 So we can move on to the next speaker. So thank you, Julio. So what I forgot to mention was that we had many people applying for this. There was a stiff competition. And it was pretty hard to make a decision on who won the prize. And the second running up was actually Rachel Glade. Rachel is a student here at the University of Colorado. And she's going to present her modeling work on blocky hill slopes, erosion of blocky hill slopes. All right, check. All right, just before I start really quick, I want to thank my advisor, Bob Anderson, and also Greg Tucker for really helpful discussions and collaboration work that I'll present today. So if you look at Earth from above in this picture, in a plane or even from space, a lot of times you can see signatures of geologic structure and rock type on Earth's surface. So for example, this is in Fort Collins, just an hour or so north of here. And here we have tilted layered sedimentary rock. And you can see that you can see the rock type stretching across the landscape. And resistant layers of rock set the relief of the landscape. They set drainage divides that can migrate over time. We can see, if we zoom into that same landscape, not only do you see soil, which is what you expect on hill slopes, you also see these discrete large chunks of rock that fall off of these resistant layers. In other locations around the world, this is in Moab. You have horizontal layered rock. Here's a place where the hill slope would be tightly coupled to the channel. You can see debris on the hill slope. This is Shiprock, New Mexico, where now we have vertical rock that's actually basalt. So this is a vertical basalt dike. You can see these big blocks have fallen off the dike and they mantle the hill slope. And one thing you can notice in this picture is that the presence of the blocks ends pretty abruptly at the base of this slope break. Just for fun, I think it's a cool criterion to show that your work is important and curiosity has a selfie with it. So you can check that box, that's good. So in general, there's sort of characteristic features of these layered landforms are two or more rock types that have differing erodibility or weathering rates. So here is in Morrison, we have a layer about tilted about 30 degrees of sandstone. And that layer produces probably some soil but also these discrete large blocks that fall off. And that's overlying softer shale that produces soil. You can see this side, the dip slope pretty closely mirrors the bedding plane of the resistant rock in this case. Then the other side which we call the ramp is a hill slope developed in the soft rock mantled with this resistant stuff that fell off from the hard layer. And this ramp is generally linear to concave and profile. So steeper at the top than the bottom. And again, we have the blocks ending at the base of the ramp. I just wanna point out that one unique feature of these landscapes is the shape of this ramp. So in general on hill slopes, we often expect them to be convex upward. So steeper at the bottom than at the top. And Hilbert in 1909, believe it or not, came up with a really elegant explanation for this. If you have bedrock that's producing soil at some rate and that soil needs to be transported down slope, you wanna maintain a constant form and you assume that the transport of the soil depends on the slope, you need to move more and more stuff as you move further and further down slope. So this slope has to increase. So we really wanna ask two main pretty fundamental questions about these landscapes. How do they evolve over time and what is the role of these discrete blocks that aren't usually treated in hill slope models? So we can come up with a conceptual model of the ingredients that we need to think about to try to understand these landscapes. We need to think about block dynamics. So how the blocks are released from this little cliff, how they move down slope, if they move down slope and how they weather over time or break down and become smaller. And then what are the feedbacks between that and then our mobile regular, their soil production and transport down slope? Because hill slope processes have happened over long periods of time, hundreds to thousands, millions of years, numerical modeling is a really good approach to try to understand that. So we developed a pretty simple numerical model of hill slope evolution. That's kind of a hybrid continuum discrete model. The continuum part, we have a rule for soil transport where soil moves in a diffusion like way according to the topographic slope modified by how much soil you have. So if you have no soil, there's no soil flux, which is good. If you have a lot of soil, then more of it moves. We also have a depth dependent soil production rule. So bedrock turns into soil at a rate that's dictated by the soil thickness. If you have thin soil, it happens faster than if you have thick soil. And then sort of the discrete part of our model is, yeah, okay, we have these discrete blocks. So the way this works is that each model cell either does or doesn't have a block. It's a binary rule. If it does have a block, then it has a weathering rate that in this case we said to be constant and a transport rule. So our rule is a very simple terministic rule where we say this block will move down slope if the relief between the block and the next cell is equal to the height of the block. So this would indicate that big blocks move more slowly than small blocks. And we could show analytically actually that with this rule, we should expect an exponential decay in block size with distance from the crest of the hill slope. So just before I start this, this is our model run, a video of it. On the boundaries, you'll see incisions. You can imagine these are streams or something like that. And everything that's white is shale or something that's very rotable. The red layer is a tilted resistant rock, like something like sandstone. You'll see that blocks fall off of this cliff when the relief becomes high enough and they'll move down the slope and decrease in size over time as they weather. Let's see if this will play. And I wanna point out that the parts of the hill slope that don't have blocks follow that Gilberty expectation that I talked about before, where the slope is higher, further down the hill slope. But if you notice over time, the part, the ramp, the part that's covered in blocks, develops this opposite concave up shape. So the slope is much steeper at the top here than at the bottom, where the blocks are small. So here's a printout of the model over time. And you can see if we zoom into the model, what's actually happening is that when these blocks sit there on the hill slope, they act as obstacles and soil gets stuck behind them like a dam basically. And you can see that the, here's an example in the field, if you looked at the sort of average slope, what your eye sees when you just look at this from far away, that is much steeper than the local slope that the soil feels when it's dammed up behind a block or this effective slope. You can see that this is happening in the model. This lower line is bedrock and the top line is the soil. And we showed in our password that the slope is directly related to block size. So because the blocks are bigger up here, the slope is steeper and as they get smaller, the slope decreases and you get that concave up shape. But something unexpected happened in our model that we didn't build in and really didn't think would happen. So if you look at the time slices, you'll notice maybe the last three of these look pretty similar in the reference frame of the hogback. The amount of erosion for a given amount of time is constant and the shape of the hill slope and the length of this ramp looks pretty similar. So how does this happen? The, whoa, yeah, how? The model is a simple model, but we have a lot going on. We have blocks moving and weathering down slope. So it's kind of surprising that we get this nice result. And if we turn, just take a step back and look at sort of a geometric analysis. Let's look at this side of the feature. So the dip slope, we have a given vertical erosion rate at the base of this hill slope. And we want to see how this could maintain a steady state where the consistent layer is keeping up with this vertical erosion rate. We can do just really simple trig and say, okay, the dip parallel erosion rate has to be that modified by the dip of layer. We can go into our model and check that this actually happens. So is it really becoming steady? Yeah, it fits pretty well. So we changed our model to have dip angles from 10 to 83s. And you can see that the dip parallel erosion rate fits with our analytic expectation. But how is this happening? Well, if we look at some different model parameters, these are all plotted with distance from the crest. So down ramp, you can see this average topographic slope measured over a length longer, bigger than one block size decreases. This is our concave up shape you see. The block size is strikingly linearly decreasing with distance from the crest. If you remember before I said that the rules we put in the model said that it should be exponentially decreased. Something weird is going on there. If you look at vertical erosion rate, we also see that it's decreasing somewhat linearly with distance from the crest. But the flux of soil is increasing as you would expect from a sort of normal hill slope theory that you need to move more and more stuff. What is going on? Well, if we go back to our little picture and focus on this side of the picture now, if we imagine that our definition of steady state is that the shape of this ramp stays the same through time and the reference frame of this dip direction, then you can see that, okay, if we have a vertical erosion rate down here, where the slope is steep, you actually need a much higher vertical erosion rate to keep the same form of the hill slope. And then as the slope gets shallower, you get closer and closer to this vertical erosion rate down here. So we actually expect the erosion rate to be decreasing down ramp for the feature to maintain a steady state. And we can cast this as another just trig relationship. So trying to tie this all together, what do we know? We know that block size decreases because I told it to you at a constant rate. We know from our previous work that the average slope goes with block size. So bigger blocks lead to steeper slopes because of this damming effect. We know from this geometry thing we just did that the vertical erosion rate should also decrease down ramp to maintain a constant form. And then with the block velocity rule we have, where the blocks move according to the vertical erosion rate downhill of them and their size, because these two terms are both decreasing, it ends up actually self-organizing to give you a constant block velocity, which is why this is such a linear decrease. But even though smaller blocks shouldn't be faster, they're seeing a lower vertical erosion rate than the ones up here. And it all kind of evens out. And then this also feeds back into the slope and gives everything this linear look that it sees. This is also interesting because it points to the importance of depth dependence in our soil transport rules because even though the vertical erosion rate is decreasing and therefore soil is increasing down ramp, soil is not actually accumulating at the base, the flux is still increasing because of this depth dependent flux rule. So now that we think we have a better understanding of what is happening in our model, we can look at ramp length. So you have to imagine that this feature was projected up into the sky before and a block was released. And the ramp length will be how far that block moves in its lifetime before it's gone, plus the amount of cliff retreat you have during that time. So we can cast that quantitatively and just say, the blocks might go an initial distance when they first fall off the cliff. This is a time scale for one block lifetime. If this is the block weathering rate, and then it depends on how fast they move. And then this is just the horizontal retreat rate of the resistant layer. I don't have time to go through all this, but if we include what we learned from the self-organization slide before and just sort of rearrange this, we can see that we can relate ramp length and the dip of the layer and the topographic slope and initial block size to the vertical incision rate at the base and the weathering rate of the blocks. And this is cool because those are two things we'd like to know often, pretty hard to measure. And these are all things you can measure in the field. So we've taken this simple 1D model, try to understand it really well and come up with testable field predictions that now we can go test. And just to make sure that we actually understand what's going on and just make that up, we can change the model to have different ratios of this incision rate to weathering rate of blocks, then we can measure the ramp length and the model, all these parameters here and calculate what it should be and you can see that it fits really well. So we really have confidence that we understand the behavior of this 1D model. And I just want to really quickly wrap up and point to the work I'm doing now with Charlie Chobh, who's another grad student here. And he is a model of fluvial incision in the presence of blocks. So big blocks land in the channel, they inhibit erosion through form drag and cover effect. And we've put both of our models into LAN lab. So I went to 2D and he also went to 2D and coupled them. So now my hill slope produces blocks that fall into the channel and we can see feedbacks between them. Just as a teaser, here's some cool looking, colorful output from our model. The river is on this edge and there's base level is proven in this corner. So in conclusion, I hope I've convinced you that blocks matter in landscape evolution, particularly in layered landscapes. I also hope that you can see that sometimes it's useful to really analyze what your simple model is doing and really try to understand it. And it might have even surprising things that you didn't build into it. And that is, I hope, useful for coming up with field testable prediction. And also, if looking at smaller scale morphology and processes can help understand larger scale landscapes. Also, Charlie and I have dual posters tomorrow. Thank you. Thank you, Rachel. Are there any questions? Rachel, thanks for a great talk. So my question is looking at your nice one dimensional model, what I wonder is how either rate constant or the shape of the hill slope might change if you had a two dimensional hill slope surface or soil to creep around the blocks. I'm curious if you're proud about that. Yeah, so I mean, that's what this is. We, the blocks exist. It's basically the same as the 1D model, but soil can flow around the blocks. So I was wondering, you know, would we see the same effect? And yeah, we do. You can see in the background here actually it's concave up slope. This slope has more blocks than this one, which is closer to base level. You can see it's sort of, this one is, you know, becoming not blocky, and this one is still feeling the presence of blocks. So it still has an effect. Thanks for the great talk. So I was wondering, looking at your conclusion slide, and that input you're showing us, can your model tell us anything about the spacing of where the rivers are cutting through, how both river and the slope are going? Yeah, so that's a great question and that's the other thing that I'm working on now that I didn't have time to talk about. But if you notice in the background here, there's this sort of scallopy pattern for it, but where the rivers are actually cutting across the hill slope. And yeah, I think there could be some explanations for the block release mechanisms. You know, this side of the hill slope has felt the channel incision sooner than this part. There are places you can go where the shape of these is sort of curved up here and then very linear. So the middle of the hogback hasn't felt this sort of cross cutting channel. So I think there is probably an interesting dynamic in there and I do plan to do that. So that was a great talk and you definitely convinced me that you guys have that whole phenomenon, well constrained and understood. As far as the actual why it matters, could you like put some sort of constraints around what the like compare it to the null hypothesis as if all of the slopes are those Gilbert type convex ones. What are these blocky slopes doing? Are they arresting the morphology in place, preventing transport of soil down to the valley, et cetera? Yeah, that's a great question. So in steady state, which is what I focused on here, the presence of the blocks significantly increases the relief of the landscape and the persistence of topography. So I don't have any slides here, but if you compare to a control run where you have a resistant layer of rock that weathers more slowly, but it doesn't produce blocks, you end up basically getting rid of all your relief much, much faster than when you have blocks. And they're also important. I focus on layered landscapes here because I like that they're kind of simpler and you have a line source of blocks, but also in the mountains here you have a lot of big chunks of granite and stuff like that. So I think they do definitely cause steepening of the landscape and not in steady state. They can significantly lower erosion. So one more question. Thanks, Rachel. That was a really great talk. It's somebody that grew up just north of Morrison. It's really interesting to see these. It seems like your prediction or what fell out of the model in terms of the size of the blocks decreasing linearly with distance from the crest is really fascinating. That seems like something that with some TLS scanning and some machine learning maybe be tested, is that something that you've looked into or has anybody else actually tried to extract block diameters from laser scanning yet what the actual distribution is and what the ramifications are for the modeling might be? Yeah, I think that's a great thing to do. I haven't done that yet. I have thought about doing sort of drone surveys to measure block size. One thing is that I understand what's happening in the 1D model really well, but the rules for block motion, for example, we really don't know how they move into life. I made up this simple rule that now I can understand. So what we'd have to do is really have a better understanding of the mechanisms for that and then it's pretty easy to sort of redo the math I did with a different rule and see if you could compare that with field data and see if it fits. Okay, let's thanks Rachel for her talk.