 Okay. Thank you for having me here. So yeah, I'm Marco Tangia, postdoc researcher at Polytechnic Milano. A couple of words on my background, basically the group research group I'm coming from the Environmental Intention Lab. It actually works on strategic planning and management of entropic alteration on a reverse system. So we're talking about like large scale development of reservoirs accounting for multiple competing objectives. So typically energy production, flood protection, tourism, even malare prevention, something like that. And the idea behind my model was to go a step forward and add sediment connectivity disruption in between this objective, which now then can be accounted for strategic planning and management. So what is reverse sediment connectivity or disconnectivity? We can define it as the connected transfer of solid material between all the areas of erosion in a river system to the areas of the position. And we know that the natural river sediment disconnectivity is connected to the well to the health of the river system so the ecosystem belonging to the river, but also the availability of ecosystem goods and services provided by for human use. We also know that the introducing any kind of alteration on the river system, whether they be the construction of reservoir, so directly on the river system also changes on the catchment like land use chain, the forestation are bound to have many effect on the natural flow of sediment in the river system. And these effect in turn are responsible to a variety of often unpleasant effect on the river system which are felt across a large spatial and temporal scales. For example, loss on in the degradation of the ecosystem, both on flood banks and on river ecosystem, but also erosion of the coast coastal landscape because of sediment starvation, then feeling, etc. And this is a huge problem because one of the problem we have is that no river, especially large river, under a single type of alteration usually we have multiple alteration on the same river system that are difficult. And this cause the alteration of sediment disconnectivity to be to have an accumulative effect. So basically, it's difficult for us to understand the big pictures or what happened to sediment disconnectivity if we look at the single event or the single alteration. We can try to turn to models to help us to understand what happened to sediment disconnectivity, but we can use it all, but also to increase our knowledge on the river system itself. But the problem of these models is that since reverse sediment disconnectivity is such a distributed and time varying property, it should keep a wide spatial temporal scale in order to account for the, what happened to the entire river system and all the different sources and things in the river system. It might also must be flexible in order to be used also in data scars environment because the majority of the alteration we found nowadays often in large river system where we don't have enough data to apply very complex models. We also must be able to run quite fast. This is because to account for the uncertainty modeling this kind of this kind of processes and also to account for the different portfolio alteration we need very fast reliable models. It also must be able to determine the causes and effect of the alteration order to understand why this alteration happen and what what is causing it. And now, if you look at the available model we have now this is a gross simplification but there with me for a second. So traditional morphodynamic model so 2D 3D base model are a little bit are very good and predicting what happened a single stretch of the system but for the scale we're working with are not very effective considering also that requires a lot of data. The 3D base model are still good and very effective in this kind of works but sometimes we need something else runs a little bit faster and we even less data. So we can turn to conception numerical frameworks. These are exploratory tool that trace part of the information and data and the preciseness of the other two categories to have a faster more reliable more flexible framework. In particular the model of my model is based on the original cascade model, which was devised for the introduction of sediment disconnectivity as an objective for multi multi objective planning and management. The idea is to have a conceptual model that combines the concept of graph theory and empirical sediment transport formulas to have at least a picture of sediment connectivity at the base in scale. The cascade models then published then we publishes a freely available toolbox. The idea behind the cascade model is to first of all reproduce the river system as a 1D direct graph composed by reaches and nodes. You can see it over here. This is defined as the core modern unit in cascade is defined as a homogeneous same homogeneous stretch of the river system, which is characterized by a set of features, which are added by the user. Then we can add the different value, different contributor of sediment and different barriers on the river system. Then we describe the sediment transport as a combination of individual sediment transfer processes called cascades, each of one trace by their provenance and carrying up specific volume of material downstream. In this way we will we have information about the provenance of the material. And also if you look at the rich scale, the quantity of the material itself. In my research I went a step further by adding a dynamic component to the cascade model, the system model. The idea of the dynamic component and so to move from a static representation of sediment transport and sediment connectivity to a dynamic one was to was to be the idea was to be able to account for a logical variation. But also the morphodynamic evolution of the, of the rich features of a decade centuries, but also to be able to apply this model also for the management of anthropic alteration for example reservoir. So, to do so we nested the original cascade loop, which looked for each rich and river system inside a discrete daily time loop. And this also caused us to change the representation of the rich features from a static feature that doesn't change over time. That's what we had an original cascade model to a dynamic feature that changes over time both according to data given by the user but also data that we can obtain by via specific modeling components. So the idea behind the cascade loop is first of all to, to identify the mobilized sediment by using using empirical sediment transport equation. Then once we know how much material is being transported every in that particular time step, we can change the morphic feature accordingly if we want to if we have enough data to change them, and then we can deliver the sediment downstream using empirical sediment velocity equations. So here you see a simplification of what happened to inside decascading each time step. So here we have the, the rich number for the color represent the provenance of the material, we have the incoming cascade so the incoming volume. Then we calculate the transport capacity, and then we calculate a new volume that would be will be carried downstream to the neck for the next time step. The model is very simplistic representation of sediment transport. So one idea we had is to be able to add the specific modeling component that we have called add on component, which are component that can be that works at the rich level, and are able to account for processes are more difficult to represent in a 1D structure. And behind this component is still to be able to have a flexible modeling environment where if we have enough more data we can add more complexity to the model representation. And in general these are the component may receive as input in each time step, the data, the data that we obtained from the cascade for example the sediment delivery, and then can use this data to change morphological structures of the reaches for example you can imagine a 2D country component and changes the width or the gradient of the channel according to the sediment erosion or the sediment deposition. Here I'm presenting a case study the first case study we applied the cascade on the ideas was to have a case study which is quite large but very well monitored in order to see if the cascade was able to represent the material. So we actually took this case study in Australia the bigger case study. The bigger the bigger system was characterized by a massive hydro morphological shift after European settlement in the 1850. Basically, what European arrived in the 1850 completely cleared the land of vegetation, and they channelized and drained the result was, first of all, a massive channel expansion in the lower part of the river, and then a release of material from these former swamps that moved downstream as a sediment slug. So the idea was this. Can we apply the cascade on this case study apply the same drivers we observe on the on the field with the correct sequence of event and then see if the model is able to reproduce these changes. And if the first step is successful, can we then use the model to predict future changes in the river system. So we of course don't have daily data of water discharge for each time step for 150 years, but we can, but the bigger system is a flood dominated system where the majority of sediment movement happened during limited flooding events. So we decided just to simulate these flooding events where for which we have extensive record of. This comes with a lot of uncertainty still, because of course we need data for the whole river basin and not a single point. So we devise a four different discharge scenarios accounting for different events to cover a wide range of possible discharge condition. We also had to do two components that changes to other components that changes the channel with and according to the erosion of the sediment deposit. So that account for the fact that during a particular large event that some of the flood the flow can go over bank and so decrease the potential for sediment erosion in the bank in the in the river itself. So here you see the results on the x axis you see the time in years on the y axis you see the sediment storage in four different reaches in the lower part of the system. So let's see first of all that first we have the erosion of the original sediment volume, which is accompanied by a massive expansion of the channel. This is model but the point in black over here is actually are actually field data. And then you see the arrival of the sediments like, which is represented with a different color since we indicate we're able to trace the provenance of the material. And then try to validate this result using field data, for example, sediment deliberation in a particular section of the river or channel volume or the amount of material. So we're trying to match the black line with the different scenario of discharge. You can see that the patterns that we observe on the field are closely matched by the majority of the scenario we are simulating with. We try to use the result to see what happened will happen to our case study in the future so we devise one under scenarios of independent generator flood events without accounting for climate change for now. And we change the feature the characteristics of the river system. So for example we account for a scenario where exotic vegetation vegetation is kept in the river system and so the sediments like is being kept trapped in the river or we need another round of deforestation in the future. And you see how the sediment delivery changes and said we move and change over time in these two scenarios. To finish up I want to show you a new case study which is very different from the original bigger case study. It's a cost study we're working now we're trying to we're publishing this paper right now. The idea is, since now we're able on the with the case study on the bigger system to validate our model, we try to go a little step further and go to a state case study which is very different from the bigger system, which is the bigger system attribute area of the Mekong. This case study is very data scars. We have no information at all about the sediment delivery we have some limited information about the sediment yield in the catchment but they're very, very few data and very high uncertainty. And what we're trying to do is to apply the cascade in order to have to do strategic reservoir planning and management including also the use of drawdown sediment flushing to reduce the impact of sediment trapping caused by reservoir. So, just to show you this video over here that you will see, you will see the water discharge, the color in the rich will represent the amount of material transported in each time step. And here we have a situation without sediment without reservoir and with three larger reservoir in the lower part of the river system. So, if we start the video, you will see how the sediment movement is represented in the cascade with and without reservoir. Here we are, we are in a monsonic region. So, you see, here we are going approaching the dry season so no sediment movement is being registered. But if we move towards a new monsonic season you will see how the sediment changes over time especially during the first flooding events. Thank you for the attention. So, if you have any question, please ask away. Thank you. How do you account for the percent of the position you have at the dams, the efficiency of the dam? Okay, yes. On the reservoir is actually, I thought to go as fast as possible in this case study is actually more complex because we consider in this case study over here we also consider sediment deposition inside the reservoir. We have, we simply have, it's actually a little bit more complex. We represent the sediment, the water storage in the reservoir as divided into different component compartment, one for each flooded rich. And then we have the deposition coming inside the reservoir. We change the features of the rich according of the flooded rich, for example, the width, the depth, etc. Then we use, we still use empirical sediment transport equation, but since we change so much of the rich, rich feature that the fact will be stark the position of the material. You can imagine that we do once, for example, we do sediment flushing, we completely draw down the reservoir and then we rest, we reestablish river condition inside the reservoir and this will cause sediment movement to be to restart and then deliver the sediment downstream. So I don't know if I was, but if you have any question, again, I have a lot of slide on this idea and I just want to touch on this case, because I think it was very interesting to see the potential of this model. Again, the idea is to have a very, is the idea behind the cascade is to ask ourselves, what can we do in this case study where we have no data at all, but where they're still building a lot. And can we inform at least a little bit of the decision of the decision maker accounting for sediment transport, even if we don't have data to set up more complex and precise models.