 speaker that is Professor Erkan Iston Hulu-Logu. He is a professor at the Department of Civil and Environmental Engineering at the University of Washington. His research revolves around improving understanding and modeling interactions among hydrological, ecological and geomorphological processes in watersheds. He's been a long, lasting collaborator with like CSDMS and like been someone who's come to the meetings quite a bit, so many of you will know him and he's one of the PIs on the LandLab project too and sort of has been advocating for ecology and land use in the LandLab toolbox. Today he will talk to us about ecosystem processes and landscape evolution specifically, sticking with the theme of the meeting. Thank you Erkan, the floor is yours. Thank you Irana for a great introduction. So in this talk I'm going to address this question of linking ecosystem and earth surface models in three fundamental to me fundamental questions and the first question is that how do climate and topography control ecosystems and that is essentially how can we reproduce the observed patterns of vegetation on topography like the aspect patterns and whether it's corresponding different types of vegetations. The influence of flow convergence, flow control in rivers and the elevation control and these are done essentially predicted by distributed ecosystems and eco hydrology models in a fairly detailed way. There are some limitations of these models for example the disturbance, feedbacks like fires and insect that needs to be improved. There are competition, the competition, establishment, mortality of vegetation, dynamics has been limited and there are computational scalability but we kind of have the general structure and the capacity to be able to predict sort of the topography and climate controls on the vegetation patterns and the next question is that how do climate and ecosystems regulate geomorphic processes so that we can essentially link an ecosystem process and ecosystem state variables with the state variables of the landscape like elevation, soils and rock properties and so on and the properties that fit into this category is that the hill slope processes like bioturbation that includes tree fall, pocket golfer mounts and borrowing at animals and so on. There's landslides, there is the influence of vegetation on the flow properties and rain splash and sheet wash erosion and there has been a lot of work that essentially to quantify these processes and I'm going to categorize them into four different categories. There were implicit formulations that use the geomorphic transport laws that are used in the landscape evolution models. There have been mechanistic formulations that essentially integrated a state variable in the ecosystem domain with a state variable in the geomorphology domain within the realm of diffusive processes. There has been great literature on that and more recently and there has been a lot of work in fluid mechanics and eco hydraulics to study water and vegetation interactions flowing and vegetation interactions and they're having a lot of work in the mechanistic landslide models but if you look at these different models the first two is maybe long term processes and the second two are more like short term processes and that relates to the shorter term timescale so one challenge is that how to link all these in a landscape evolution model. The third question is what are the emerging outcomes, emerging patterns of a coupled ecogeomorphic system and I will highlight a few patterns that I can think of in one such pattern is that the influence of mean and precipitation and vegetation grow on slope gradient. There seems to be an increase in slope as vegetation becomes as biomass increases on the landscape in this first study from Jeffrey Adolf from the from the Central Andes. Similar type of behavior is observed in New Mexico as well in a lot of St. Mayer landscapes that with more vegetation with higher amount of biomass and root biomass on the landscape there seems to be a steeper slopes and more converging topographies and a similar pattern is also observed from the same elevation but across aspects seems to be a polar facing slopes with different type of vegetation maybe more vegetated polar facing slopes have steeper gradients than than the equatorial facing slopes. Another study this one is from the Golan Heights that shows that northern aspects have steeper gradients and they also have more vegetation. Another emerging property of this at the latitudinal scale at the global scale perhaps is the observation that this patterns are pervasive it pervades in the steeper polar and record this paper published by Polis that shows that northern facing slopes or polar facing slopes is always steeper between the 40 north and 40 south latitudes than the equatorial facing slopes so this is an emergent property. And another emergent property are at the pattern at the smaller scale at the pattern scale that's the pattern of pattern vegetation these includes bendy woody plants gap patterns labyrinthine patterns and spot patterns but the very interesting observation in the literature is that the models that are developed to study these patterns are very different than the cohydrology models and the dynamic vegetation models ecologists have used dynamical system models including reactive and vacuum diffusion models going all the way all the way back to Turing's initial work kernel based models and hybrid models that includes some of the two and what they found is that a simple principle relating a growth to lateral water transport explains the variety of self-organized patterns so it just seems like the water is the key thing in all of these patterns as well and the secondary influence was the plant facilitation and plant mediated changes in soil and nutrients have have that secondary role in these informing of these types of patterns so then looking at the landscape evolution at a various scales then how do we deal with model complexity so i'm going to divide this to a land diagram and i have landscape evolution models ecosystems models and hydrology models and then we look at the way these models have progressed over time and get more complicated the initial set of models have been more disciplinary the landscape evolution models large look at the geological time processes and there was there was not hydrology in them and similarly ecosystems models the first set of vegetation pattern models they have used the advection diffusion and reaction rules but there was no hydrology biogeography biogeography models they looked at temperature thresholds and so on and the hydrology models they were interested in essentially predicting the the hydrograph right and maybe getting some of the saturation patterns and over time models have developed became more interdisciplinary and became more complex in all of these different disciplines and then more recently you start seeing models popping up in these intersection areas and what i call this the intersection of the middle between the landscape evolution models ecosystem models and the hydrology models is sort of an optimal or maybe an ideal model complexity and but what i mean by that is that these models should be sufficiently captured character characteristic space and time dynamics of each couple state variable and and the fluxes that are exchanged by them perhaps a model like landlap or toolkit like like landlap could have the potential to develop models like this because then we can update each state variable at their based on their characteristic time scales i'd like to give a couple of examples of this from a child landscape evolution model in studying the influence of essentially the influence of biomass influence of shear stress in a coupled eco-hydrology and geomorphology model this study was conducted in a sort of a similar type of climate like new mexico and the idea is that there's an eco-hydrology model that represents the characteristic soil moisture and evapotranspiration dynamics and there's a geomorphology model that relates the biomass to the shear stress and then does the does the geomorphic model the model was tested at the soil moisture time scale i get to test it at based on observations there's an example of soil moisture evapotranspiration leaf air index calculations model and annual runoff and sort of the runoff and return period relationship they have been sort of backed by by observations and then the model is used for long-term simulations to study the influence of influence of aspect and solar radiation when we take this model and run it across latitudes remember that one outcome i talked about was the the the role of the the latitudinal variation of the hill slope asymmetry and when you model when you run this model across different latitudes that we observe this behavior north facing slopes over the polar facing slopes gets steeper than the south facing slopes and this pattern running this model across the latitudes given different precipitation rates we can see that this pattern sort of pervasive across the 40 degree north and 40 degree south and and within this different precipitation regime still the observation holds and another interesting outcome is what surprises do we get from the coupled model now we use this model in order to predict look at the sensitivity to an abrupt climate change and i'm going to talk about an experiment about that uh run by an arm and sink at all um in senate antenna false laboratory and they have an experimental mountain setup um there's an uplift provided to this and and there's a rainfall simulator and they've studied they looked at the change in climate what you're seeing is their observations in a slope average slope angle an average erosion rate domain okay at steady state starting from an initial climate with a certain slope and steady state erosion rate on the domain an increase in precipitation leads to a rapid increase in rapid increase in erosion because there's more water and there is loss of slopes the landscape becomes flatter and then as the weather climate continues then there's a transient period erosion rates decline and the slopes decline and then at that point they they change the behavior to a dry climate uh and and the cycle goes back to where it started and what they see is a counter clockwise response in the slope and erosion domain to wet and dry climate cycles now if we take our landscape evolution model and do a similar experiment like this this regarding vegetation assuming that the landscape is bare or maybe very little vegetation in it we have the same response the same sort of pattern emerges um in the system but when we put the dynamic vegetation on the system with its ecological uh variability and so on we get a totally different type of behavior um the the behavior becomes more complex now one interesting observation is that during the wet period there's deposition there is deposition in the system because of the vegetation growth lush vegetation grows and it has an influence of the flow shear stress partitioning and and the sediment deposits from hill slopes into the channels and the slope stick line um when the drier climate change when the climate changes to the drier then the vegetation dies and there's a there's increase in erosion initially slopes increase and then slopes get back to as the as the landscape goes back to the steady state and the slope stick line and erosion declines and the very interesting behavior emerges uh in the model it's just because of the vegetation dynamics and its significant behavior another observation in this is that um but if you take all these points as independent data and fit a line through this with the dynamic vegetation this sort of resembles the mean erosion rate mean basing slope relationship that have been widely published based on looking at the long-term erosion rate and and mean slope data so a couple of notes here is that dynamic vegetation in the model that governs the transient landscape response it's key to sort of getting these cycles um in in the response of erosion and slope it reverses the geomorphic response to climate change compared to the bear case or in the insignificant vegetation cover days and it lowers the variability of the spatial mean erosion and slope over time so the landscape actually becomes more stable and one last note on surprises is that is the question of savannah stability and one big question in ecosystems is that are savannah stable over the long term maintained by disturbances this is a land lab application uh using in a semi earth climate it has a storm generated in it there's solar radiation soil moisture vegetation dynamics and cellular automaton spatial competition has been used between plants plants seeded with certain fractions and the model runs in the model there are fires and there's a linkage between the fire the fires are generated by lightnings and if the if the if the lightning sees a grass on the landscape it has a higher chance of burning it if it sees a trees on the landscape it has a smaller chance of burning there are um you know like the thousands of years this landscape became like a like a grass dominated savannah but at some point some gradual increase um some gradual increase in in the tree cover led to sort of a critical uh canopy structure and that's sort of interacted with the with the fire model because when there's lightning it was seeing more trees and grasses and all of a sudden there was a catastrophic shift from a stable looking uh grassland savannah system to a more um tree dominated savannah system so that's sort of an another example of like a surprises in a coupled model but the important aspect of this is that if we have landscape pollution model running underneath it and I think the patterns will be will be very interesting to see in terms of these geomorphic outcomes so we discussed three fundamental questions we proposed that the model should sufficiently capture your characteristics space and die at space and time dynamics of each coupled state variable and fluxes and we presented several examples and with that I'd like to thank the uh watershed dynamics group and the land lightning thank you thank you erkan um we'll we are starting to get short on time I'm not um at the perfect gatekeeper and I also wanted to like give people the chance to talk to our speakers a little bit even if it's through the chat so we'll take two questions and I'm hoping you're going to stick around for muriel's talk in a little bit of like an overtime um and so that she will have her time too um so are there any questions for erkan I see like a lot of positive comments that they liked um your like connection between like the vegetation and like how landscape evolution processes um coming I don't see any pressing questions yet oh here's one from Greg can you elaborate on what was behind the sudden shift in vegetation type so the sudden shift in vegetation type is just the the connection between the fire fire frequency and and the and the landscape existing vegetation type on the landscape the little changes in tree cover that leads to that that sort of separates the fire cycle because the trees are um less susceptible to fire than than grasses and therefore as a as a tree is attained sort of a certain threshold of cover then and it reduces the fire probability and and the fire probability leads to more spread of trees through the cell automata process and then and then there's a further reduction in fire return period and that leads to sort of a complete shift from a grassland landscape to a forest landscape um right now the the questions are starting to fill in like really quickly erkan you'll have like a little bit of a chat job after after we move over um maybe we'll take zele's question uh what do your morphic processes and properties do you think are relevant to prediction of the kato scale vegetation and carbon cycle oh that's a very good question i mean i think it would depend from side to side and it would depend how large your watershed area is and so on and i think like rapid erosion and deposition would be one thing to look at