 Okay, we're coming in this like next plus or minus or like our plus a little bit. We have an exciting new part of the CSD mass meeting. And this is a combination of the spring school that we just taught in a week before the CSD mass meeting. We had 22 participants here and focused, we're focused on cyber training skills, basic GitHub, basic Python, learning about these tools that CSD mass puts together and the community puts together. And one of the things that people then do as part of this course is team projects where they try to apply what they just learned and like take advantage of tools that are already there. All of them are scientists. They come with different backgrounds and different topical interests. And you'll see that through today, but we'll have like each team up on stage and like show us how far they got in their team projects. The idea is that these are contained projects. So we tried to like beforehand define or like as a team they defined an idea of like this would be something that students could use as an to introduce a concept or like a fairly like exploratory first step, but they are presented as notebooks or teaching modules that later on people in the community can use. There are five different groups. The first group consists of Alex Marina, Bunti and Jed. And they all were interested in long-term landscape evolution or tectonic processes were more deeper processes and river incision, etc. They have a modular course set up. So they'll like each walk you through like one of the like a sort of process modules that they have made available through a new notebook. So I'll hand it to Alex. There is enough time for people to ask questions. So like be aware like we'll have time for questions and for like interactions. Hi, everyone. I'm Alex. I'm a final year PhD student at Imperial College. So just to introduce what we what we're sort of aiming for and what we've produced. So we're basically aiming to produce like a teaching resource to allow students to develop some intuition on that long-term landscape evolution. Okay. So the main thing we're interested in allowing students to understand where the impacts on different erosional parameters you might use. So erodibility diffusivity, spatial patterns of uplift and how that impacts long-term steady state landscape, the impact of our graphic rainfall and then using the new little layers package. The impact of differential rock strength and then how some of those things might be recorded in provenance records. So just to sort of give everyone a primer. We are a sort of the model set up where assuming it's just a square grid of 100 by 100 plummeters and all of the model of the fume diffusivity according to just a standard diffusion equation and then an effective effective landscape evolution model. So just like a primer and then I'll hand over to Jed. We're going to talk about erosional parameters. Yeah, so kind of as the first step and this is also just as much for us to kind of get a grasp of the right range of variable values to use in the model. We ran the same model with all the same parameters except for changing between a range of five different diffusivities and KSP. So for rock erodibility, just start looking at how the kind of base land lab model of again fastcape eroder and linear diffusion results in different landscapes. So we can see here those different combinations of erodibility and diffusivity and how that and the main parameter that we extracted. Drainage density just for seeing again, building kind of an intuitive sense of how these change the overall form of the landscape so can see moving from very high erodibility and low diffusivity where we get much more dissected landscape landscapes with high drainage density and then moving into very high diffusivity with lower erodibility where we get more subdued landscapes and overall lower relief over the course of and these were all run for the same number of time steps in the model. And I think I will then turn it over to. Okay, so. As Alex said, uplift rate is one of the factor control the evolution of the surface right and. Different assumptions result in different circuits landscape evolution and in this model I show some of the basic assumptions of uplift rate when first we think about like the constant uplift rates and then the second binary when they are to uplift rate different at the top and at the bottom and in the third case when we use the linear functions like uplift rate is related to like the linear function with the y axis and the fourth one don't share when I assume the public functions between uplift rate and y axis and the last one, the more fancy one when we adopt the components in land lab called normal fault to control the uplift and you can see that when we use different assumptions, the result is different topography and also different drainage area and inside the models of student can. Try to use different assumption and understand like the results and okay. So the next component was the our graphic precipitation component so here we wanted to kind of explore how to incorporate different precipitation patterns into the land lab framework so in order to do that you need to add a water flux into the model. That then will go into a discharge component taken by the fast keep a rotor and algorithm. So, and we device for scenarios so the first one is just a binary one where the top have has a higher precipitation than the, the lower half of the grid. Then we did a linear components where it's just linear with respect to the y axis values. And following on that we have as the orographic kind of more orographic scenarios so the simple case here is just the precipitation scales with the with the elevation values, they're normalized to fit into the range of precipitation values that we would expect. And I took Boulder as the reference place because we're in Boulder. And then the final one which was the most complicated one was them. I, you know, I could have done, you know, solve all the physics equations for the orography that would have taken forever. So I didn't want to do that. So what can I do. I was like, okay, well, I need to include some sort of proportionality with the elevation. Then I took the maximum point of any north-south transect and decided that every point north of that maximum value would would have a higher precipitation value that all the points south of the maximum point at each north-south transect. And then I assumed that the wind that it would be a system created by a wind coming from the north. So it would, the north-worth side would get more precipitation at the southern side. And it, it's basis assumptions is not great model, but it kind of fits the what we would expect, you know, the northern side having more precipitation and scales with precipitation with elevation and then the southern side having less precipitation. So it's just kind of encouraging students as well to create their own modules and to show that it's easy enough to make new functions and to integrate those functions into the land lab kind of main framework. And just as like a final synthesis chapter, we, oh, yeah, we also use the litholayers package to explore the impact of, if you sort of are uplifting lithological layers with different strengths, the impact that might have on the total sediment flux. And then by tracking what rocks are being eroded, you can also sort of generate the synthetic provenance record through time. So I'll just skip over that quickly and just sort of describe the, there's a final synthesis chapter where we're trying to put all this together to sort of mimic a real landscape. So here we have both like a domal uplift stratigraphically led rocks of different strengths and orographic precipitation. And we explore how that evolved through time. So here we just have the rock strength broadly increasing with depth, but a little bit of randomness to simulate sort of natural variability. And then we just uplift like an anticline. And then just going through time, we can see different rock types being exposed to the core and topography evolving. And because of the orographic effect, it's sort of asymmetrical. And then gradually we build a more complex landscape using different rock strength, variable rainfall, yeah, all in land lab. So the idea is that by the end of the course a student is able to build sort of quite complex things having built it up module by module. So yeah, happy to take any questions. Very, very nice work. So just out of curiosity, are we these these vertical things that we see stick out in the south? Are they then the next step? Are they diffused away or what? So actually with if we haven't quite worked out how to integrate the little lab with the fusion, because the you'd need to have that basically like a particle tracking system with the little package of the perfectly like like grid based and you can't assume a lateral transport. So actually, this isn't technically this doesn't use diffusion in this particular model. But yeah, in the real world, they would diffuse out and soften at the time. I guess they're like equivalent to the flat lines. Thanks. Get up. So the URL here, if you want to use them, they need a bit of polishing. So we're sequences, the teams like totally choose their own topics, right? We sequence them a little bit in a source to sync order. This was very arbitrary, but like that's how I propose to do it. So the next group focus on upland landscapes and was looking at like susceptibility for landslides. And they're going to show their codes that is for landslides and failures. So this was team landslides susceptibility. Anybody who is going to speak for them. Scott, you're the fires, fires and landslides, or are you then still combining the fires and landslides. No. So this is Joe, Nick, Sebastian and Emily is over there. And David. Yes. So hi, I'm Sebastian. We are very excited to present. Oh, sorry. I'm Sebastian. We are very excited to present our very small project. Well, we all know that climate change has a great impact on the occurrence and intensity of extreme events. So for instance, fires, and also extreme rainfall events and landslide. And so we were very curious to see how the combined effect of fires, extreme fires and extreme rainfall events and landslide would result. So we made explosives by combining land lab components in the smaller notebook. Everybody I'm Nick. So, as Sebastian said, we were curious about how wild finder wildfires changing soil, cohesive cohesiveness, cohesivity, and increasing erodibility would impact the occurrence and size of landslides and then as sediment flux as well. So we used the, well we used a stochastic fire generator created by another CSDMS team that we modified for our model here. And as well as the highlands bedrock landslider component and the space sediment transportation component. So here we're just initializing some of our steps and creating the burner function. So the way the fire generator works is choosing a random grid cell from the DM that we load and a random fire radius, and then any cell or grid cell that falls within that radius has its erodibility boosted by a set factor. So here we initialize our components and use the BMI topography component to load a 30 meter DM of the Oregon coast range which we chose because it's a well studied area, a lot of parameters that we could find and include in our model, as opposed to just choosing something. So this is the DM that we started with. You can see there's a well developed drainage area with some pretty steep slopes. And here we set a bunch of the parameters for our model. So we're using a time step of one year and simulating 10 years of fires with a high, a higher currents interval of the fires at about one year. Our soil properties. We had an initial soil depth of half a meter for the coastal range. And then our boundary conditions allow for flow out of the sides of the models. That was a choice we made to simplify the model process. So here is our fire function running as you can see, um, seven fires occur over the 10 year simulation, and it ran for about a minute and a half so fairly quickly on this DM. And here is a plot of the fire locations and the erodibility change after each fire so each fire centered around or at one of the centers of the circles, and then basically multiplies the erodibility of that cell. Or all of all the cells within the fire radius. This is the sediment flux after the fires. As you can see it mostly follows the major drainage pattern up to the north, but there are a few going out to the sides here. And then after this we run the bedrock landslider and space components to sort of move some sediment around and see how that will affect our sediment flux. In this image, we have areas of deposition in green to blue and areas of landslide occurrence in the warmer colors. And as you can see, we've had quite a few different events move some soil around. And then again, we plot our sediment flux and it looks very similar to our first plot but some of the channels are a little more defined and if you were to review the scale bar here, you can see that there's a lot more sediment being mobilized after the landslides have occurred. So that's what we modeled this model did not take into account recovery of vegetation did not take into account recovery of vegetation due to, you know, that would increase soil cohesivity or sort of a proxy for erodibility over time. And we didn't have the model recreate more soil because one it's over a 10 year period which is not really a time scale for bedrock to become soil. And I think that covered everything. Thank you. Questions. For the fire function you have a perfect circle kind of where the impact is right. What you know is, can the fire function take into account wind direction or kind of topographic boundaries that it counts. Sometimes fires are stopped by riches or anything. Can you tell a little bit more about that. So, I'm sure that could be incorporated but our model was very simplistic in the location of fire. Just to stand next to you Dan. Just from my understanding is it so that the material that's deposited by the landslide that you give that a different erodibility and that's why you you increase the sediment flux. Yes, so the erodibility is increased by the fire and basically modifies the parameters that the bedrock landslider uses to move sediment. We didn't change the erodibility of landslided material, but increasing the erodibility increases the number of landslides. Okay. So this component is made to simulate deep seated landslides, but we found it more user friendly and fit our needs better for our shallow seated landslide so we used a soil depth of half a meter. And the tool can do deeper seats landslides, but we didn't utilize it in that way. Just to follow up on Danos question is the mechanism that increases the flux farther downstream from the landslides the fact that you put a lot of material in the upper parts of the creek profile and steep in it or. Yes, so basically we've moved extra material into the stream channels, which we can see here on the green depositional areas are largely at the bottoms of the slopes, which is increased soil thickness in the channels which the space component moves to the outlet. So it's, it's not transport limited, but limited by how much stuff is there in the creek bed. I don't think we took that into account. Thank you. So the fire model do you have a propagation of the fire from the source of the, I mean the initial ignition point. No, we just set a radius and have everything within the radius increase in erodibility by a set factor. It's kind of interesting how like everybody's minds, like straight away goes through like all these like more complex add on like functionality. Right, like, I think I literally when we set up the team project and told people, I told them like we in the course itself there's only like 10 hours or so that you like literally doing team projects but there was a weekend and people crammed through a weekend a little bit more and I think So, this group is also like more in upland environment and basically they started as like abrupt event was a really large group of people and we thought of some of the same processes but with a different application. And so it's my name is Archana I'm a third year PhD student at Montclair State in New Jersey. And today our group is going to be talking about using the land lab toolkit to simulate sediment flux after force fires. And so briefly force fire intensity is increasing with climate change, especially in the western United States. Fire season is lengthened due to warmer springs, longer dryer season and burn area which then impacts the soil and vegetation making a dryer. There is less fire, but the events that happen are larger and more catastrophic and we also know that sediment flux increases after fire events. And so for our project we wanted to ask this question and figure out how might climate change in two different scenarios, a medium emission scenario and in extreme emission scenario. How would that sort of general scenario effects sediment flux after force fires. Hello everyone I'm Scott the anima 25th year PhD student at the University of Nevada. And so our, our kind of methods here are we're going to set up a GitHub repository as we've kind of looked at here we're not going to go into the details just we have it in a presentation format, but that all this stuff we can learn in this past week and a half. As our China said we're going to have two precipitation scenarios and we're really just simulating for 100 years to kind of see how these are going to affect sediment flux throughout our our model domain or our landscape. And then we also developed a stochastic fire model we didn't use the one that was currently in land lab because circles are hard and we like squares as you'll see. And then we're going to use this stream power with alluvium conservation the space and entrainment the space model, Sam absolutely crushed it on getting that up and running. And so we're going to have increasing scale, increasing chaos and that's going to just scale with our fire. It's a binary switches and on off, as we'll look at. And then we're going to just use these output radiuses have some relationship with cohesion, and then we're going to run stochastic precipitation across that fire across our model. So this is what our fire distributions are going to look like and so on our y axis here this is just number of fires it's just a distribution that we're sampling from. And then on our x axis here we just have our fire size proportional to our grids is just a length relative to what our model domain is. And so we've got these these small these these these moderate scenarios, and then we've got these large catastrophic wildfires. So an important bit is this is what our fires are going to look like in model space where in this, this more frequent smaller fires we only have five per our time step, and they're going to be small and then for our, our less frequent but large catastrophic wildfires, we're just going to but as you can see they kind of take up a huge, huge area within our model. And then, so we chose to climate scenarios, and as as Marina put it we chose boulder because we're in boulder. And we've got this climate one climate to fortunately they're they're roughly the same but we're going to still sample from these stochastically, and just kind of look at primarily the effect of fire over time. I'm Sam. I'm a master student at West Virginia University. Okay, so where did we run this model, you may ask. The watershed is really fake. And what we did is we made a like kind of tilted with some random noise and then we ran a stream power model over that to give us a drainage network. One kilometer by one kilometer, you can see in the right corner. This is the hillshade of this watershed. So then we put this in a raster model grid, a land lab grid function. Sorry. We added so that's a bedrock elevation and then we added soil depth to that we just added two meters arbitrarily. So here's where we instantiated all of our land lab components. We use linear diffusion, just for simplicity, and then like Scott said before space. Here's the time loop that we ran. So you can see for each loop we choose a random precipitation from our climate data, and then we also update the K values, each time from the stochastic fire model. So here's what we found. So this is the control there's no fires in this. So we can see that sediment flux kind of makes sense. It's where the drainage area accumulates, just to make sure it all works. So here's one of the model runs with fire so the red squares are fire. And you can see, sorry, easy gift maker wasn't working this morning. So we're doing this. But you can see where it kind of increases sediment flux and if you notice the color bar, it actually increases that quite significantly. So here's just another fire run. So yeah, the watershed with the most fire kind of has the most sediment flux, which is cool. And here's one of our larger fires. I guess it's actually more than one larger fire. But yeah, notice the color bar again, just the maximum sediment flux is pretty significantly larger than the smaller fires. So what does this look like with precipitation, I guess, okay, this is it without precipitation. So fire is on the bottom. And you can see that sediment flux tracks pretty well with fire. And then when we add precipitation, we lose that effect a little bit, because we see that precipitation is actually pretty important for the sediment flux. And we kind of lose the effects of the fire, which was an interesting result. So here's the same thing with larger fire. So just notice the middle plot we have on the y axis, the percent of watershed impacted by fire goes up to like 50%. In a similar result, we didn't do any statistical analysis, but I think it's pretty clear that a lot of the precipitation trends are reflected in the sediment flux and the fire trends really aren't. Yeah, so clearly this was a simplified model. And so for our future work, we really do need as someone asked earlier, we need more realistic fire habit. And yeah, like I said, squares are nice, but incorporating elevation vegetation, wind direction, all those kinds of things we think would would work quite nicely. We also need to better constrain the relationship between our fire magnitude cohesion, and then our changes in our chaos or our roadability value. And I think that it was just an on off switch, but it would be nice to kind of to to trend that with our fire growth and actually scale it properly. And then we also want to apply this to a real location. We need a real DM real fryer projections and then a little more complex climate data. And then we also want to compare it to actual measure sediment flux going forward. These are references and we also just wanted to thank our instructors for this week. This was really fun as we can all attest so maybe give them all a round of applause because it's been really great. Any questions. Thank you so much for your experience fire by how much did you change the roadability order of magnitude no idea that's right. Gotcha. So if I understand it well that the precipitation had an impact on the sediment load because it could transport it right, but does the precipitation also have an impact on the fire will it, you know slow down the fire or or, you know, stop the fire or no. So how we how we did it was we just ran a couple, ran a fire scenario, gotten initialized landscape and then gave it a precipitation and saw it out to the fact that's a great point. So you said that you thought that the impact of the fires was not that great on the sediment load precipitation was the greater impact. Why do you think that is why are fires not having a bigger impact. Maybe, because we only changed the K value by an order of magnitude, we didn't really play around that much. If, if only everyone had seen the like fire talks at the meeting before getting started on these projects right. I mean, one thing that I find interesting. I mean, some of the questions that you asked today are questions that could be added in like sort of a discussion topic for students to like ponder at the end of like doing these simplified experiments and say like, what do you really think that is like an additional complexity or what if we would and do in hypothetical on that. Like each of these are in the end like presented for like students to be used or learners to be used or like as an example for a set of like simple notebooks to build upon so we have a whole bunch of people who are interested in coastal behavior. And they took not out of the landlap component set but they worked with a model that's called pie delta RCM and the two developers or the two active developers that are working on that model are here. So like they'll get like a surprise and like may have some feedback. Or it can be super proud because their model was accessible enough and easy and UV enough and documented enough that other people could start like tinkering with it right so So that's a group that worked on vegetation effects on coastal systems. Elena is a more Liz Kelly and Lexi are presenting. Hi, yes, my name is live, and Lexi and I will be doing the presentation and is more and Kelly will be filling your questions. So today our talk is titled exploring the interaction of salt marsh and mangrove vegetation on river delta evolution. So we know that vegetation influences delta morphology and evolution with so our research question is does a delta that grows with mangrove vegetation evolved differently than a delta that grows with salt marsh vegetation. This is motivated because vegetation eco to eco tones are shifting due to climate change. So it's likely that coastal areas currently occupied by salt marshes may transition to mangrove. An example of this would be south Florida. So in this image you can see that currently mangroves dominate southern Florida, but we think that possibly they will start shifting north where salt marshes are in the north. To address this question, we use the delta RCM vegetation model. So the vegetation model was built firstly on the delta RCM model. So the delta RCM model is the delta river delta formation and evolution model with channel dynamics. The vegetation component was expanded to include vegetation effects. So in this model that was published in 2018 vegetation colonizes grows and dies. This increases a bank stability and improves resistance of flow and the original parameters for vegetation such as stem diameter carrying capacity, logistic growth rate and rooting depth represent marsh grass grass type plants. So what changes do we make to this model. So the parameters are now defined by a gamma file. We converted X range to range because the original model was written in Python two and wanted to run in Python three. And we use mangrove parameters for comparison to salt marshes. So in this table you can see that we changed some of the parameters from thought marsh like some diameter from point zero zero six meters to point one four. So the amount of the capacity is much different between salt marshes and mangroves as well as the rooting depth. I'm going to pass it over to you. Yep, so I will be talking about our Jupiter notebook and how we kind of set it up. So at the very beginning, we didn't have pygamel installed so we had to install it. And then our next statement just imports, YAML which we use to import the YAML file and image IO which we use to create the GIF. It's really helpful for you. And we import the delta RCM class from the Python file. And then the next line of code, we actually import the YAML file and what that looks like, at least our YAML file has all these parameters in it. Which these were originally defined within the Python file itself so we just made that easier for users to change their own values and we decreased things like the water particles and sediment particles just to make it run faster. And our total time step is 5000 which represents about 150 years. So then our second to last block is just setting up the model so instantiating the model with our parameters and changing our values for either salt marsh or mangrove like what Liz was saying from our table, and then running the model, which takes very long time. And then our last block of code just creates the GIF files. So here's one example of out like just one of our simulations which was a mangrove simulation. And the top left graph shows the elevation so the white is your elevation and then the darker it is the more negative it is. So our top right graph shows the vegetation, which is from zero to one based on a density of vegetation density. The bottom left is discharge and the bottom right is water surface elevation. And so we we decided to save time steps at approximately 25 year intervals. So that's what you're seeing here is the evolution of 25 year intervals over 150 years for a mangrove run. And so kind of transition. This is what the four different scenarios we use kind of based on the original paper parameters. And this is specifically for salt marshes. So the way this works is that from top to bottom, we have a decreasing flood frequency. So what that means is the 75 is 75 time steps in between a flood. So 100 time steps between a flood is less frequent than 75 time steps between the flood. And then on the x axis, we have the sand percentage. So 25 is 25% sand 75% mud, and the right is 75% mud 25% sand. And really just what we want to notice is that the flood frequency didn't really seem to impact the vegetation or growth that much they're pretty similar. And then the like in the marshes we have a lot of small channels forming off the main channel. Whereas when we go to our mangroves you can see there's a lot less kind of like white space in the vegetation so we saw that there were a lot less channels forming. But these are again the the same 25% 75% sand and flood frequency. And we also notice that when we did the mangrove vegetation for 25% sand we had a lot less colonization around the entire semi circle. And then we're going to do further analysis using the top left scenario. So that is the 25% sand and 100 day flood, which Liz would talk about next. So yeah we compare those top two left plots, which were the base conditions in the original model. Okay, sorry. Yeah, so I hope you got that. So if we compare the results between the salt marsh and the mangrove deltas between the vegetation and the mean free surface elevation we can see that there are differences. So in the first plot on the left, we have time in years on the x axis and vegetation cover in pixels on the y. The salt marsh and mangrove vegetation start out a little bit different, but they do converge around the 60 year mark and then slightly diverge again at the 150. In comparison, the on the plot on the right we have again time in years on the x axis and mean free surface elevation in meters on the y. And again the salt marsh and mangrove vegetation deltas start differently, but instead of converging the differences only grow. And so if we can tie this back into our research question, does a delta that grows with mangrove vegetation evolved differently than a delta that grows with salt marsh vegetation. We found that yes, we found spatial and temporal differences between vegetation cover free surface elevation and as well as the other variables. So these are key findings and implications. First that we were able to simulate delta growth under different dominant vegetation conditions using the delta RCM vegetation model. Secondly, we found spatial and temporal differences in delta evolution between the vegetation types. It's important because the evolution of coastal areas under different types of vegetation is important for planning and management of coastal areas as well as thinking about all those ecosystem services that mangroves and salt marshes provide. So with that, they will be happy to take your questions. Great job. My question is thinking about your plot that showed the differences in elevation between salt marshes and the mangrove simulations has that been observed like in the real world, do we know. Hi. That's a really good question. So it really. So I guess I'll also answer it a little bit with the lateral accretion, because mangroves do accrete laterally and so you do see those increases they're known as land builders right so they both go in X and the y direction is the way that I like to explain it to people. So those differences. I'm not so sure off the top of my head in terms of salt marshes, but I do know based on your comparing trees and the tree root systems I wish we had a picture of a red mangrove, because I can explain a lot better. But you can, you can think of red mangrove root systems as kind of like spider claws is what I tell people, and or spider legs and so they're able to trap sediment but also the organics come into play and they're able to. Their leaves fall and kind of degrade within their mud systems, they're able to trap those leaves and it doesn't really leave the system so they kind of do accrete up as well. So you do see that sort of but I'm not too sure. Yeah, I can expand on that as well so this free surface elevation. The differences there are really related to the vegetation properties itself so like if you think of a salt marsh that it's like really high density and there's like lots of plants. They're everywhere, whereas the mangroves are trees so they're like a little bit more spread out. So what that graph is showing is not like. It's really related to like how the water is like interacting with the vegetation itself so more of like a hydrodynamic impact, but this model has yet like a smart was saying has no vertical accretion for either the salt marsh or the mangrove, which would be really important to include in a model of Delta growth because typically in like a salt marsh that organic sediment that Ismar was explaining makes up like between like 30 to 80% of the material that's actually on the platform. Depending on the mineral sediment loads those areas so yeah it's a bit more complex than that this graph shows I would say. Super cool. It's fascinating work I think you guys should write a paper. But I'm really wondering about how the water surface elevation is higher in the mangrove cases that you were alluding to the hydrodynamic properties but I would have guessed, since there's so much more rarefied that they'd be less resistance to the flow and so the flow would be less effectively channelized when it looks like it's more. I'm wondering if the deeper roots and the fact that it makes it harder to kill those. And so they're more on a longer time scale or morpho dynamically confining the flow more than hydrodynamically. What do you guys think. Yeah, I would I would agree with you. Yeah, I have. I can talk about this all day. So in the Caribbean specifically just off the top of my head we have four different species of mangroves right. But the main to that a lot of people focus are on our red and black mangroves that have completely different types of root systems right. And so that's not really simulated in the pie Delta RC and veg model, but depending on the root system that you have the type of mangrove in terms of which type of mangrove you have, you're going to get different types of channelization, and also the hydrodynamics of the system so depending on how much salinity for example or soil stressors. You have there's going to be different kinds of mangroves and different parts of the delta egg region right and so those root systems really do affect the channelization. And as I explained with the spider leg analogy, like mangroves have proper roots which I explain as like forks sort of. And so they trap sediment, sand sediment completely differently than red mangroves would. So channelization is different depending on the species as well. I just want to make you run. Awesome talk. I have a question about so at the beginning you talked about how in some areas you know mangroves are might be overtaking where salt marshes used to be. And so you ran these as separate separate model runs. And so I don't know if you had any comments about how, you know, if use a mix of properties how the delta changes or if you start with like you run the mangrove simulations but start with the marsh salt marsh elevation and symmetry how if you thought about that at all. Yeah, that was actually like our initial goal was to have the mangroves like encroach on the salt marsh in the model itself. But we just ran out of time. But yeah, I think that would be really interesting. Awesome, given this is given the time this was an awesome project. Our last team also focused on delta processes. And it shows you one of the merging conflicts in our community where they use a different version of pie delta rcm the more updated version that did not have yet vegetation characteristics so like got to do something. This team. Include a 10. Donna Lawrence, Katie and Juliana are going to talk about delta rcm and delta processes. So we're just going to introduce ourselves real quick. My name is Caitlin or Katie Turner and I'm at Louisiana State University and I'm a PhD student. I'm Julian Davis I'm at UNC Chapel Hill and also a PhD student. Hi, I'm Lawrence Willis I'm a PhD student at UC Irvine and also a PhD student. Hi everyone I'm Jenna left it up to you. I'm a master's student at UT Austin. And Tian Dong who's a postdoc at UT Austin who had to leave early on here with us today contributed quite a bit to this project. And then we would also like to thank Ethan, Josie and Mark for all of their help for getting helping us and arena for helping us get this model to run. So, one of the main things we really wanted to do was connect real life situations that are happening to our Jupiter notebook. So we used wax like Delta as kind of a case study using the pie delta rcm model. So one of the things that coastal Louisiana is really experiencing right now is land loss, and that is due to a range of factors such as subsidence sea level rise and as well as this lack of sediment that is happening partially can be due to the upstream damming that's been occurring in the Mississippi watershed. And so we actually experience in the wax like Delta Mississippi area, 50% less sediment rates than their early historic values. So this is definitely due to damn trapping trapping, trapping sediment. And the other part that's going on is when we have the a chaff ally river having this kind of pipeline of levees, we're not getting sediment going into the surrounding areas which is why, if you've ever looked at the Louisiana coastal master plan there are a lot of diversions coming in. And so this was kind of our, our inspiration for this project. And so you can actually see with wax like Delta, it was created it was a man made flood diversion. And so you can see the growth of this delta over time. And so we wanted to see how if we change sediment, how is this going to impact delta delta growth. And so the first part of this was to characterize stream flow so we wanted learners to understand what, what is happening upstream. So our first part is looking at using a Python package called data retrieval which takes USGS stream flow and showing them different ways to calculate the flow frequency curve, as well as the exceedance probability and intermittency factor, because the intermittency factor really tells us how long we're going to have flood waters that will contribute to growth. So it walks learners through what all of these processes are and what they mean, as well as how to plot them. And so we can see a few of the plots that they'll be able to replicate themselves. And yeah. Cool. Yeah, so then there's some educators, educational questions and comments here like, can you comment like what the intermittency factor white meat might mean etc. And then we wanted the students to use a more for dynamic model to basically simulate delta growth. And in this context, PI delta RCM which is reduced complexity model for delta growth was the ideal candidate because it can run within the Jupiter notebook and basically build a delta in a couple of hours. So actually the developers are here. And it's available through land lab. So we basically set it up in this notebook and then designed a couple of experiments that the students would be able to play with basically different set of characteristics corresponding to different real world damming scenarios and Julian's going to walk through that. Finally, we wanted to simulate different effects that dams can have dams can take a range of forms but here we're kind of just reducing the complexity of the problem to think about an earthen dam, like you see on the right that doesn't allow any water or sediment through, and then a dam that has a slew skate allowing some water suspended sediment and bed load to continue moving downstream, like you see in the image on the left. The table below shows the scenarios that we have run. As Katie said it's inspired by the wax like Delta but we're not claiming to be reproducing that system. So we are looking at two different sediment concentrations which you see in the rose. Originally, a sediment supply of point three and then point one five. And then in the columns were changing the amount of sand versus mud so changing the sand fraction to see what would happen if bed load is reduced, and as you move to the right in this table, the delta is totally made out of finer grain sediments and mud with no sand present. We call those baseline and then a through a will be using that notation again so just as a logistical thing. We do take a bit of time to run as the other vegetation group said, which is why we have do not run this step in our educational module. We recommend this as a lab with a homework, the component where these students run the models overnight and do the hydrograph things on day one day to look at their output. If you don't have that much time we have also uploaded the results of our modeling that students can look at. So we then ask them to follow through some of these steps, import the data and begin doing some different visualizations. This figure shows our outputs and those format is the same so the top left is our baseline. The bottom row has the initial sediment concentration. Bottom row sediment has been cut in half, and as you move left to right, the amount of sand decreases until you only have mud, all the way on the right. This shows delta elevation and values in light green to yellow are elevations above the water surface. We ask the students to make different observations about what they see in these deltas. What we encourage them to look at is the frequency of different elevations that they observe like which deltas are more effective at building upwards versus outwards, and that's what these histograms are showing. We then ask them to look at bifurcations and the number of channels. So again here, we are showing those six simulations and the colors here are flow velocity, higher velocities in yellows. And we ask the learners to compare these different deltas, think about how the changes in sediment could result in the changes in delta morphology that they see, and then make comparisons over time. These show the deltas at the end of the model runs. And here, we look quite a few time steps earlier, but smaller deltas and ask the students to kind of think about how these deltas grow and change through time and how that varies with sediment characteristics. We also invite the learners to make some predictions about other parameters they could change in this model. How could they simulate subsidence or sea level rise. We ask them to the documentation and encourage exploration of other ways that they could change the deltas that they can make in pi delta RCM. I'll pass it back to Katie to wrap it up. No, we're good. So part of this learning component is to link it to like the broader impacts as well. And we finalize it with just like a general discussion section where they can look at the broader impact of the entire watershed. This is just an image of the Louisiana or Mississippi River watershed area, and we're showcasing the cascading effects of dams along the watershed, where resources are extracted all throughout. We also provide a link to the master plan for the Louisiana coast, where they can look at restoration projects that are currently active, and the different scenarios that they can use such models to kind of forecast for the master plan also like reduction in budgets and mitigation efforts as well. So we want to like invoke them to see other broader impacts that link science to society as well. And that's pretty much our model. Any questions. Thank you so much for your work and building this really cool educational module. And, of course, being up in Minnesota, I appreciate doing on the Mississippi. And I was wondering if you guys have any insights into how much the locks and dams in the upper Mississippi actually affect the settlement supplied to the lower river considering the sharp dropping gradient to the lower Mississippi and the large amount of sediment buffered in the valley itself. There's actually a reduction of about 50% of historical values, which if you're thinking about land building, that's, that's a lot. And so it's starting to cause, we have other issues down in Louisiana, I go to LSU so I get to study it a lot. And so subsidence is also part of our issue and sea level rise, but it used to be that the sediment could actually help counteract that but because we don't have those same sediment, it's actually getting more difficult. So even when those diversions go in, it is going to be kind of a diversion game of, okay, we can open this one but we can't leave it open too long because it'll create shoaling on this channel. So it's still, it's the diversion network that they're creating down there is incredible. And so it's really kind of stemming from this issue that there is a lot less sediment. If you had another day or two. Where would this go. So I personally think it would be cool to see how the stratigraphy changes. So if there's a way to like incorporate more grain size variations, see how the stratigraphy would look like in a vertical sense, while the delta is building up. So sometimes we see the delta is pro grading versus a grading. And I think it would be cool to explore that aspect. As other groups have said we also had, you know, an original idea and then what was feasible in the time given. So something we talked about originally would be running a delta for a certain amount of time and then starting to modify the characteristics, instead of starting with an empty basin. But we think that would require modifying the source code and that was a bit beyond what we felt we could get done and present. So I think that's another direction we'd like to take this. Did you want to add more. Okay. What do you envision as the age level for the learners. So we were actually given an approximate age level which is helpful. And it was more advanced undergraduates and early graduate students. I was thinking about this it really seems like a hydrology class or geomorphology class because it really links those two together and a lot of the times if you take a class and say civil engineering and then one in geomorphology, you can kind of miss those links so it was really for us to link it together. I want to thank everybody for such hard work that was done to like get to these projects. These participants know that you will be nagged by like the people from the integration facility a bit to like get this into the educational repository and there's like 100 other people that now know that these exist, right. So that may want to use it in classes or they may want to use the material like for their own like hydrology class geomorphology class etc. I think these are wonderful I think some of you like really showcased, like sort of the progression of how a student or a learner or like someone who pulled this from a repository would interact with this and have questions and things. I just focused a little focus a little bit about the contributions that you did make changes to code or like small improvements to like how users would interact and I just appreciate like how many in engineers creative like solutions you found for all these like obstacles that were there in the in the process of putting a notebook together like this.