 Thanks James for inviting me to give a presentation about the X-beach It's always nice to be here. This is the third time I've been here in Boulder, and it's it's always snowing. I know this so I don't know Maybe the summers are different So X-beach is is a open source model that was developed for the Army Corps and it was specifically developed to provide more dynamic modeling capability to look at impacts of hurricanes on coasts There's a long history behind it, but it It was initially developed by a small relatively small group of people And the PI was done a moving He's at UNESCO and then involved people at Delta Iris Delta University and then when I moved to Miami it also involved the University of Miami and what you see at the top is example of the new X-beach website Because the old one was falling apart So they developed a new one and this is where you can get all the information that I'm not showing you It has YouTube videos, etc. So the other thing is as it from its inception It has gone from a relatively Small group to a lot of people that are developing and that are using this model. So I'm happy to be part of that. I adopted the title for from Essentially from James. He told me to talk about spatial scales and Temple scales, so I thought that was a good start for the title and instead of going to larger scales I keep going back to smaller scales. So We'll see where we end up We have a lot of support on going support and there's also European support, which I haven't mentioned Okay, so it has almost no equations in it, which is which is good But I have to talk about the the processes. So the key process That we included in this modeling was wave groups. So to simulate erosion under Strong conditions instead of just using a mean wave height. We added a shorter scale process On the timescale of wave groups and wave groups have timescales of about 25 seconds So what you see here is a depiction of wave groups that are Incident on the beach These wave groups because of their modulation in energy and momentum They generate if a gravity waves and the if a gravity waves is ultimately what runs up and down the beach causing a lot of Sediment motion so for us that turned out to be really important part of the process the other part is Avalanching so as these waves run up the the dune face It mobilizes the sand and if the sand if the slope gets too steep It avalanches and the sand gets deposited from the dune into the foreshore and where it gets transported by waves and currents So avalanching is important Wave groups are important to test the model we compared it to a number of observations and This is an experiment that was performed in the Delta flume in in the Netherlands. It's a prototype flume pretty large scale not true prototype but pretty large scale and We compared the measurements with observations what you see here is the initial profile is in black And then you see two blue lines and the dashed line is the x-beach model computation and solid blue line is the the observation And this is after one hour two hours four hours and eight hours and essentially The model is very good at predicting dune erosion. It doesn't do such such a good job at The bar that's developing in reality The model is not capable of doing that. The other thing I want to point out is that if You exclude avalanching it does a I was gonna say crappy job. Okay, it doesn't it doesn't do a good job In predicting dune erosion if you do not include avalanche If you leave out the wave groups It's equally bad. So apparently These two mechanisms are important That also meant that we needed to include smaller spatial and time scales to get something that is occurring at a larger time scale another example is breaching so this is based on an experiment that was done by Paul Visser Where they went into the field they built a dyke of about 250 meters long and they had an initial You call that Cut into it not very big but at least that it started the flow and then in within a within one hour We go from a relatively narrow channel to a deeper channel and then as time progresses So time is on the horizontal axis in minutes as time progresses we see that the channel is Starting to widen and after one hour we see that more or less the channel evolution has come to a stop Because the water level inside and outside are similar on the right. You see a few pictures of Not of this specific event, but this is what we observed in the summer. We did a field experiment in Carmel and Yeah, I had Carmel River State Beach And this is a femoral river and supposedly during the summers that river is closed Except for this year. It was really wet and rained a lot. So while we were doing an experiment, which was not focused on femoral rivers We had a lot of water coming in from this river. It would breach and then it would close up again it would breach again and so I think we have a Really nice data set that that we're going to explore with X beach. This is a really cool result that was obtained by Robert McCall He looked at modeling of overwatch at Santa Rosa Island during Hurricane Ivan and what you see is Model predictions of The barrier island at the beginning of the hurricane. So this is where a lot of the waves are reflecting This is a snapshot of the surface elevation and then during the storm The overwash is being initiated. So you see water coming from the The Gulf of Mexico side into the back barrier Bay here. We have full inundation and here we have The situation where the water level has dropped again after the hurricane has passed This is data from the used. Yes Where we have the initial bad profile. You can see the road over here. So this is the Gulf of Mexico side This is the back barrier Bay the brown colors are high relatively high. There's nothing high in Florida. I have noticed So relatively high This is the model simulation of the erosion after six hours where we see some initial dune erosion This is where the overwash starts kicking in and this is at the end where you can see that The front of the island has lost a lot of sand and there's deposition on the island and in the back barrier Bay typical of an overwash event at the bottom are the observations and As you can see the patterns and the distribution are actually quite close and he shows that for this specific case We get a good model skill and This is another example where you see the pre-storm parking lot and The post-storm parking lot. So apparently this model has some skill in predicting. Let's say destructive Events But that's the thing it does it. It's so it's very good at predicting what happens. It's at short-time skills with really high energy But we need we would like to have the capability of also looking at what happens are longer time scales It means we have to say something about beach recovery So we have two problems one is we need to go for longer simulations And the other thing is you have to go for beach recovery In the corner, I didn't mention it But typically so I give the typical dimensions of the problem that we calculating space time scales time steps duration And you have to take into account that this is all based on my very old laptop So you can do a lot faster if you have a better laptop So this this is some examples that the down a roofing Calculated for the formation of an app title and flood title Delta in a situation that's sort of typical for the Dutch Coast and Key here is to get to longer time scales We use a morphological factor and there's actually a lot of other things you can do to speed up the process And it gives you really interesting results. So in this case, we have a tide that's just Coming in and going out without any waves. So you see a nice flood and F Delta and if you compare that to The scenario where you have waves You see that there's a lot of so the waves are actually from the let's say the north East no the northwest So they're driving the little drift is directed from West to east and you see this bypassing of the sand and so it is the model approach like this is capable of predicting long-term evolution, but the problem is that it does not add any functionality so if The beach recovery phase for example is not included in this model approach So that's a bit of a problem and this is the big problem that we have so we have all these short-term Processes that are responsible for moving the sediment around related to waves and currents and and that just Gave another presentation about this and of course all of that is a function of the overall Configuration which is determined by stratigraphy vegetation Whether as urbanization, you know You can go on and on and on and all of that feeds back into the sediment transport and then into the evolution of Barrier islands tell title Delta formations and etc So the key question is how do we resolve this and I want to have So I started looking at well what what happens if we take the barrier island recovery as a case So we know that x beach does a good job in getting rid of all the sand Now we'd like to get it back. That's that's the key. So Initial profile and then we have the black dashed line Which I call the post storm profile and then we need to have some mechanism that actually brings the sand from offshore To the coast so it can build up the wind can pick it up and we can get a new barrier around So we need on short sediment transport to restore the barrier island. Otherwise, there's no kind of work Now if we use x beach for this because it's so good at predicting erosion It'll erode the beach And that's related to the fact that we have a wave related mass flux that's bringing mass onshore in the surf zone Creating a return current or undertow to bring it back And as a result, we typically find erosion at the shoreline and some deposition Further offshore Okay, better hurry up So that's a problem. So in to solve this problem. We go back To even smaller scales. So instead of wave groups, we're now going to look at individual waves So these are x beach simulations that were done by Arnold for lawyer where you see the run-up of individual waves on the beach and Hypothesis is now that the turbulence which is at the front of this bore that is running up the beach is picking up all this sediment And it's moving it onto the beach and leaving it as the wave goes back So to solve this long-term problem We're now going back to the interwave timescale counterintuitive to say the least Okay, so we have some neat experiments that we've done about two years ago Could be three in Monterey Bay and what you see here on the upper left are two surveys of The bathymetry and Monterey Bay is characterized by a beach that has persistent rip channels and shoals So this is a rip channel over here. This is a show. There's another show here And this is a 10-day difference. So you can see there are differences in the in the bathymetry This these are the conditions during that 10-day time period So this is a tide tidal elevation of here We have the significant wave height offshore which is the green line and then we have the measured wave height At the pressure sensor one which is modulated by the tide We have the mean period which is about 10 seconds relatively constant And this is a fruit number that says something about the strength of the circulation of the rip currents So in the beginning is relatively weak and then it sort of picks up and it peaks at low tide and this bottom plot is The grain size the mean grain size as function of time and space and this is obtained by using Edie Gallagher's digital image Processing gear This Which shows you the grain size distribution or the mean grain size in millimeters So it ranges from about a hundred micron to 1.5 millimeters And what you see is so this is time and this is space that there's a spot of core sediment which is located around 45 meters which is Sort of where the brakes the waves are breaking So this is where this the foreshore connects to the beach and you have a strong change in the slope And this is the shore break So there's a vigorous breaking at the shore break and we observe just by looking at it that at that location You see a lot of sediment being entrained into these breaking waves that then travel up the beach So that was the main motivation to start looking at this Now if you look at the individual profiles, so this is I have two transects one S and one N If you look at the individual profiles over 10 days, there's Significant variation, but if you look at the along shore varied average profiles You can see that over the 10-day period even though there's a lot of variation in wave height and tidal elevation It's more or less the same. There's some small changes. So it suggests that the beach profile is Stable so the question is for us now Can we model this because we know that x-beach always erodes everything can we Like it taking into account these additional processes. Can we actually model the stability of this profile? And of course the answer is yes, because otherwise I wouldn't be keep I wouldn't be telling you this story So but to do so we use a multi-sediment class formulation into x-beach So I've defined 10 layers And 10 different sediment grain sizes. It's all sand We can calculate the sediment transport associated with each of these sediment classes And then for each sediment class we can calculate what the corresponding changes in the bed elevation So that way we can calculate the change in the bed profile And we can also calculate the change in the great in the grain size distribution What you see here are the results. So on the left is the measured profile The blue and the cyan and the difference between the two correspond to this dash black line over here So that's after 10 days. So this dash black line doesn't change in time. It's just the final profile change Then what you see is the model calculations Which is the grain size distribution in color So the initial grain size distribution is a combination of all these different grain sizes that we put together But then you see that there's some erosion at the shore break So if you remember where we are it's sort of where the waves are breaking So we see a build-up of coarse sediment at the location where the breaking is most vigorous we see Finer sediment being transported offshore and finer sediment being transported onshore So we see a nice sorting of sediment which occurs at relatively short time scales The bottom plot shows the comparison between predicted and model grain size distribution at the surface So this is compared to Edie's Camera observations and provided that the rip currents are not very strong It seems to be doing a pretty good job. The other thing is that the changes in the profile So this is after 10 days Both in the model and in the observations are relatively small. So by including this process we can Stabilize the profile however So now We wanted to look at recovery and now we're looking at interwave processes So the time steps that we're looking at are less than a second. The vertical distribution is in the order centimeters So we have added the dimension. We have an horizontal dimension But now we also have a vertical dimension. So we made the problem just a lot worse to get to where we want to go So that seems like a very long way to go before we can get any any good results So the key is to aggregate these small-scale results to a scale Where we can use it to predict longer-term changes And my argument is that that is and of course that's what everybody is doing In trying to decide what kind of model they want to use and what kind of physics they want to use The key is to get the processes that are relevant for the long-term evolution What we can do at the small scales is that we can give you ideas about what is important And hopefully that we could translate that into something that operates on a longer-time scale And we can have more physics in the longer-time scale model So We need to couple these systems to enhance the functionality and this is Another example, so this is something what people at Delta hours are working on so the beach recovery phase So this is another example of what the erosion and deposition During a hurricane so all the red dots are actually houses that have disappeared So to recover to be able to predict the recovery you need to have on show sediment transport I said it already Burn building because once you have sand on the berm the wind can pick it up and can move it along So wind is important Then of course we get all these other things that are important like vegetation Buildings, etc But at this point we're looking at this part of the problem and not so much at what happens at the next level And to do that This is where systems comes in this is where We can couple all these different sub process at their Level at which they are optimal which they can work Well for you to get something that works on a much longer time scale So there's an experiment ongoing at Delta hours where they're coupling June Which is an aeolian sediment transport model with x-beach for the times when the beaches eroding during storm conditions And then del 3d for the more moderate condition The first test case is a acidic island in collaboration with you. Yes And of course this is not done by myself This is a large group of people and I'm sure I forgot a lot about other people that are involved Thank you. Thanks at the perfect timing. Thank you And are there any questions we have time for some questions? This is going to seem like a crazy question, but In the tropical environments, are there things like I don't know Cementation iron things like that that are going on that are That happen in these tropical environments that may not happen in the more northern US European environments as x-beach Been tested like in I don't know Indonesia in places like that Short answer is no, it's not those kind of environments. It's been thoroughly tested in European environments Of course, we've used it in sub tropical environments but Anything that would affect the behavior of Yes, sediment of course that would Look forward to Discussions about the most effective way to incorporate what you learned with the smaller time scale in spatial scale Modeling into longer term But I wanted to ask you a separate question about a detail Seems to me that the onshore transport that has been missing from x-beach is mostly Velocity asymmetry that could bring sand from deeper you were talking about moving sand on shore in the swash zone but you need to also bring sand from the lower shore face to to get recovery after a storm and Could you get get that in there? I blasted you I don't so actually The sediment transport modeling that we included in this example has bed load and suspended load the bed load has the velocity asymmetry in it so using Nielsen's expression to look at the differences in shear stress and progressive waves But in addition so we found that if you don't have the suspended part, which is related to the bore Suspension You don't get the result that we observe So it needed an additional process to get to get it right I had a question rather similar to James's and can you extend this you mentioned that this was good Or you used sand grain sizes. Can you extend this into gravel beaches? You can and there's a group in the UK that is working on it The thing with gravel beaches is that the groundwater flow becomes very important. So There's an additional mechanism that are to be taken into account, but x-beach can cope with groundwater So and it seems to work well any other questions No, well, thanks Ed. Thank you