 Shelby Arndt, who will be talking about river morphology and flood risk in the Pacific Northwest. Take it away. Hi, I'm Shelby, and I work with a team of researchers at the University of Washington to understand how river morphology or the way a river is changing shape affects flooding in the mountain systems, the Pacific Northwest. And in particular, how do these morphologic changes compare to hydrologic influences such as large rainfall and runoff events that produce high flows. And I'm going to use river gauging data to investigate historical changes in both of these aspects. And what I'm ultimately going to propose is that channel change impacts on flooding are comparable to hydrologic changes. The specific role channels play depends on location and their geologic context. And we find five different categories that impart different ramifications for flood management. And the greatest of these ramifications are forced by extreme natural events such as landslides and anthropogenic modifications such as dams. In this study, I am focusing on Western Washington, where we have a high density of people living adjacent to rivers flowing from mountain regions such as the North Cascades into the ocean. This is Seattle for perspective. And these communities are frequently impacted by flooding. We also have evidence that this region is experiencing changes in hydrology. For example, we're seeing seasonal shifts and flows as the snowpack decreases under warming temperatures. More fall and winter precipitation is delivered as rain, hence we see a rise in flow volumes shown by the red and orange colors here. Folks have also studied regional peak flows and have found evidence of positive trends which increases flow hazard. But these rivers are also very dynamic and mobilize sediment from various inputs such as the toe of headwater glaciers. We have landslides from neighboring river valleys and from existing stores in the lowland. And on top of all that we have humans damming and dredging the network and influencing the sediment dynamics. But how do we know whether this sediment that might affect a river's channel shape and conveyance compares to those increasing flows. And to study this I'm going to share a method for separating and comparing these two effects. If the channel is affecting flooding we leverage the idea that at a gaging site there is a relationship between the height of the water and how much flow is going through. If this relationship changes meaning more or less flow can come through the channel at the same stage or water height, the size of the channel must have changed. We can also measure this concept quantitatively through specific gauge analysis or by looking at the field measurement data for a river gaging site, documenting flows at different water stages or heights. And this blue rating curve defines a relationship between water level or stage and flow for some period of time. We're not seeing less flow or points below this curve. At given stages we know that the channel has lost capacity which causes flow hazard to increase. And we look at these residual values over time to quantify how the capacity has changed. In order to compare these changes to hydraulic we'd like to know what's happening to these types of flows or the flows that are right around the threshold at which the channel will flood. We look at this quantitatively through flow duration curves, which map a sort of mean daily flow to a return period or how often will that flow occur. Natural channels typically fill their banks about once every one to two years. And by looking at changes in the flow that occurs right around that return period we can gain insight into how hydrologic changes modify flow flood flow behavior. We do this for a suite of flow duration curves built for different time periods across our data set. And in particular if we see an upshift in this flow volume at that return period flood hazard might be increasing because there are more frequent higher flows. And the range of this variation kind of describe the extent to which hydrologic change may modulate over bank flooding. So what happens when we apply these analyses to 48 river gauges. Well in the morphologic analysis we find that channels are changing in a variety of ways which impart different ramifications for flood hazard management. First we have some channels that arguably aren't changing very much so here I'm plotting three examples of such changes which show the pattern of channel capacity over time. As a fraction of the flood flow and so if these values are are decreasing the channel is losing capacity and there's a heightened risk of flooding. But in these sites noted in gray on the map there isn't significant change in the channel capacity and so morphologic influences are likely to be pretty small. Other sites had significant long term changes in channel capacity and these changes will still fairly small might be important to consider as long term shifts and conveyance that could have impacts on future flood hazard. There were also channels that oscillated between periods of greater and lesser capacity throughout the record on decadal to multi decadal scales. And this was the most common behavior intended to be a bit higher in magnitude than the first two categories. These sites also raised the question of whether there's some forcing mechanism functioning in some river channels that causes capacity to oscillate such as periods of high and low sediment supply. There were also rivers that display prominent channel change in response to extreme events. For example in this river. There was a large capacity reduction after the Mount St. Helens eruption caused lahars to slide into the channel. These changes tend to be in basins with higher and steeper slopes perhaps contributing to sediment instability and they're often very abrupt. We also found a large response in some channels where a combination of existing sediment supply and flow regulations from dams interacted to have significant morphologic influences and often resulted in drastic reductions in channel capacity. These shifts were often long term and a part important consideration for management particularly downstream of dams where regulators may release volumes contingent on the capacity of the channel. So how did these channel changes compared to the hydrologic changes? I'll here I'm showing the range of those changes for all of the sites in each category and the orange box plots are showing how the channel change is affecting flooding in these blue box pots to find how that near flood flow is changing. So in all categories channel changes appear to play a role that's comparable to hydrologic changes. Arguably in these first three categories flow variations may play a slightly greater role in flood variability. But once you start having identifiable forcing mechanisms such as extreme sediments, pie events or dams channel changes start to become more significant. So ultimately channel change is important when considering controls on flooding and categorizing these sites into five distinct responses which is plotted here spatially on the right offers a framework to manage flood risk understanding that channel influences on flooding are based upon site specific eomorphic history and not regionally synoptic with the greatest of those being these extreme events and anthropogenic modifications. So with that, I can take any questions. Thanks Shelby. As people write questions in the chat or raise their hand. I have a quick question. So you. Oh wait, I'll let someone else go. Okay. You measured flood risk as how often the river exceeded the banks. Are there as right. And other measures you could have used or how. How would that have affected your results. Yeah, so I, this is a little bit of a tricky plot, but I was looking at those fluctuations in the flow volumes that are occurring right around when the river pens overflow its banks. Some people look at peak flows and so those previous studies looking at trends and the peak flows show how you know floods might be increasing in severity. But I'm sort of interested in right at that threshold where the flow will overflow so it was moderate floods that occur more frequently, but I could have looked alternatively at the highest flows for sort of an intensity metric. Great. Thanks. Okay, so here's a question from Greg to that's neat Shelby. Can you describe an example of one of the sites where there was a strong sediment supply impact. Yes. So, um, Greg is this, I hope, I think these are the sentiment those extreme sediment supply results or the dam sediment. Supply impacts. Um, well, in the extreme sentiment supply cases this lowland gauge received is representative by this capacity pattern. And it's actually receiving lahars that slid down the river after the Mount St. Helens eruption and sort of resulted in a drastic reduction in capacity that they then have to go in and actually dredge out to increase the conveyance of the channel and then further slid in. But ultimately the channel sort of reworked it back into a equilibrium state. Great. Thank you.