle deformation so we have extension alfalting that then creates a depression called rift valley that we can see here, which is good place to then deposit sediment. We also thin the lithosphere underneath where we have upwelling asthenosphere that creates this region called the necking zone. And because we have warm material upwelling into this region, this also works to further localize the deformation so we continue to really focus our defamation into this small region. And so as we continue to pull we continue to thin this lithosphere until eventually we completely break the lithosphere and we start exhuming mantle material. This then forms an oceanic crust and we get into seafloor spreading. And so what I want to point out on this one is, even when we have a seafloor spreading, we can see that we still have some of these faults in this defamation that occurred that is kind of recorded in these passive margins so this is kind of the history of the rifting event that they will take seismic data to look more into. And so people that have looked into this have found that there's actually some trends and how a rift system thins the crust towards seafloor spreading. And so if they look at this thinning, they can actually divide it into a few different domains that they see. First, when you don't have much thinning at the crust but you still have some faulting, this would be known as the proximal domain. Eventually you begin to start to drastically thin the crust. And this will be when you're into this necking domain. And eventually you get to hyper extended very thin crust and that kind of rate changes a bit and this would be a hyper extended domain. And then after this in some cases you'll assume the mantle lithosphere that's underneath, and then eventually into seafloor spreading. It's important to see this because we see examples of relict rifts and also active risks throughout the world. So an example of one of these that is active as the East African risk system down here, where we have two active rift segments that are splitting apart the African continent and so this is in kind of those earlier stages I was talking about we're still breaking apart the lithosphere. And if this goes on long enough. Can everyone see my mouse. Just a yes, yes, we can see your mouse. Okay, it keeps disappearing on me I was confused. But so this goes on long enough you eventually get to your mid ocean ridges like the Atlantic Ocean. And then in these zones if we then go to near the coastlines we can find these passive margins so for example the new foundling and the variant or conjugate margins that would record the history of this rifting event. And so there's been a lot of previous work that has looked into how these rifting works and models, whether this is looking at 2d models that are trying to see what exactly creates asymmetric rifts or maybe how risk and deactivate and reactivate in different areas, or even large scale, 3d models that are trying to look at how rift armed interact like in the East African risk system. But one thing that is important to consider is how we look at the surface in these models. So there's a few different ways we can do this. One of the most simple one is to use a free slip surface so this is where the material can flow along the surface but you can actually deform the surface with the stresses that are there. So a lot of models actually use a free surface which you can see on the left here, which then allows the surface to use these stresses and to form through time. But one thing you might notice here is that we get these unrealistically high topographies at the surface and also unrealistically low values that form this because there's nothing that can work to deform the surface through time. And so additionally because a lot of this defamation happens near the surface these changes to loading based off this topography topography can affect how the defamation evolves. And if we were to include some surface processes will no longer have this issue and we actually redistribute our surface through time, and this can then affect the brittle defamation that's happening at the surface. And so there has been some work and rifts also that has looked at adding this surface processes to these models. And this is kind of done in two different ways. What you can see here in the top is where they use just a one D diffusion equation, which is kind of short range transport of the sediment and they put this on here to see how these basins everything affect this defamation. And the second way that's becoming more common now is to couple these tectonic or geo dynamic codes with dedicated landscape evolution codes. And so here's an example of this down here where we have so Paul, coupled with fast escape. And this has the added benefit of having more complex surface process laws, such as the stream power law. But one thing a lot of these models have found looking into this is that surface process generally increase the localization on a fault and this has the effect of prolonging the fault lifespan. And one thing I want to mention is a lot of these previous studies they take all kind of a qualitative look at the model evolution so it hasn't been much quantifying how faults change based off of these surface processes. So during this talk, the main questions I want to go over are how to surface process effect and evolving rift and it's fault network in 2d. And this is something we want to take a quantitative look at the fault network. The second question is can sediment loading drive flexible substance along strikes the fault. And the third which will be rather short section is just what effect may surface processes have and large scale 3d models. And the method for doing this is going to be using a method that I worked on during my PhD which was a two way modeling between the open source codes aspect and fast skate. So aspect is a gdm code that has been used for subduction models plume models and riff models. And one of the selling points of aspect is that it has this adaptively refining mesh so you can have the mesh that changes over time and keeps refining the areas that you're actually interested in. So aspect handles deformation of material within the model is through two different ways generally the brittle deformation and so this is where you'll have your faults. And this is something through like the buyer leave law or the drug or Prager law where near the surface when you're cold and brittle, generally you'll focus defamation to very thin shear zones that could represent faults. As you get deeper into the system you have higher pressures you have higher temperatures, so you can no longer have these faults forming. And in this case you begin to have rock that flows more like a fluid into the ductile defamation regime. And here we use a harmonic average of two different laws the Newtonian diffusion. And Newtonian diffusion or a strain rate dependent dislocation law. And so we took this model and we coupled it to the landscape evolution code fast gig, which is available in quite a few different languages for our purposes we use the FORTRAN version. And so this is just something that applies a set of routines to deform the surface and a road and deposit sediment. And this is just a quick example video of this, having I believe uplift in the stream power law applied to it. So fast it can deform the surface in quite a few different ways, and this depends on whether you're above or below sea level. And so if you're above sea level, the change in topography over time is related to the uplift rate and the infection term. And so these are two things where aspect will send these velocities to fast escape and then fast people use them to change the surface. And then just into this we have we apply the stream power law which is how rivers are incised into the system, as you can kind of see over here where we have these river systems forming. And this is based off some erodability coefficient kf and then the drainage area and the slope. After this there's a sediment deposition term which is just to say that this these rivers holds some amount of sediment and out of the sediment some part of it can be deposited back onto land which again you can kind of see in this image with some deposition occurring down here. And then lastly it has the hill slope diffusion term which is just a diffusion the short range transport of sediment. Below sea level it works similarly but there's a little bit of difference the main. The similarities is that again we have the uplift term the infection term so we move our surface. And then we have a diffusion term as well but we'll use a different coefficient in this case. And the major difference is that there's this continental sediment flux term and so what this says is that all the sediment that was eroded from the stream power law that did not get deposited based off the second term will end up in our oceans pushing a lot of sediment into there. And then lastly for these models we had some constant background ocean oceanic sedimentation rate. So the first question I wanted to talk about is how does surface processes affect an evolving grip dance fault network. And so for this we set up a 2d aspect model that's tied to a 2d fast state model. And the way this works for these 2d models is that the velocities from aspect surface will be sent to every node along why. After this you'll run fast escape you'll find the difference between the new surface the old surface, and then you'll average this again back along rely down to one single point and send this back to aspect surface so we can deform. And so for fast escape we have fast escape on top with all the laws I just talked about and then we have our aspect model where we have lithosphere that's 120 kilometers thick and then 35 kilometers of crust, and then we pull this at a maximum 10 millimeters per year to see how it deforms. And so what we're really interested in is before I talked about we have these three different rift types so we wanted to know how this changes with wide rifts asymmetric rifts and symmetric rifts. And then also we want to know how surface process affected. So we either have no surface processes at all or we change our kf value, the rotability of our stream power law to simulate very low erosion of very very high erosion. And so here is a video of the symmetric model evolution and this is kind of like a qualitative look at how you would watch these models to see how they would evolve so we have the top portion of aspect and then our fast escape surface above. So what you notice is we immediately form very large faults in our initial rift Valley and then over time we can begin to perform progressively smaller faults toward continental breakup. And then also when we look at here we can see that this would be our passive margin over here that's recording the history of these older faults. And so as I mentioned previously we wanted to have a way to kind of quantitatively see how these faults are changing through time. To this end we use the fat box of fault analysis toolbox as written by T. Lerona. And so what this does is first we take our plastic strain field and so this is where we've accumulated brittle deformation so tells us where we have an active fault or we at least had an active fault at some point. And so we take this and we threshold it down to get these shear bands, and then we discretize these down to individual line segments. So without this, the toolbox will apply a few routines that will separate these line segments into distinct faults based off of splitting and things like this. And then it'll apply a label onto each of these and then we can track this label and the fault through time for every time step. And then on this fault label also we can keep track of any property really interesting seeing such as a slip rate the displacement or the dip angle or anything like that. So this is the exact same model, but this time we're going to show the fault that worked through the, through with the toolbox tracks and in this it'll show an active faults and black and then red and blue will be active faults. So again first we see we have many faults that quickly localized onto a few large faults. And then first these all stay active until eventually becomes so weak in the center that we shift deformation inward and deactivate our outer faults. And then finally during seafloor spreading we just get a few faults that are active a time that are also quickly replaced. And so we wanted to see how this actually changes through time and we look at these fault parameters so for this we're going to look at just the cumulative active fault parameters, specifically the displacement the length and the number of faults that are active in the system. And what we find is that there's actually about four or five distinct phases that a RIF system goes through as it evolves. This is the early phase we saw in the video where there's tons of active faults that are competing with each other, but eventually these localized them just to onto a few major faults which is the end of this first phase. After this we saw that we had all these faults active but we're continually placing new faults into the system. And this would be our fault system growth phase. But again as we saw in the video eventually we get so weak in the center that these outer faults start to turn off and focus deformation inward and this would be the start of this fault system decline. From this point on we begin to continually replace these faults with smaller and smaller faults, until eventually we enter C4 spreading or continental breakup where we just have a few active faults that are replaced at a time that we can see over here. And so the next thing we want to know is well this is how this model evolves but how does this change if we change the amount of surface processes we have are the roadability from our stream power law. And so with this we go from having absolutely no surface processes up to using a very high surface processes where basically eroding any topography that can form. And what we generally see is that like the previous studies surface processes works to enhance fault localization, which leads to longer live faults. And what this means for the models is first that because you have these faults lasting longer is less need to create new faults so you end up having less complex fault networks with more surface processes. The second is that it actually delays continental breakup so we can see with no surface processes reach continental breakup around 11 million years, which then goes to 13 to 14 up to even 18 million years when we have very high surface processes. And so the other thing is again we applied this all three different right riff types. And what we found is that generally, regardless of the rift type we see the same four or five distinct phases. The only major differences between this is in the wide rift where we have deformation spread over a much wider region. Our fault system growth with all the faults active last much longer. And then for asymmetric we actually have a fourth phase between the fault system decline into the continental breakup. And this would be a rift migration phase we're really building the asymmetry into the system. But we also found that regardless of these three rift types surface processes had the same effect where you ended up with less complex fault networks in a delay and break up with more surface processes. And so the final thing I want to talk about for this part of the talk is that we mentioned how there's these three. We mentioned how there's the, what was I saying, the rifted domains that they see in seismic data and we want to know how these phases compared to these rifted domains. And we found that generally they compare quite well there's phase one distributed definitely similar proximal into the fault system growth. That's like the necking and the pulse of some decline that's like hyper extended into the oceanic. The only major difference we saw was in asymmetric margins where when we look at in the direction of rift migration. We kind of can't tell where we go from phases two to four. And the reason for this that you can't really see here, but there's a nice video of it. That I don't think they'll be time to show unfortunately but as the rift is migrating you actually begin to break up all the passive margin you have here and it gets translated to the other side so essentially we're eating up all this history that we could have taken a look at while this rift is migrating. And so we didn't compare this to any real world seismic data but there is a paper border at all that is maybe released now that is trying to compare this to the Norwegian margin. And so the question for the beginning this was how does service process effect and evolving rift and fault network and the main conclusions with that delays continental breakup reduces fault network complexity and this is because it works to prolong fault activity. So next I want to talk about a specific region, which is the Andaman Sea off coast Thailand. And so this is a region that initially was an extension so there was written that curtain this region, but then it shifted into a strike slip regime. And so what we're interested in is this South Seagine fault down here so this is a strikes the fault that formed in this thin lithosphere from this older rift zone. And specifically along this fault is a basin called East Andaman Basin, which is a laterally extensive basin that exists on both sides of the fault. And what's interesting about this basin is that when you think of a strike slip basin, you think of a polar part basin which has two offset segments that because of the orientation create extension between the segments that forms a kind of narrow thin base. But in this region there's actually no, no offset segments that there's no really way to form a polar part basin. And in addition to this, the basin is kind of interesting that on one side of the fault it's very thin and uniform. And on the other side it's thick next to the fault but then thin strike perpendicularly. And so if we look at the region during the time the fault was active there was something called the Mergari Ridge that was sub aerial. And so the idea is that maybe there was a lot of sedimentation applied to just this one one side of the fault. And so the question we had from this was, well, if we have a large sediment load that's being put on to this weak strike slip fault, can we drive flexural substance along strike slip fault. And so for this we set up a very thin 3d aspect model where we have thin with the sphere that's 40 kilometers thick with very thin crust of eight kilometers. We then move one side while fixing the other so we can self consistently form a strike slip fault in the center. And then the two sides of periodic which is to say, even though this is a thin model it kind of works like you have an infinitely long strike slip model, because material that goes out one side will come in the other side. And then on top of this because the whole area was underwater, while it was active, we only have marine diffusion going on but we add sediment to the system through constant uniform sediment rain. And also we assume that we have some offshore sediment source like the Mergari Ridge, and through diffusion push sediment into the system from one side. And so here's just a quick video of the reference model. So you can see we quickly localize a strike slip fault in the center and at the same time we're pushing a lot of sediment in through the side. What you'll notice is that the sediment is being pushed in actually not too much of it is making it to the other side of this fault. This is because this fault is working as a weak zone where once the sediment gets up here, the loading actually deflects the crust and let the sphere down so we end up with this thick basin here and thinner on the other side. And so one of the important questions for this to is that we have this thin lift sphere in the region, but how does this work if we would have thicker lithosphere so we took our lithosphere thickness and we buried it from 30 to 60 kilometers. And what we found is that when you have this thin lithosphere like 30 kilometers, you get a very asymmetric basin that's much thicker on the side of fault that has this high sedimentation up to about five kilometers. And if you thicken this up to 60 kilometers, you can already see that there's a much, much smaller basin forming with the maximum depth of about a kilometer. And you can also see this dashed line would be our initial surface initial surface so we're actually just accumulating a lot of sediment on top without actually deflecting the lithosphere. And then so just for a brief comparison comparison between the basins. Here is the interpretation of the east and then basin from the seismics, where we see the defining features are on one side of the fault we have a thin uniform basin. On the other we have a basin that's thickest near the fault and things towards the sediment source. In our reference model basin we still in very similar we have much thicker shears and of course from our fault, but then we have a relatively uniform basin on one side and on the other side we have basin that's thickest next to the fault and then things towards where we were pushing sediment in. And so the question for this section was can sediment loading drive flexural substance along the strikes fault. And we found is that yes if you have enough sediment can actually create a flexural strikes the basin. However, you need to have a pretty thin lithosphere for this to happen if the lithosphere is too thick you just can't deflect the lift deflect it at all. And this is because the fault acts to it works as a weak zone it decouples the two sides the lithosphere and allows accommodation space to be created. And the last pretty short something I want to go over is well what effect does this have we put it on the large scale 3d models. And so this is just an example of a model, an aspect in the same fast scape surface on top of it. And so to start with this. This is an older study I did and so this was just checking how offset riffs connect. And we found that based off of their ex offset and the crustal strength they either connect through an oblique connection, or if you get farther apart they begin to overlap and you rotate the central micro plate. And so when we were first testing this coupling, we just took one of these models and put it on top just kind of to see what it worked. And specifically we took this transform connection model. And so here's just a brief video of the fast scape surface during this evolution. So you can see we form our riff valleys and a riff links that have a lot of sudden being pushed on to them are often them especially into that basin. And so what we'd expect from the free surface model is that these two sides would propagate forward and then they would connect to a transform But we actually ended up seeing in this case is that we're not actually seeing this transform both form we're actually seeing rotation which is more into our micro plate regime, which is what we'd expect if we actually had a larger ex offset than what we have here. And just to really showcase this so this is to the exact same model one run the free surface that forms this transform fall as we expected, and just by adding the surface processes we actually change this regime into this micro plate that we didn't expect from these ex offsets based off the free surface model. And so I want to know that this is more this was just a test so it wasn't really the most realistic setup, but I think this just illustrates that surface processes definitely can have an effect on these 3D models and something to look into in the future. So the major questions for this talk or how does surface process effect and evolving rift and it's fault network, can sediment loading drive flexible substance along a strikes of fault. And what effect may surface processes have on large scale 3D models. And the final conclusions are that when you have service processes on a rift system, it works to delay break up and it also reduces default network complexity. And then for the strike slip case, we saw that if we have a sufficiently thin lithosphere we actually can form flexural strikes the facings. Assume we have a lot of sediment going on top of it. And finally, we haven't looked much into it but it seems like surface process may even become more important large scale 3D models. Yeah, okay. Got a little lost there. Okay, that should be it. Wonderful Derek thank you very much. I didn't say so at the beginning but you can ask questions. If you want you can either unmute yourself or type in a question in the chat. So we've got a few minutes for questions here. And while people are thinking about their questions. Let me start off with one. So you showed, I think it was the strike slip case where you showed sediment loading on one side of the fault right and, and you indicated that it would. I think it would keep the fault longer active. I was actually wondering, does it change the. I'm not a fault person. I'm not a tectonic person, but does it change the magnitude, I guess, of the of the of the fault being, I guess, the movement of the fault itself by loading it more on one side, would that accelerate the strike slip movement. It's an interesting question I'm actually not sure I would think. I think it would depend because a lot of the strike slip movement. I mean generally you're going to you, we'd have something called plastic weakening so we weaken it and this probably determine the maximum movement. I think at this point when this is applied, it would already be maximally weakened, but this is something especially if you don't have these like set results says are kind of not weakest strength is set in the model and also resolution base so in the model it might not make much of difference but if we could, you know, theoretically have infinite resolutions or be closer to reality maybe this could make some change by weakening the material as it's moving as well. And I see that my got his microphone and muted. Yeah. So, looking at the riff breakup models in terms of sediments what came to mind was that a number of riffs like in South America and the Red Sea. There's a huge amount of evaporates that accumulate and so suddenly you're getting an input of a large amount of sediments. Very quickly, and I'm wondering your thoughts about how that might change tectonics or affect the timing of breakup. Yeah, that's a good question to. I guess I don't know enough about the difference between like evaporates and sediment was put on but if we just assume it's like sediment and we put it on very quick. I think the main change that it does is that initially you know you're putting a lot of cold sediment on top and this actually thickens kind of the brittle layer. So this will keep these faults active and keep the riff going longer. Okay. But over time, of course, when you're putting a lot of sense you can also warm it so it's a bit of go either way I guess. Right well and salt has a very high thermal conductivity so that it would promote the heating up again. Yeah so maybe this actually would would not be as much of a difference than if it's heating it as well. Interesting. But it's an yeah it's an interesting thing as to how that might because it's it's you know often right around break up. Okay Mike. There's another question from Lester in the chat. I can briefly read it out to create dark dark. Lester is interested if there is some sort of a script available where you couple both the models of escape and aspect and if that's available for the community that that's one question and the other question more than one practical question is about. Do you think that slow evolving landscapes, like at the combine nor chili could have a strong implications in the structural evolution. Okay so for the first question, it's not quite completely on aspect so that the plan is eventually you should download aspect version, and then you'll already have the coupling script ready to go. You could find it on my GitHub, so it is freely available. And we're still working getting a push but if there's any questions you could just email me to and I can help. And for the second question so, could you say it again. Yeah. Do you think that slow evolving landscapes, like the other comma in north chili could have a strong implications in the structural evolution. It's interesting to I would think, generally, when you change the rate that you're pulling the rift, it doesn't really change the structural evolution that much they end up pretty similar. But in a case like this, and this is maybe similar with the evaporates to is if you're doing it much slower and you put the same sentence on this going to be more time for to warm up to forming. So I could see maybe surface processes wouldn't change this as much, although maybe also you have more time to accumulate sediments so. Interesting, we didn't try this at different velocities, but I would be curious to see. Wonderful. Thank you Derek. And thank you very much for your presentation. If people have more questions about their Derek's presentation. Please put them in the chat and maybe Derek you have time to answer them. Don't see will present. So we're going to move on to the next presentation. Don't see will present today on responses of mangrove forest to sea level rise and human interventions using bio morpho dynamic models. And the floor is yours. So, very interesting to talk about Derek for it's a very large scale, a very long time scale. So, the difference that I'm going to present now it's kind of like where we really smaller scale compared to Derek. And it's my PC topic rate to coastal mangroves, and we folks on the how mangrove respond to sea level rise and human interventions. So I'm going to start with some very funny figure. There you go. Yeah, so for some who never see the mangroves, these are mangrove seas that are generated from their mother trees and can be distributed by the wind, flow and even animals. And as long as this sees like deposit or strength on the ground, they will try their best to get that loose longer and strong enough and therefore we can become seedling and become young mangroves. Of course, some seeds will be fresh away by the strong currents. Well, it will take some like decades for the young mangrove to come this mature mangroves. Yeah, if you have time you can release a boat here and start to swim in the in the main group for us. And you wouldn't be surprised to say hi to these animals, but definitely not get too close to me. Yeah, it's kind of dangerous. Yeah, so I was always intrigued by the complex system when I saw in the main blue system here. So this blue system also called brazing lose, or we say new metaphors because these lose, they can take up oxygen from the atmosphere and they can filter salt from water. So this extra secret that why they can stay on the sea water for a relative long time. But one thing I would like to highlight if you look into this profile ship here is there are more sedimentation in the mangrove sites compared to non sedimentation or let it be sediment in adjacent area. So I think it is the power of the mangroves which can trap sediment and keep them like I'm flooded by the sea level rise. So this is one type of mangrove roots that we just saw. And there are some other type of mangrove lose so this is like the red one on the left, and this little black one. So we call this opportunity and these are even dense lose but these roots are more like a pencil. So this one is more like a brother this one is more like a pencil sticking out of the ground. And one thing we can tell directly from these two figure is that the loose on the left hand side, it's much more dense so basically the mangrove density can vary with different species, different age and their size. Something we're going to talk about this more later on. So mangroves are also under pressure nowadays, so they can be easily thrown by sea level rise because you can imagine them when the sea level start increasing. So mangrove can endure a relative period of inundation but the sea level goes too crazy they can still be drowned and even die. And human intervention for example they remove mangroves or building the coastal construction somewhere can further constrain mangrove coverage. And we also find now that lost mangroves can further exaggerate the coastline retreat. So nowadays more and more research is focusing on the fate of mangroves using different techniques, like the geo let's call data analysis, meta data analysis or geo imagery analysis of the observations, etc. And all these research can be nailed down to one general questions how do mangrove response to accelerating sea level rise and human interventions. And you know like you numeric model can be an ideal approach to to give you a very special 10% spatial predictions in the future. Well when I look into the Google Scarlet, I found something interesting so when I see okay mangrove numeric model or so much numeric model. So this is one of those very important coastal pools. And I surprisingly find that much less studies on the mangrove numeric models. So this is definitely something that our scientists need to need more emphasis on this stuff. Yeah. So, to capture the main group response to the changing why we need to understand how mangrove interact with their surrounding factors, at least what we call bio more dynamic feedback. And this has been widely adopted by the current coastal weather models. So for the bio part, we need to consider vegetation dynamics for example how vegetation colonize grow and mortality in the end. The more dynamic part is actually the process that we need to think about how salmon transport together with the water movement and how salmon deposit or be loaded from the ground and the bed level start to update. And of course that leads to interact with how vegetation influence the salmon transport and how the flow influence the vegetation grows. So there's a tiny animation. I made to show how my model works. So many vegetation colonize on some suitable area and they start to grow and compete with each other and some die and some may keep on growing. So sea level start to rise and statement got more time to deposit on the ground and start to bed level increase. However, if the sea level goes too crazy that mangrove will die eventually because like they cannot endure very longer in nation. Now we have the model. The first thing I would like to test is how mangrove responds to different sediment concentration in sea level rise. So I'm going to use one second to introduce how we can read this figure. If you look into the right hand side of fear. It's one day model. So everything's very simple. Each line represents the coastal profile. This is your site and this is language site. And you see now there are three different color dots and each color dots location of these colors represent the location where mangrove coordinates. And the right one is a kind of mangrove. They prefer a longer and Asian periods. And this blue one represent what mangrove and let prefer a bit shorter and Asian periods, while the black one was somewhere in the middle. And I'm running the simulation in two constructing simulations. And if you first look into this one low sediment concentration and high sea level rise rate. So this system looks very stable. You didn't see a lot profile evolution, while all the mangroves kind of shifting downward. So we can say in this systems in a low sediment concentration and high sea level rise rate mangroves go hanging hanging seems like mangroves is dependent on the sea level rise rate. If you look into the insert, which represent the total extent of vegetation. So the extent of vegetation remain stable because there are no change on the profile everything moved inland. So not a lot changes expected. While if you look into the high sediment concentration on the right figure. So there are a lot of sedimentation happens and the coastal profile start to build up and the movie in the mangrove seawall. And there are slightly language retreat for this white mangrove. So in general, the total extent of vegetation will be extended. However, when we include a die on the land will part of this coastal area or coastal regions. A lot difference happens. So if you look into similar like a low sediment concentration high sea level rise rate. All mangroves start to retreat land growth. Why we didn't observe like the blue dots, which represent what mangroves cause like the what mangroves they are learn more space has disappeared. So we can kind of see up the reduction of the coast mangrove extent. Why if you look into the high sediment concentration and low sea level rise rate. The vegetation start to expand in seawall. Of course, there are no language retreat, but the overall lip lip behave like quite similar to a scenario without barriers if I'm going back force and I can compare so this is the high sediment concentration with barrier. So actually something similar happened here. Yeah. So one thing I would like to highlight if you look into these regions. As we're interested in a species replacement occurs during the vegetation expansion. And I'm going to quickly expand, explain this within this tiny product. So during the vegetation expansion and sediment transporting to this location is reducing because the sediments providing actual risk and resistance. So less and less and can be deposed in this area and leading to a smaller sedimentation rate in this area. While sea level rise, keeping on increasing eventually will increase these locations, higher appear higher. And for local species, they cannot endure longer and higher period and start to die. While there's some new species which can. Endure a longer species works of Coronas. So this makes a species replacement happening this location. Yeah. So as I mentioned, like the, the density vary with the different main group species ages or locations. So I also investigated these impacts of density in my model results. And now I run two different scenarios one is the way sparse rules and one was doing dance rules. And some interesting regarding the sediment concentration along this profile. So when sparse rules you see that more sediment can be transported to inland area. So the black one is this side here. You can see more sediment can be transported from offshore to its inland area. But if you look into the dance rules, make basically most of sediment is, is retained in offshore area. So the different sediment concentration behavior along this profile will eventually shut up different profile shape. So this is like class your term sex of this profile. And if you look into the more dense lose in there will see the provides much more propagating seaward and less more deposition in the scenario of them. The sparse lose scenario. So what interests me is like the different look density will shape up different vegetation diversity. See, like this one, you have a very equally disputed vegetation species under the sparse rules, while in the dense lose situation. It's kind of like whole system is gradually dominated by these red mangroves, which means like the diverse is kind of losing. Yeah. In the end, I would like to present this diagram to summarize these findings. So I determine all the mangrove condition is two different scenarios. One is low in one pressure, one is high in one pressure. In low in one pressure, which is the high sediment supply and low sea level rise. So the high sediment supply and low sea level rise and main group calories will both increase. But what if this brings is that when you have the dense lose the vegetation diversity will reduce as we observe like the eventually evolve to a one speech dominated systems. Well, in sparse lose them a scenario. So all the vegetation diversity diversity will remain stable because we see a relatively stable mangrove species. Well, in terms of the high in one impression, which is the high sea level rise rate, but the low sediment supply. So if there are no barrier, the system will be controlled by the sea level rise rate because we see all the mangrove species retreat to landward hand in hand, and both the main group coverage and diversity will remain stable. But if there are barrier like build it just after the main mangroves, you can see like something can actually disappear and the total mangrove stand will reduce. So this is a very simple model, and we focus on only one title range systems. And next, I'm going to extend my scenario to multiple title range systems. So this is a family photo to study in the main group responses under different title range systems and the sediment supply. So from left to right is increasing title range and from the top to bottom is increasing sediment supply. And now like to first on focus on the first row, which means that different sediments different type of range, but the same sediments of prime. The first thing is that we realized the profile remain relatively stable and they are not much like the vegetation expansion or retreat. And we see like a very comparable vegetation extent. And now if we increase the sediment concentration, you can kind of define that. They are more like sediment accretion, a proper accretion build up and the coastal soil, it kind of become gentle. And it over always turns extent is increasing. Now if you look into these seawall mangrove age, mangrove tend to colonize a bit higher. If you look, it can bear here mangrove is right close to mean water level. As we increase the sediment concentration, mangrove tend to like corner speed higher location. And I call this area, the distance from the mangrove seawall age to mean water level as in a nation buffer space. So in the field we actually find in a nation buffer space is increasing based a different title range and this is very important for mangroves. So when sea levels start to rise, the seawall mangroves will not be inundated immediately because you have inundation buffer space. So now I'm going to include sea level rise here. And they are only two different kind of dogs. So this green dots represent scenario right before sea level rise and the blue dots represent the scenario after 100 years of sea level rise. So this is again like similar family photo you have increased tight range from left to right and increasing sediments right from top to bottom. So I would like you to focus on the first column. This is the micro tight range of various small title range which go upside and down one meter. And you will surprisingly find all the mangroves start to retreat language immediately after the sea level start to rise, even under high sediment concentration. So here is the high sediment concentration. And if you look into the large title range systems, like this scenario, even in a small sediment supply, there are less language retreat of all these mangroves. So if you compare these two scenarios, we can actually draw a conclusion these mangroves in the micro tight range system probably will be more vulnerable. And it's interesting to see like they are also stable mangrove seawall age in this scenario here, even under the inter-title, even in the inter-title scenario, in the intermediate sediment supply. And I explained this with the inundation buffer space. Like as I said, the mangrove colonize at high location, which means you have like a lower relative hydro period. When sea level start to rise, what level increase, there are more frequent inundation appears in this mangrove side. Well, this relative hydro period will still remain below the stress hold, which means mangrove steel like survive. So mangrove will not be inundated, will not be inundated immediately within this study period. So a very quick technical measure of this study. So the people think about the mangrove extent will be increased with title range. But from our model is actually not true. So the mangrove standard can remain similar among different title ranges, because we need to consider the coastal slope. And the second thing is that we also observe a stable mangrove seawall age. And this is something actually people always neglected because we need to think about whether they are inundation buffer space right before mangrove seawall age and everything is very important. Because the whole world is saying, okay, mangrove is under the high pressure, but is this the real true case? We don't know. Yeah. And the last thing is that the mangroves in the micro title system is actually more vulnerable in other systems because we already see like the sea level rise will be inundated mangrove forest and making the mangrove restrict language. Yeah. So something different. As I said, like a whole world think mangrove is so importantly calling for to protect mangroves. Well, in some cases, like the New Zealand, I think mangroves are actually trouble makers because they muddy layer coastal regions. So what local people start to do is to remove these mangroves with the hoe they can restore their sandy beach. But the things that they have no clue about the outcome whether after after after they remove mangroves. So we wonder if we can use our model to to tell them the results whether or what can happen after they remove mangroves. So this is our study site in the in the New Zealand and North S3 and the bottom three and S3 can see the how fast mangroves is spawning seawall and where the location where mangrove are removed. So we built a model. We first run a model by adding a model to a load mass supply and to mimic the pre disturbance period, which means you have a low mass going to a systems and a very small main slow mangrove expansion. And in the end, we subsequently will also add high mass supply to represent the pose disturbance, which means you got a great amount of sediment was transported to a systems. And together you have a very fast vegetation expansion and stuff on this point we start to diverge our scenario to three different directions. So the first we continue the model run with a continued high mass supply and and that one is that we remove vegetation just like what they did now let's try to remove vegetation to see what's happened. And also the last case it will also run the reduced mass spread from catchment, which means you have reversed the mass spread from high to low and to see what's going to happen for the systems. So, before I step into further about my model results, I would like to think about why local government will think about removing mangroves is potentially a reliable approach to restore the sandy beach. So, we are also informed by at least local skill knowledge, the presence of mangroves will slow down title currents and title current so that title currents will that accelerated mud deposition and mud deposition potentially create more space for mangrove to colonize so this is actually like, I would say a vicious cycle cycle rule, right you have the vegetation deposition and more mangrove more mangrove more mangrove. But now, as long as we remove vegetation, it seems at least the binding effects was removed, and you have a stronger title currents and you expecting to lower down the mud the creation, and therefore you have that money regions reduced. So everything sounds very reasonable to me. However, when I see my broader results, I was surprised at the model results behavior. So this is the line graphs to show like the muddy regions of the of them of the basin in under different model testings. So today we're going to look into this green dots was to represent the mangrove removal scenario. It seems like a mangrove removal you got to the more muddy regions. And I was confused at first but now when I look into this for everything makes sense. So, this is a mangrove channel, and you have the mangrove colonized on these two sides. So the presence of mangroves will limit will converge the title currents within these channels, and all the sentiment was moving. And together with the ties going upside and down the sediment is not allowed, of course, not not well, the limit sentiment can be transported across these types of banks. Well now when we, when we remove mangroves, all these binding effects was removed and sediment is allowed to transport further, and you don't have the energy to be converged in these channels seems like the energy was slow, and they can have more sedimentation in these channels. So this is like schematic figure to show my hypothesis, but to demonstrate my hypothesis, I calculate the sediment sickness in different locations of the office spacing area. So today we're going to look into the five different environments we look into these three different variants with the future scenarios that just model. The purple one is that we continue to have high mass supply and the green one is that we remove the mangroves and the blue one is that will reduce mass, reduce mass supply from the upstream. We're first looking to generalize area, which is the least part. And we see there are more muddy sediment in this area from our model results. And this actually can be explained because all the title and it was distributed to other location and the, and the title and it always had a dynamic energy is as kind of smaller in this area which allow more sediment deposits in this, in this channel bit, you increase your amount of deposition in these channels. Well, if you look into a title phrase that this location is kind of closer to channels and this location is further away from channels. So if you look into this further away location further away from channels you see had that more sedimentation layer. So from the purple to the green you get more sedimentation. So which means our hypothesis cracked. So you have more sediment moving from the channel to the further locations. So only we can do is to reduce mass pride if you look into the blue, blue variants, both the channelized area and the title phrase they got a bit lower mass thickness if you reduce the mass pride. So in this case, I would like to advocate my complex model feedbacks. I mean, it's even complex now so I've caught like answer by more of a dynamic feedback. So initially, if we look into the local scale feedbacks where station slow down currents and encourage sediment deposition. Right. And we'll increase what potentially allow more mangroves to colonize so this actually inspired local people to remove vegetation constantly trying to reduce them modification or their systems. However, we from our model results we show when we remove vegetation last three have a more sedimentary distribution with you got a larger extra in feeling. You potentially will have more intertidal area and the creating putting more area for mangrove to colonize. So what we need to do is that we need to zoom out a bit further by looking to upstream, we need to conserve lands. And we need to reduce or slow down the decline the default, the frustration upstream and making less sediment being transported through the rivers through the channels to the coast. And this is what I'm going to do instead of like looking to the coastal area so this is actually the scale dependent feedbacks for the coastal management. So in summary, we should stop cutting off mangroves in the coast, but we need to manage the upstream land activity. Yeah. So all the code that I talked about today can be finding my pages. If you stand on this code here, you can easily get very cold and how like cardboard and a detailed instruction how you can start the model. Yeah, I think at least all my presentation. I hope I didn't exceed the time too much, but I'm happy to hear if you have any questions. Thank you very much for your talk for your presentation. Thanks. We're a little bit over the hour but if people can stick around and will be great. Now's the time to ask questions to Dungham. And I see Mike and muted so go ahead. Yeah, I'm just mulling over, trying to take in all the things and just. And I'm sorry because I work a lot in Bangladesh and one of the things where they're doing or thinking about anthropogenic changes of increasing the embankments is they're also thinking about mangrove forestation around it and the question. One so one question is whether or not that will work in the face of rising rising sea level. If you plant a thin strip of mangroves in front of an embankment. Will they will they survive and help grow the land or or not. Yeah, it's it's a. So the question is like the in the in the background where we have a sea level rise the planting mangroves how will they survive available less alive right. Yeah, so I think it's a very, it depends on locations we see so many mangrove restoration projects. Yeah, and it kind of kind of story like when you plant the long mangroves in the in the location you probably will kill that mangroves because like you know, as I see I said show like mangrove species that red, black and blue if you remember just so different mangroves they have different like preference on the inundation period. Right. Yeah, so for some if you plant like mangroves into a short period that definitely will kill mangroves. So, like the different that cornering strategy some mangrove will colonize very quickly and can help them to stabilize themselves very quickly and then eventually evolve to the seedling and and young mangroves while some mangrove can be easily be fresh away by the strong title current so it's a lot of uncertainty going around so it depends on the site but I would say, yeah, we have, we have since a lot of failures happening in the mangrove restoration sites. So you're going to be careful about that. Yeah, you got to choose which mangroves very carefully. Yeah, and also dependent on the local condition. Yeah. Right. Yeah. Are there any other questions? If not, I have one question now. Okay, so your model kind of takes into account slow, I called a slow hazard so sea level rise right it's slowly propagating it slowly getting you know your your sea level getting increased, potentially affecting your your mangrove forest or mangrove trees. What about, you know, short duration high impact like cyclones would so in real world mangroves are mostly positioned around, well, around a large band around equator. That's also where cyclones take place right. So would a cyclone damage mangrove forest such that it would kill them or would it actually strengthen the mangrove trees or forest by by affecting them but then then give them a new boost to recover. Yeah, yeah. So I'm wondering if that would how that would impact actually your simulations. Yeah, I agree. So the cyclone impact cycle on the mangrove forest. It's actually like happen a lot if especially like what you say that in the tropical subsoil area it's a very we say cyclone active regions and they will also like kill mangroves at some point because like you have very strong waves flow currents like can be easy fresh away or directly practice mangroves. Yeah. So, but I mean like this is a very short term impact the study I'm showing is like the long term how mangrove can gradually adapt to sea level rise of course like in the short impact you have suddenly mangrove be fresh away by saying the case during cyclones. But at some point mangrove will, if the condition is allowed, the mangrove will start to recolonize and and of course like dependent on the we say the windows, like if you allow mangrove to grow to the similar state to the similar mature state or you have another cyclone to coming up again to remove everything away. Right. So, a very similar example in Australia in the north Australian they got an event that suddenly had got like tons of mangrove was dying. I think the whole coastal mangrove was was was was died during the very large trout event instead of the cyclone. But the people think about maybe mangrove will recolonize in this in this location. And I have a master student working on this and we actually find it's not that easy because like the frequently cyclone occurs in lead regions would kind of like to say stop mangrove to recolonize because as soon as you have a baby mangrove colonized layer, the cyclone comes will destroy this mangrove scan so it make it bit it's a bit hard for mangrove to recolonize. But of course it's again it depends on the locations in some location for example you read you the mangrove was died, and the cyclone is not like recurrent. It's, it's a possible for mangrove to regrow and become the mature again. Yeah. Wonderful. Thank you. If there are no other questions. I think we can close this this webinar session. I want to.