 Appreciate it. It's great to be here. It's my first proper systems meeting and regrettably, I can only be here for today because the family issues I was hoping to spend the whole week here. I think it's a great community, exciting mix of science in these clinics. And so what I'd like to touch on a little bit are the different scales of inactions within solid earth processes perhaps is relevant for the community here. This is the sort of thing that that I normally work on trying to understand the planetary evolution on on the largest scales. The forward models we're trying to explain certain data sets in certain regions, and in ab initio models that try to explain the general plan form of convection. And when you look at something like this sketch here from SC 40, then all that usually becomes this one error which says tectonics and And since I'm excited would be nice right because then I didn't have to worry about you, you wouldn't have to worry about me we could all be in our little happy silos but it's not quite as simple and of course we know that everything is linked in the sense that mantel The question is the reason for plate tectonics there for all tectonic activity, including a rodney which was pointed out by by homes. Interestingly, right decades before plate tectonics was actually accepted and we know more specifically, for example from work from the fall down and others and meet and Conrad that there are certain settings. The rodney in the Andes where a good case can be made that the plate velocities sense what is happening at the interface, and at the growth of the mountain as shown here in paleotopography constraints leads to an increase in the sheer stress forces on the plate boundary and that perhaps slows plate boundaries down. So we know that there are some interactions. And it's kind of useful to then ask, well, which, which of the driving forces which of the governing terms are the most important. And if you do some sort of local analysis as to how plate tectonics works then it comes down to a balance of viscous dissipation and slap hold the first. And the thermal boundary layer models of oxberg and turquoise and others in the 60s did that. And so you can consider slab going down here. And then if you balance the slab pull force that is written here as a thermal, thermal buoyancy force with dissipation in the convective mantle where we have the mantle viscosity here. And that sets any sort of timescale. That's the stock's velocity the velocity of the sinker that has a certain density and it's modulated by the ambit viscosity. But when you then do things a little bit more carefully and you consider the lateral viscosity variations that I expected, given the temperature depends on viscosity. And then claim Conrad and others have shown that the dissipation within the bending plate, which is this term here that strongly depends on the bending raise radius might be quite important the stronger the slab the more important that term is. And then there's this other pesky thing which is the interface shear string, where if you consider the brittle shallow interface the stresses are perhaps quite small. So you can consider the lower shear zone component of that plate boundary and then this term can matter. And the nice thing about this force balance is, we can do simple estimate. And so, for example, when you then predict the plate velocity as a function of slab viscosity compared to mantle viscosity then particularly for strong slabs, you have a reduction in the plate velocities and that reduction depends on the bending. And there have been decades of work trying to understand subduction models based on isolated plates perhaps an overriding play trying to explore the role of bending but as it turns out, slabs may overall not be very strong based on temperature dependence alone. There may be orders of many orders of man who's stronger than the mantle, but it appears that there's some process reducing the strength and depending. If we look in seismic tomography slabs are bent, you know they're sort of squishy. And we can think about the micro physical mechanisms that lead to that strength reduction and plastic behavior is one one such mechanism where in terms of an instantaneous behavior we can look at the Christmas tree diagram. And we can see that there are these regions where piles plasticity might be active, and sort of taking a chunk of the list is for strength profile. There's some evidence from geodetic inversion that we're seeing such a reduction in viscosity in this abducting plate based on to hooker okay post seismic data. We can also worry about so the second order features. And recently it's been suggested that an interplay of brittle and ductile damage might lead to a segmentation of slabs, which then might be reflected in some of the seismic tomography images we're seeing and in the end the behavior of the slab is sort of like a toy snake that can be pulled and it's quite easily bet. And so what does that mean. Well, if the slab is just a couple 100 times stronger than the asthenosphere, then we have very different forces. Here's fairly old work by my former student. Adam hold illustrating the horizontal stress field where red means compression blue means extension for two different time steps where there's a slab going down here and we'll be interested we're interested in this region. In this weak slab, that's only 100 times stronger than the mantle, you're very different behavior for storms that you can also see that the stresses in the overriding plate at time. Right, so if we're talking about this tectonics arrow, then how that arrow works and how the forces are transmitted depends on the slap strength, the lower the slap strength, the more important the overriding plate, and that whole things time to independence can be quite pronounced if you have things like slap folding as we saw in the seismic tomography. So you have variations on time scales of millions of years perhaps even short. Now what does that do to our plate velocity. If this lab has viscosity relative to the mantle of around 300. Then we can explore what happens if we change the sheer zone viscose. So, as the sheer zone viscose the increases for certain lengths and aspect ratios of the sheer zone. You can have a significant increase decrease in plate speeds. Sorry, and you can have an increase in the actual shear stress accommodated in that deep viscous shares. So this is a very simplified exercise based on force balance it doesn't include sheer heating other feedback processes but it's intriguing because it says that if slams behave, as we think they might know relatively weak sort of fashion interface matters. And so what Whitney bear and I suggested is then, if you then look at different types of interfaces, then you can, you know, explore based on field observations on PT, PT conditions, and based on laboratory strains on creep laws for slams that have basalt turning into ecologite or slams that are covered oceanic plates that are covered by sediments, you have very different viscoses. And so this is the plot turned on the side and this would suggest that for a particular slap pole a particular length of the slab. If you have agglotide of the surface has higher viscosity, you have fairly, fairly low plate speeds, and if there are sediments coming in those sediments can serve to lubricate plate motion. Not a new suggestion that has been, you know, proposed again for the Andes by lamb and Davies and others but what that means is that now the interface itself isn't just important, but it also matters what comes into the interface. So the geology in the sense has an effect on plate speeds perhaps. And this might matter on short time scales, but it may also matter on long times. If you look at the convergence of India with respect to your Asia here plotted in terms of absolute and orthogonal velocities, and there's this well known increase. And these, at some point, there's collision. What we suggested is that it might be that this is perhaps associated with an equatorial bulge here, where the Tethian lithosphere might have had a carpet of the logic sediments lubricating the plate. This raises an interesting questions as to what might be happening in terms of the overall evolution of the planet, the formation of the continents, or Agni increase sedimentation rates, mass flux changing what is happening at the plate boundaries leading to a feedback with a deeper and that's quite interesting. And some of the things were suggested by a civil fm brown year later. And so, since then, we've done a bit more work on the role of the interface, and these particular computations use the aspect mental connection code, but which we'll hear some later in the clinics. So it turns out when you run these little dynamic models when we started here, particularly interesting, they have, they have a have a layer of different viscosity crust in it. And then as a function of the interface viscosity here 10 to the 2120 and decreasing to 10 to the 18 which would then be roughly what, like 100 times weaker than the upper mantle. You can just roll back, and you also have an increase in, in the introduction. So the nice thing is, you can then take these fully dynamic models, and bring it back to an analysis that is akin to the force balance analysis, even though those are fully time evolving there's rollback and other things and you can show the convergence velocity here as a function of the interface viscosity mental viscosity for different shears on aspect ratios, you have again this decrease in convergence velocity as a function of interface. So rollback plate speeds are enhanced by the sediment cover, and you know, the more the sediments the faster the rollback and also the faster the convergence speed. Now we are currently exploring these variations in three dimensions. This is work by by no heart and others are not involved in this and trying to do similar tests looking here at the stress like this, the stress long strike your topography and there's incoming incoming plate material with different costes and you can get similar responses and the next step is then to take a look and see how surface transport matters and how the mass rebalancing at the surface then interfaces and to explore these feedbacks between aerogyny and rollback. Now, in terms of mass transport, even that I'm at a surface process focused meeting I thought I show one result from work where we actually did explore the role of surface processes this is now quite old by Boris Kauss my former student Claire statement where we have a setting here there's a pretty complicated lithosphere coming in with sediments upper crust lower crust and then the mantle lithosphere, there's some sort of overriding play there's a free surface. We then explore the role of erosion and sedimentation it's a very simplified description it's just some sort of hill slope diffusion thing that smooth is out and transports. The mass away without any consideration of deposition, which of course we should and so what you see is then in terms of the large scale behavior here, there's the slab. Here these layers of the crust and so those are models that were a little bit tailored to Taiwan and we were very interested in the exhumation and just comparing the case without erosion. With very fast erosion you see that there's a bit of a change in the large scale dynamics bit of the change in the rates and that's due to what's happening at the interface but the major effect really is not surprisingly perhaps that you have way more exhumation. You have like a vacuum cleaner, cleaner sucking up the mass on top of it so the the shallow dynamics of course very much affected by surface processes and mass redistribution but the deep dynamics, perhaps not so. With this in mind we then continued to see what happens, in particular in a subduction setting, looking a little bit at what's going on at the mantle wedge where we have all these interesting processes different we have the building up of an accretion a wedge. And we have perhaps underplating recirculation and we want to understand these, these different pathways, starting with trying to do a better job at representing the, the accretionary wedge. So this is, these are results and work by Sylvia but it's a former postdoc of mine, who worked a lot with Ilona one dinner and this is a large group of authors and so these models and theory can really go across the time scales. And we're not concerning ourselves here with the earthquake timescale we're just running these models with fairly complex layering of really different realities and different materials, and changing the sediment thing. Right and so what we expected based on these other models and sort of the general observation of force balance is that we would see an increase in plate speeds, if the incoming sediment layer is changed in in terms of its thickness and we saw sort of the opposite. And what happens is that the average velocity of the slab actually decreases and this these colors here. These colors here are sediment thickness the thicker the sediments, the slower the plate speeds, and you can imagine it wasn't deeply fun to just have one paper saying the opposite of the other that we just published but what turns out in in these models is that you see all of these is really big accretion where and that leads both to an increase in the shear stress due to the increased overburden, and more importantly, the increase in the length of the plate interface. The secondary effect is that there's a bit of a reduction in slab pool, you know, depending on how you set up the mall. So, when you look at wedge dynamics, there are certain scenarios, be they realistic or not, where thick sediments can decrease the plate velocity by increasing the sort of interface, we're going from a small to wide accretionary wedge and and having having a stronger level of shear. And so, that's interesting that's a complication that you want to understand better and it might apply differently different regions but slabs also don't live in isolation. They interact with each other. And I want to sort of close by a couple of comments bringing things back to the larger scales, the largest scale we can think of a global scale and turns out that plate velocities and different plate reference models shown here are showing a natural rotation of the lithosphere with respect to the deep metal so everything moves a couple of centimeters sort of westward like the Pacific plate moves for example in hotspot reference frame. That matters. If you worry about where slabs are going relative to the surface what's happening with rollback and things like that. It turns out work by my former student Melanie Jerome, that I kind of forgotten about is that the weak zone viscosity based on these two dimensional models, having slabs and keels as drivers of flow to explore the role of net rotations, affect the net rotation so the stronger the weak zone viscosity, the more net rotation within limits there sort of a soft, there's a sweet spot here in the weak zone with also matters and it's always a trade off between the two. I was reminded of that work when she's a young pointed out this paper to me this is work by my own song from two years ago where they run global circulation. Put in density anomalies based on different approaches and you can match the plate velocities and then they show that if you have lateral variations in the weak zone. If say the Western Pacific is weaker than the Eastern Pacific, then you can have a change in plate velocities in particular you have an increase in this degree one to idle flow which is the net rotation so they're global effects the whole lithosphere is doing something differently. If you have lateral variations in sheers on spike. And this is just to say that these cost scale interactions matter, because if you're looking at things like advancing and retreating trenches in a region such as in Japan to understand as shown here, the deformation patterns if you're trying to understand the global stress state to then further analyze what's going on over the cycle or doing a Roger need you need to understand these global reference frames and the global reference frames themselves, depend on what is happening in slabs as shown here in this old presentation by she is young and Mike Ernest showing you that how rollback changes. Once and if the slab penetrates through 660. So we have these cross scale interactions where slabs are being controlled by sediments. They are trajectory throughout the mantle then affects large scale pressure fields which can change the net rotation and that will change the way the slabs make it into the mantle. Before this we're trying to do so in an earthquake context with recently funded and as a fresh project where we're also running summer schools so if you're interested in these things in an earthquake, earthquake focus, I encourage you to reach out, but I want to conclude with this recognition that slabs may well be relatively weak. And if this is the case the interface may matter, not so much the brittle interface but the deep ductile share zone. The plate boundary stresses for sure matter to moderate plate speeds lateral variations may matter on global scales and the sediment transferred conserved to lubricate or come up the interface. And there's these possibly global effects of local plate boundary mass transport. Thank you. Hi Thorsten for like giving us sort of like huge picture of like deep earth. Are there any questions we'll take at least one way maybe have to. Hi Thorsten thanks for the nice talk. You showed plots kind of parameterizing the rollback in terms of the sediments and also eclegite there. Is there no effect of the slab age. For sure this lab so that the slab age. This is all for constant slab age. Okay. And so the slab age sets. Mainly the thickness, and that will affect the slab pole, but also the bending in a complicated way this is why these bending analysis are kind of annoying and so those were all at fixed radius and then for these malls, you know, it probably says somewhere like 100, 100 million years. And so it, it will, the interface will always matter, but then the thickness will affect the specifics of the trajectories. So you can think of the age of setting like a background. And then on top of that you vary this, the interface strength and it will still have relatively the same effect. Absolutely it's going to look different. There are other questions to Mara. Thank you for the talk was super interesting. My question is related with the stress, if these models can help us to see the stress changes in the overriding plate, because I'm trying to link with the like what happened in the surface and how, how that can impact the topography that we are studying a surface process. So if what is happening in the deep, this lab and how does changing the stress overriding plate. Yes. And so here is one example, which shows the horizontal compressive stress when this is negative and red it's compressive and vice versa. And that just illustrates that in the case of the weak slab, you have way more compression, for example, than in the case of the strong slab. And in general, the rule is the more messed up the slab, the more important the overriding. For example, if you have classically weakened slab, the thickness of the overriding plate matters because it affects the pressures at the interface. And so yes, so the answer is yes, it's kind of complicated. This is a follow up question. And because I think it's interesting and hard to answer how these models can help us to understand the time scale of those changes. If a slab is weak, how, like in how long we will see that reflected in the overriding later. So, you know, this kind of stuff, where we work with Tars Garand, Dave Bercavici, you can, you can look at the segmentation here, for example, and that segmentation makes predictions in terms of the offset of faults in the C4. And so you can measure that, and you can quantify that, you can go a long strike where the plate age changes and you can see how that's basing. So that would give you an idea, for example, of the intensity. Right. But then, in terms of the time scales, right, it's a it's a it's a very good question right for certain problems. All I'm talking about might just be that one vector right that says tectonics for others you might have to take these interactions into account. And I think if you're talking about building a mountain over five to 10 million years or something like that, for sure, right, be talking about very short time scale things, then, then maybe these interactions are less important. And as you know, right, it's not just the time it's also the spatial. So, for example, there's, there's going to be a clinic right by Joe Nabilov later, where people have put service evolution malls on top of aspect, and there you can you can answer that question empirically. Right, but of course we understand analytically some of the answers and I think others are just more complicated. That's where the exciting work happens. Thank you. That was.