 Okay, as Irina said, so I've been working more at a sequence stratigraphic timescale, sort of the hundreds of thousands to millions of years, which means you have to sort of very much abstract some of the processes that are going on and all of the detailed process work into what's the long-term version of that. And so, as Irina said, I've been working on this with particularly with Eric and Greg here, trying to build this into into LandLab. And I've also been helped by David Olegon, who at the time was a Bronx High School of Science student to help me with Python programming. And he's now a, he's finishing his freshman year at MIT. So the original, and I should say this project actually got its its genesis the last time I came to the systems meeting and met with Eric and Greg. And we created a proposal to work on an area in in western Turkey. So in western Turkey, Wems is Arabia is hitting Turkey, which is is moving out west. And then you have a subduction zone here that's pulling this way. And in here in this kink, what's happening is Turkey is undergoing north-south extension. So this ragged edge, western edge of Turkey, is a result of north-south extension producing all these bays and promontories. And the collaborators that I have in Izmir collected some beautiful high-resolution seismic data there, where I hope you can you can see where there are various sequences that we believe are the 100,000-year cycle mapped out. And some of them, like if you follow this tier one, you can see it going down due to the subsidence because there's a riff basin just out of the page here. But then it comes back up and here's the clinaforms up on the shelf and then it's dropping down. So the sequences have obviously been warped by the tectonics. And so, I mean, obviously this is subsiding, but is this part simply stable with other parts subsiding around it? Or is this actively uplifting? This is certainly related to a transpressional fault across here. This is the location of the line. And so, the idea was to use sequence stratigraphic models to model the stratigraphy and try to back out the tectonics and relative subsidence rates. And, you know, I previously, as part of Stratiform and EuroStratiform, created a sequence stratigraphic model that was the basis for this one. And it is, you know, a completely undergoing bit rot. You know, it was built with a GUI that no longer works on any modern computer. But some of the basic stuff underlying it, you know, it produced models that at least made pretty pictures where you can see things like how the shore face is prograting with the sea level fall, as well as the depositional shelf edge. And you get interesting effects like, you know, even as sea level rises and the shoreline starts regressing, the depositional shelf edge can still be prograting, you know, being driven by the sediments that are being eroded by the transgression. The pictures and it like many other models can produce sort of the three main erosion surfaces that you see in stuccointigraphy is the subarial erosion surface, which is the sequence boundary, but it also produces a marine erosion surface that's usually mostly bypassed, not very much erosion, across the shelf when sea level falls and a transgressive ravinement surface as the shoreline advances and erodes across the top of it. And we're able to use that in a couple of places while it still worked. So in the Yellow River on, you know, sort of the tens to hundreds of thousand year timescale and some models and the millions year timescales, they said it, you know, is is gone in terms of effect will be able to run. And so the idea was to create this now within LandLab and take advantage of, you know, Python programming and and all the capabilities that are built into LandLab to hopefully make it easier to port it to something that's that's usable. Now LandLab is generally a plan form based model. I'm at the moment still doing a two-dimensional cross section so the way we do that in LandLab is by making a grid that's a three by a large number. And we set, very simply set the closed boundaries on either side so all the sediment has to simply go through the the center line. So that's what we we make it and yeah advantage in LandLab is I can simply tell it these are closed boundaries these are open boundaries and it worries about all the boundary conditions and making sure that everything's correct at them and saves a lot of of very painful time and effort in the programming. So here's for instance an example that's produced by the the modern version where on this I've just covered colored the fluvial, short face, shelf and slope. It's done with a slightly asymmetric sinusoidal sea level curve where you can see the again the shoreline prograding out the shelf edge moving back and forth not quite in sync with the with the shoreline. In this I'm still not completely happy with the algorithm I have for picking what the shelf edge is. It's not quite the easiest thing to to do simply and this is so this is a running with 100,000 year cycles so there they're actually 6,000 layers of sediment in here that it's kept track of here. I can also produce wheeler diagrams so this is looking at distance versus time and the blue is showing you where there are sediments that are being produced with the darker colors being the thicker sediments and the gray represent representing sediments that were deposited and then were subsequently eroded. And the one on the right is is similar but unlike the older version we have now a very simple but working two lithologies so we just have sand and mud yeah producing a very fairly simple variation in lithology across the model okay here for instance is a model where instead of a relatively sinusoidal I put in the oxygen isotope curve from bisecti and ramo where I've just converted it to a directly to a sea level curve even though I know their temperature effects and it's not the actual sea level curve and you can see now that it you know has these tongues of of prognation that are being produced with each cycle because generally each cycle is asymmetric with a very slow long slow fall and a very rapid transgression and so you get these long tongues that come out and all these little wiggles in the sea level curve are producing a lot of back and forth within the prognation with all these erosion surfaces that you may or may not be able to see from from a distance and again you can produce a wheeler diagrams going to shoreline and the and the shelf edge with the shelf edge again not quite being in sync with the with the shoreline but obviously a lot more a lot more complicated with all the variations in in in sea level so the the basics of the model includes well the the big thing is is how to represent the sediment transport and deposition across the model which I'll go into in a in a few minutes you know there are sediments coming in the landward side where you specify the volume of sediments percentage of sand and then that sediment influx can vary with with sea level for instance you read in subsidence through a file sea level either through a simple sinusoid or or or a file and we have flexural isostasy compaction we have the routines already exists but we haven't connected it up I was afraid to break the model before this meeting because I'm sure once I do there will be all kinds of things that will have to be repaired and so there it's still in a state of of of flux where there's a few more things that we have to you know add in adjust and debug but I hopefully soon it will be you know publicly available through through systems so the input parameters come in through a yamo file which is yet another mock-up language you know so there are various sets that talk about how to time step through the model the shape of the grid out at where to output the file which come out as net cdf files you know where to get the the symmetry and the sea level the subsidence various parameters for the sediment transport and the properties of the of the sediments probably more parameters will get added as time goes on so breaking this down that's real awesome essentially on the coastal plain I follow Chris Payola's long-term model of essentially a linear diffusion with as you you know increase the river flow from precipitation towards the coast the diffusivity increases towards the coast and you can set a basin scale for how much and then I use alanita rotors model or a non-linear diffusion for the shelf so the transportation on the in the marine regime is obviously some kind of non-linear advection diffusion scheme and people like john swenson used a more effective model I've used a more diffusive model I think both of the models produce a lot of very similar things because I think a lot of the features that you see at the sequence or at a graphic scale are relatively robust relative to the details of the model but this model does have some nice features that I'll show in a moment the you know failure on the slope we haven't yet added in but we'll probably add in some kind of simple scheme for for collapse of the slope when it when it gets too steep and I'll show right now we have just a almost a more of a placeholder for the for the hemipelagic mud deposition that I really like to improve and I welcome anyone who wants to help me improve it to join me in this so so onshore they said it's simple a non-linear diffusion with a linearly increasing diffusivity when I go and and forgetting the lithology onshore we it's a this is a scheme that was produced by Juan Fidelis and Chris Paola we're essentially as we you see earlier today you know channels tend to accumulate the sands and grow faster than the floodplain and so based on the model parameters it calculates a channel boat aggregation and a floodplain aggregation relative to the mean sedimentation rate and you know and assuming that channels evulse when they become super elevated it calculates then the number of evolutions per per time skip to figure out how the sand and mud are distributed and calculates a an average sand fraction you know this is a two-dimensional model so this is essentially averaging across this width of the of the of the basin offshore it's a said it's a this non-linear diffusion that alanita rotor devise we are basically diffusivity is a function of both distance and and water depth with some you know with small terms added in to help keep things from blowing up near the near the shelf edge and then I've added in an exponential decrease in the diffusivity below a wave base the one of the nice things about this essentially makes a dynamic diffusivity that varies as the as you prograde and and and and regress so what you find is that when the system is prograding the shorefaces tend to be taller and steeper when they're transgressing they tend to be shorter and flatter and so when you transgress you often just a road off the top of the shoreface and the lowest shoreface tends to be preserved in the upper shoreface eroded which you know fits with what you see in a lot of outcrops so i've been so you know i don't know if there's a better schemes out there but this certainly produces things that that seem realistic to me um and then that exponential decrease in diffusivity once we pass um a wave base that you that you set um enters the sediment transport and gives you this rollover of the depositional shelf edge or the the clina forms or pro deltas or whatever you want to call them um the only problem with the scheme as it's implemented is that then um very little sediment goes past the clina form um and so the idea is that we will take the amount of mud coming out of the river mouth and use plume to calculate an initial distribution and then need some kind of advection diffusion scheme to distribute that properly um and i could use some help from people who know about those kinds of processes right now as a placeholder what we've done is just a simple scheme where um the mud um so it's getting deposited from wave base and increases and then feeders out as you go farther from shore which makes a nice tail onto the onto the system but clearly could be you know very much improved to produce something a lot more more realistic so here's um that same example again and here's a uh just showing it in a in a movie let's see what's going up so you can see silver going down and the shelf prograting out in front of it um when the shoreline turns around you actually continue to probe for a little bit and then it and then it comes back in you can see the shelf is having going back and going back and forth um so this is just taking that that file um the output of what the final stratigraphy and since compaction isn't in here i'm able to invert and calculate the stratigraphy at every at every time step and run through it you can see the the sink in here that's due to flexure um so we have flexural isostasy where you can set the flexural rigidity where if it's stiffer the deflection from the weight of the sediments is broader and if it's weaker it tends to be larger and more closely focused this is a fairly low one but one thing is at this time scale flexure is not instantaneous right so you know at best it would um go on at sort of a glacial isostatic time scale you know of sort of thousands of of years but you know the the actual gia response is also wavelength dependent so the time scale should be longer when you're going down to a basin scale instead of a instead of a you know ice sheet scale um and that has the potential to um have have influence and i think you know estimates of what the time scale should be is not very very good so initially we put on the sediment load um right now we have nothing initially you could put on just a pure elastic load uh deflection which would be uh relatively small and then it sinks into it and so like here's something where i have a relatively short time scale which is what i've been using in most of my models um but if for instance the response time becomes very long becomes long um comparable to the um sea level cycle then what happens is you know this is programing out and the area behind here is still subsiding you know when when sea level was up here and you were depositing up here it wasn't um getting isostatic load and so you can see more fluvial deposits being accumulated because it's subsiding after sea level has after the shoreline has gone by and you get sort of the opposite effect um in here where it doesn't subside and then it's um it's subsiding during the transgression and so it's getting more of of deeper waterfaces instead of instead of shelf so this is something where the the response time and of the isoscecy and sea level cycle can interact when they start getting close in in time here's a another couple of experiments i did where this one is with constant a constant sediment input and on the top and bottom i have ones where over here sea level uh the sediment input decreases as sea level falls so that might be something like in the Ganges Barma Putra Delta where i'm working where when the monsoon shuts off the sediment supply they are probably decreases by an order of magnitude and so you know there's more sediment at the at the high sea level and you can see the larger more fluvial and you can see the shorefaces becoming thinner and much less of the of the shelf because there's less sediment coming in then here's the opposite where sea level is where sediment is increasing as sea level falls it might be more of like a Mediterranean climate you know and so we're seeing a lot less fluvial and a lot more you know sediment prograting things at the at the low stand so again that's a user a user-based parameter and so here is another movie showing a 600,000 year run with you can follow the little bouncing red ball where you can see there's a lot of jiggling back and forth with the with all the twists and turns of the oxygen isotope curve and you can see you know coming in and you know every time it tends to go back and forth you can see lots of erosion surfaces and and more complicated pattern of advance and retreat it went back and forth you can see the little tongue from the first part of stage seven versus the main part of stage seven against some some kinks and then here comes the big final Holocene transgression across the whole thing and so you can see again these these multiple tongues related to the long-term sea level curve if i take the same model and run it not with the last 600,000 years but 600,000 years earlier in the Pleistocene when the 40k cycle was dominant i get something like this where it's lots of very fast patterns and not as much variation in sea level and you're tending to get much more vertically stacked fluvio shore and shelf deposits so i can actually yeah so you can compare the two you can see overall the over the last 600,000 years sea level has on average been lower and so the entire package is a little bit farther seaward in here and then for another variation i did this these models are run with a fairly rapid subsidence so that these nice tongues of sea level are are fairly are visible and not well you know i'm off off to a new jersey where you know there's almost no subsidence and a lot of them are just eroding and and repeating each other and amalgamated so here's one where i i cut down and got rid of the the shoreline and smoothed out the initial photography and you can see this much more of an overall progradation of the system as it as it progrades across this this ramp so there's enough of a ramp to to give you this accumulation but you can see the the difference in fact i had to move the sea level curve to the other side because it was covering covering it in terms of of tectonics we haven't yet started to model the the Gulf of Cusadesi field area i'm actually going there in in in two weeks to visit the area and collect and get a kingdom sweet version of the of the project with all the interpreted seismic lines but meanwhile i also in that area been working in the Marmara Sea where what we found is in this rift basin is that there's a series of beautiful prograding low stand deltas and what happens is most of the main rift basins are here but there's an extra one here and so this area here in the southern shelf is marine now but during low stand this becomes exposed and the rivers come and dump the sediment over this fault and so when that jump over the fault is not too high we've seen a bunch of a series of low stand deltas prograding off during low stand and then it transgresses back so we just tried making a model where i put that kind of backward subsidence in here to produce these tongues of of shallow prograding shruff face it's it's not yet tuned to to match this but was more of just a proof of concept to make sure that i can make it work i also the the marmara sea becomes disconnected from the world ocean when sea level fall when the last time when the sea level fell below 85 meters and became a fresh water lake so i cut off the sea level curve at 85 meters so these are a little bit flatter in front but you can see these series of of prograding deltas which when we when we collected this data was you know a really useful discovery because that gave us these nice hundred thousand-year timelines that we can tie and get age control over the stratigraphy within the entire marmara sea so the you know in any way the you know future plans is you know whoops is is now to go ahead and add compaction i also want to make some improvements to the ability to change parameters during model run so you can you know start a model change parameters around and continue the continue the run you know for instance in my field area i don't know if that uplift you know perhaps started partway through the time period i'm i'm looking for you know or if you know there's a river capture and sediment influx increases you know i want the ability to be able to to pressing the wrong button reason and then i hope once we you know put in the last few extra things like this to release it the sds dms and github and make it publicly available you know on a slightly longer term i i'd love some help to improve the the mud transport across the the shelf in deeper water so i want to add the some kind of slope failure and and simple turbidite and i'd be happy if anyone else wants to volunteer to help join our group to you know take part in in future improvements on the longer term we actually have a proposal to couple mock persons fluid flow modeling through this so that as sea level fall sea level changes go on we can pump fluids through the through the system to look at where there's both you know fresh and saline groundwater entrained in in in in the systems was for instance off of new jersey there is fresh to brackish water up to 70 kilometers offshore in the sediments in bungalow that's where i do a lot of work the hollow scene section in in the southwest is is primarily saline but underneath in the Pleistocene there is fresh groundwater so we'd like to try to adapt this to model those those places i talked about this for at Levant and there was a suggestion well what about you know coastal deposits can you get more detail in that and perhaps an estuarine you know module or organic matter of preservation and you know the the system is written relatively modularly so that i hope you know if people want to help develop new modules that we can plug them in relatively painlessly you know and of course then you know perhaps down the road if people want to work on on carbonates and you know greg has certainly suggested 3d i think 3d will be a lot bigger job and probably require you know a whole new proposal to to go that way but you know landlab is inherently 3d and so you know since it's built in landlab it should be you know at least possible to turn this into a 3d 3d publicly available system because there are some 3d sequence stratigraphy models or models but you know some are for shorter time and longer time ones a lot of them are proprietary so i wanted to present you know where we are you know right now and so we welcome anybody who wants to come in and help us improve this or just wants to use it to yeah talk to me later thank you thank you