 Good afternoon to everybody. This early evening we have a lecture by Dr. Vasily Zitov who is from USA and he works in the Pacific Marine Environmental Laboratory which is under the National Oceanic and Atmospheric Administration of the United States Department of Commerce briefly it's a NOAA. I mean the world knows this organization as NOAA. Vasily is one of the well-known experts in the tsunami, tsunami modeling, tsunami data assimilation and there are many many other things related to tsunami. For example there is tsunami dealing with some real-time tsunami etc. Another important issue is also Vasily is not only the excellent scientist but also he promotes the international science of tsunami. He was the chair of the International Tsunami Commission of IUGG. Also he is involved in the activity of the Intergovernmental Oceanographic Commission and they are program on tsunami and many other areas definitely including the national agency where he works. Okay Vasily it's nice to see you with us. We are very happy that you could make it because we know that what happened yesterday or even on Sunday night that your place and that's why we are really very happy that today you are with us and you will be in this lecture. Thank you. All right well thank you very much Alec for the very warm introduction. Hello everybody my apology for a little bit of very shuffling of the schedule. Due to the problem on my side we had some weatherstrike I was out of electricity for a couple of days but everything is back on track now so what I'm going to talk about is the data simulation and the data inversion problem in application to tsunamis where I called my talk Data Driven Tsunami Forecast for Tsunami Warning. So I'm as Alec said sort of a general tsunami scientist because in Tsunami such an interdisciplinary field that you have to branch out to many different aspects of physics, mathematics, geology, astronomy even I'll talk a little bit about it so it's fairly general but the main goal for the tsunami science if you will if you think about it is well when it started anyway it was to try to forecast tsunami so when a tsunami hits we can save lives pretty much. Saving lives is fairly you know pretty much the main goal of the science of tsunami if you think about it although the science of tsunami of course is also for you know the our intellectual advancement also it's a very complex phenomenon so I'll talk a little bit about that but again the general sort of application goal of this science if you will is tsunami forecast and warning so the better we do it the sooner we do it the more life will save so the reason I say I'm saying is that this you know this data simulation and inversion problem that I'm going to be talking about is as a specifics in terms of the tsunami tsunami you know problem application and I'll talk a little bit about this but again the specific is that it's all driven by this goal of of quick and accurate tsunami forecast so there is there is you know fairly complex mathematics there but is also a very complex implementation phase of these things and I'll talk a little bit about it so maybe just explaining a little bit of where I'm from after the Alex introduction so I've worked for NOAA as Alex said National Oceanic Atmospheric Administration the reason for that is that tsunami warning system in the US is in NOAA so NOAA has is is running two tsunami warning centers in US and one is sort of international tsunami warning center that helps countries around Pacific at least to provide one and we provide we I work in the research part of NOAA so we provide research we study tsunamis we develop tools that we then transfer to tsunami warning so I'm I'm between the science and and and and tsunami warning practitioners I still consider myself a scientist but but I we we work very much and a lot with tsunami warning you know practitioners to apply this science so tell a little bit about that too and how the science that that that you develop or that we developed can be applied to a very noble goal to save lives so what I want to start with is just a very general thing description of tsunamis because it's it's still a fairly narrow field in geophysics the so so I just want to do like to put us all on the same page so we we're talking about the same you know the same things and using the same terminology we start we'll start with just the the word tsunami it's actually Japanese word and it's a very definitive it's a very descriptive word the Japanese in Japanese it it consists of two characters and these two characters can usually be translated to to to English as wave it's harbor and wave so it's harbor nami is wave so that is actually a very descriptive term and I'll just I like to elaborate a little bit of that on that it's not just the wave okay it's the wave that becomes dangerous in the in the harbor only where so where it becomes larger big you you can't really notice tsunami in the in the ocean so if you I like to show the reason for you know why why this actually just little illustration how tsunami behaves and why it's not visible in the ocean but actually not why but but how this is a very small tsunami that that was generated in Greenland a long time ago it's an old video so it was a sort of small-scale tsunami that's generated by the iceberg collapse what did you see when it collapsed there is a wave propagating it doesn't really seem you know too too dangerous while it's propagating through the deep waters of the fjord and this time it's it's it's a very close propagation but when it came close to the to the shore it grows and that becomes really a wave that's that's the harbor wave for you that's that's why it's called harbor wave so you see that it's only dangerous when it's close to coastline in the harbor that's why it's harbor wave so if you put it back into the set of geophysics say terms and the definition of tsunami would would go something like a tsunami is a set of ocean gravity waves coast but but the ones that coast by the large and large is the keyword here abrupt disturbance of the ocean floor it's a large scale disturbance in the in the previous example video it was you know relatively large but but the tsunamis that we talk about will be much much larger in scale well if you convert it to hydrodynamics then the tsunami become the gravity waves that since they're long they involves the whole water column into motion and that's described very well by by the simplified navier stocks equations that's called shallow water wave equations so what the reason i said that this they actually if you translate this japanese characters on the right to mathematics that's what you get the shallow water wave description it describes this term very well it's it's the wave that's it's sort of uh very smooth in the ocean not noticeable but it's it's become real wave in the harbor okay shallow water wave equations they're very uh well descriptor of tsunamis now why do we study tsunami what's the scale of this problem uh just a little bit of the tsunami history the the the the scientists since in tsunami field they're very careful collecting all the data for all tsunamis you know combing the historical books and and and the modern records also so that's about you know 2,500 2,500 tsunamis documented since about 2,000 BC so for about 4,000 years uh so it's how you know it's it's a lot but but it's it's a long period of time too so to give you a little bit better perspective on the scale of the problem let's look at the last say 15 years uh also so since 2004 you may have heard about the couple of tsunamis you know excluding the Indian Ocean tsunami but actually there were well over 40 it's sort of outdated flight it's a slight it's probably uh um you know up to a hundred events that actually cause damage not just detected but cause damage and and and or potentially dangers uh and it it kill tsunamis kill people at this again for the last 15 years almost a quarter million deaths were attributed to tsunamis and the damage is also staggering 200 it's a quarter of of three and or for just again the last 15 years so pretty much problem that's why you know studying it and actually applying the models and and and studying and getting the data into the models which is the problem for the data simulation and data inversion is so important because we want to use that to provide the forecast quickly so to to at least lower this number but it was interesting the the last tsunami meeting uh uh Eddie Bernard which is the sort of the the father of in the modern father of this of this field he's a very well known uh drew time now scientist he uh uh proclaimed this this I mean not the goal the challenge you know zero casualties tsunamis we we have enough tools mathematics computers to do fairly good forecast which I hopefully I don't know we're still need to improve but people still die so we need to convert this knowledge into into results to with the goal of zero casualties from tsunamis I thought there was a great challenge that he posed to the tsunami community I pass this challenge to you as uh as you know the growing generation of scientists uh now to namish the this again this phenomenon that's that's generated by source and if you know the source very well then you can simulate tsunami fairly well too I'll talk a little bit about this uh but if you don't know the source you can't do anything so uh what are the sources generation mechanism of tsunamis there are multiple the seismic uh the sort of earthquakes generate most of tsunamis that's they can't you know during this historical time of about you know for for us on yes it's about 80 percent of all tsunamis are generated by by by earthquakes and that's the most often generation mechanism that you that you hear about however there are other mechanisms that can generate tsunamis the landslides actually can be very large scale and if they happen near the body of water they can generate very large waves in fact the record for the highest run-up which we call run-up is the highest point of the of the tsunami wave inundating gland so the record of that belongs to a landslide and I'll show you that at some landslide in Alaska created almost half a kilometer high inundation run-up then there's meteorological tsunamis there's you know there are typhoons and hurricanes these are not tsunamis per se but they they can generate the actual tsunami waves which is you know the definition that I showed you large-scale abrupt disturbance that can propagate for long distances as a long wave and and that has become a very active area of research you know these because they are not that large but in some locations they are very persistent you know they're very very methodical while the earthquake generated tsunamis are fairly random process we don't know what they happen next meteorological tsunamis you know for some coastline it's a returning phenomenon like almost every year volcanic eruptions can cause tsunamis and in fact they cause one of the most spectacular and and and devastating tsunamis over the history but they're fairly rare and then even asteroids can generate tsunamis they are much more rare but but interestingly it's it's not a negligible risk from that I'll talk a little bit about that also in fact I may dive into the sources of sort of exotic tsunami sources for first before we go into the sort of the 80 percent of the problem the the landslide generated tsunami that's a lituya bay and that's exactly the place where this the record tsunami was generated it's a very long and narrow bay that's separated by by a little bank from the ocean and the tsunami was generated by the huge landslides from one side of the bay on the right side from from on this water and that's splashed up to half over half a kilometer on the opposite side but then as this white area around the bay you know it's actually it's the forest that was stripped by the wave while it was propagating from the source to the ocean so it was like half a kilometer near the source but it was constantly over 30 meter high and and many places over 100 meter high all along the bay amazingly enough there were witnesses of this tsunami that that was it was after the earthquake but there were the earthquake there was an earthquake that triggered this landslide and and some people even survived you know there are a couple of people that were missing in action uh from there because the the fishing boats often find rescue in this in this lituya bay from storms from bad weather and and a couple of people actually saw this monster wave with their own eyes and survived it so that's why that's why we know that that's a little more photos of this lituya bay and and aftermath of the tsunami this uh on on photo seas you that that's this highest you know the record run up from tsunami that's where it is and then you know some some other spectacular run-ups from this from this event we we do know how to model landslides i mean they are very complex models and that's just an example of one of them and and we've studied them for assessing the risk in terms of the tsunami inversion problem it's a very tough nut it's it to crack um the the reason is that the tsunami the the the mechanism this little mess that slides into the ocean has so much variability and so difficult to anticipate what exactly the even the total volume may be that it's it's really and then how what what is the dynamic of this falling mass is going to be that's the prediction of this wave is a very challenging problem so that it's it's much more a tough problem than than say the seismic wave that we we have some handle on and i'll i'll show you a little bit so while we can model this tsunami is the forecasting it and data assimilation fortunately it is a huge challenge still and but but it's a it's a problem you know we need to assess say for some studies likely a hazard assessment for sensitive facilities like a nuclear power plant for example you do need to know what's the largest tsunami can possibly be generated so these are important the media tsunami i'll i'll actually come back to that a little later in the in the talk again they can wreak havoc in in some specific locations including qs for east coast which is normally not very susceptible to tsunamis but but there are some danger for the media tsunamis and then in mid-terranean it's it's a well-known phenomenon as in some locations in spain it's almost a seasonal seasonal occurrence volcano tsunami generation again at again it's it's it's rare but but spectacular events and it's historically it it was attributed to the you know the end of some civilizations were now attributed to tsunamis generated by landslide the the most known is the terra explosion in something about 3000 years ago that led as as many historians believe to the end of minoan civilization on crit because the destruction was so large that the civilization did not survive this this hit on the on the very prosperous harbor there and that's that's perspective view in sort of in against against the agency uh this and the blow up of the what's what's remained from this terra uh caldera um although they they're still persisting today in in only in 2018 there was the last tsunami uh that of the generated by uh by very uh volcano in indonesia that claimed lives at the same location which is krakatau volcano it's a it's a volcano underwater volcano uh in sunda strength between uh between sumatra and java islands of indonesia in uh 1883 150 years ago that it it generated huge uh uh tsunami that thousands tens of thousands people died nobody knows for sure how many uh in the close in the closest location to in the cope on the shores of the sunda strength but the the same mechanism repeated itself in 2018 which is the same sort of that's called anak krakatau it's it's son of krakatau uh caldera which started to grow in the same location exploded again and again unexpectedly and again deadly and uh many people died there so indonesia is really now trying to figure out what to do with this with the prediction of that and prediction always comes with data inversion first the meteorite tsunami that sounds like an exotic um a mechanism that uh i mean we definitely never see historically there may be some anecdotal evidence that there may have been well there's one anecdote that the the tsunami from oil and all the the impact at uh uh uh around 60 million years ago in uh uh mexican gulf uh pretty much uh eliminated dinosaurs from the earth the but the the modern problem with with meteorite generated tsunamis is is that is is is that uh the uh nasa uh trying to estimate the risk uh from the tsunami from uh the the meteorite impact in general you know there's this special group in nasa near earth uh bodies and the uh and and they they they try to to see what's you know what what is the risk of the of the large asteroid hitting earth and on the left upper upper corner you see this risk curve in in nasa for some reason the using an interesting metric for risk fatalities per year i mean it's an abstract metric but you know one can use that um so the i mean since it's definitely doesn't happen every year the meteor struck especially large meteorites but they if you stretch it over the yes you can come up with a metric and that's what they use so interestingly depending on the diameter of the impact on which is the size of the meteorites that's that will potentially hit the earth the fatalities per year it's you know they're pretty large you know for the small impact because they just happen very often and then they become large for for the ones that are very rare but very destructive like the end of civilization type events which is more than a kilometer you know large body hitting but in between there is this uncertainty and different authors have different estimates of that the reason for that is that there's uncertainty of what would be a tsunami that's generated by this but is uh the potential meteor extra and that's why we came in the tsunami scientists we were working with with with astrophysicists to try to figure out to to narrow down this this risk curve uh and uh well when if the the the meteorite is meteorite is large enough it will actually hit the earth the the ocean and that's what you see on the right lower corner but if it's a smaller one like the one that we you know the people actually saw one in uh what was that 2013 over the city in in in russia and it's a large city of chelabinsk in seberia and and and and people actually saw the meteorite exploding over the city create quite a bit of damage uh like almost over a thousand people were injured nobody died but what happened there was that the meteorite almost you know completely were destroyed in the in the sky so what what happens there is that the the shockwave is formed and that's the model of that is on the law on the law right and then the question we were trying to answer is if this shockwave can produce a tsunami also you know not only the body itself so in terms of the you know big meteorite hitting the ocean there's six impact again that I was talking about the uh we actually modeled it's it's a very recent study we we decided to see I mean there's the impact from that event was multiple and tsunami was probably the least one of the least this you know disturbances from this impact but apparently we it it had a global reach at tsunami itself you know you see the model that that we come up with scientists using very sophisticated generation mechanisms it's not exactly the the inversion problem that we're trying to solve but in a way it is but it's very simplified one we were trying to see with a uh estimated uh impact size because we know the crater size which we tried to put it into the models series of you know more less and less sophisticated models but more reaching uh to see what the impact is and and that's the model of that that's uh it doesn't tell you much but only quantitatively it took about two days to engulf the whole world by the way you see the uh that the the continents that the gray gray out areas these are the continents that's how they look like about 60 million years ago so they're different that's one of the problem you need to come up with the with the proper symmetry to see how the wave propagates uh but but it looks like it was pretty impressive tsunami of everywhere for any cost on the uh for this world for in terms of these small meteorites you know this is 250 meters for example meteorite which if it's not an iron one it will disintegrate completely uh in the air but then the big shockwave hits the ocean and that's what you see in this animation courtesy of our nasa friends and gale and gisler uh in particular uh and that creates actually the uh a fairly large disturbance of the water that propagates with the shockwave also and interestingly enough the shockwave which is the speed of sound in some areas of the ocean can propagate almost as fast as tsunami so that's there is a resonant feature there that we've explored and and put some of our tsunami models into work and and and try to assess at least some potential you know area for the risk assessments but let's talk about the seismic generation because again as I said you know about 80 percent of tsunamis is the seismically generated and this little cartoon shows just you know in general how this generated the the the earth crust under the ocean ruptured and that creates the wave on top and this wave the sort of the bulge on the top of the ocean propagates as the wave outward and that wave is the tsunami wave so to model this wave that propagates across the ocean and we want to model them to to to forecast them for for the coastline we we want to know what the source is the again and this so and that's a classic conversion problem right you know so you want you know some data the first data that comes our way is the seismic data because the seismic waves propagates about 1000 about 100 times almost 1000 times faster than tsunamis actually not I'll take it back about 100 times so seismic data will come first and you you will have lecture you know or maybe you already had one uh that speak in in details about inverting the seismic seismic data seismic waves seismic recordings to understand the source of this shaking and that's what we have to use to to figure out what is the tsunami source there potential tsunami so so is the potential tsunami so there's a bit of a problem there the seismic waves they measure the shaking of the earth right uh so they do look into the finite period of of this disturbance what is generating tsunami wave is if you will the infinite period of this spectrum it's it's the static disturbance you know the world the world the the the the earth shakes the bottom of the ocean eventually displaced and they what you see here is say on the lower left oh sorry on the lower right is the typical you know just a perspective view of this typical simplified model of this disturbance of the ocean floor and this disturbance you know is of the scale of hundreds of kilometers it's you know 100 kilometer long 100 kilometer wide very large part of the real estate of the ocean floor will be disturbed and and the models that we have you know they fairly simplified the sort of elastic media deformation models can provide you with the deformation pattern general deformation pattern that that we use and that can be inferred in fact interestingly from this seismic wave from the shaking that this disturbance create disruption creates one of the ways to do that is called you know central cmt solution central moment tensor solution which looks at many different you know a recording of the of the seismic waves and come up with this simple simple you know point source solution and two from from that cmt point source solution we can derive this what we call you know this the static deformation solution that takes you know for the static deformation this simple model that we see again for the you know that's that's Alaska tsunami that was generated back in 1996 we're trying to model that and that's the model source for this tsunami to to generate this disturbance you know to model this disturbance and we call it this this model was you know was was created a long time ago by a Japanese scientist the and we still use the same equations very often because they're very fast and and very convenient and and this scientist okada actually it's an interesting story too the okada sun wasn't actually did not derive these equations it was he actually published a paper correcting the previous publication by Benzina Smiley which has a little you know error there but he put it in such a nice you know present presentable way that's easy to use that everybody is using now this this formulations that he published and so this he become one of the most cited scientist in the at least in the tsunami field and actually I met him he was he retired recently he didn't even know that he is so famous in tsunamis because it was it was his graduate study paper that made him famous so anyway the okada formula is still used for to create in this assessment definition the problem with that I mean there's no problem with the formula but the problem with the application that's formula is that it takes you know this seven actually what is it seven it's actually nine parameters if you include the the coordinates and if you want to do it fast and again our main goal is to to do the forecast for for the coastlines that that are gonna be inundated soon and it takes about you know an average about half an hour before the big waves will start hitting the closest coastline that's that's almost a general rule about you know 15 minutes but but the big waves will probably come in about half an hour so we have this half an hour of golden time before before you can do some actions and say at least they evacuate people from the coastline so you need to do things fast and early on in the in the assessment of the of this earthquake out of these nine parameters you may know only one to three well the coordinates which is two and the strike deepen rate uh let's you see here that's can be defined in sort of fairly soon and fairly soon i'm talking about you know minutes maybe tens of minutes and remember if you have only 30 minutes to to one the the first coastline that's a lot of time but even even with that when the cmd solution available you know solid cmd solution is available uh we still don't know this the length the width and the slip because these are sort of connecting with they're not independent parameters if you will but they depend on each other and depend on the magnitude but they're not known exactly so here's the dilemma i mean we have to come up with those on the fly somehow so to to to come up with this very simple formula with simple simple uh definition of the source so again the seismic wave seam distribution as a sort of first line of defense if you will uh and then version of that it's outside my lecture but you will i'm pretty sure hear about that very well and there are some faster techniques now and much more sophisticated techniques uh that you can come up with not just a point solution but uh a solution that uh that that actually create you the distribution of sleep along the fault and that is a much more usable solution for to to put into the tsunami models for propagation and inundation estimates but this takes even longer time again it's uh uh the computers are getting faster and faster and the methods are getting you know extremely sophisticated and fast and that's the what you see here is the the finite fault solution for the the the tsunami in japan i will refer to this to this tsunami quite often in the talk because it was well it was fairly recent it was very large and it creates a wealth of data for us to study you know to to study the tsunami tsunami phenomenon further uh so the that's uh that for the finite fault solution it was seismic seismically very complex around anyway because it was large and it was shaking for 10 minutes uh so uh that solution was was uh available like a day maybe a day late which is definitely very late for the actual tsunami forecast but it's still useful for for tsunami status so it's this is you know the inversion problem is useful on on any time scale but this you know with my occupation i'm very much interested in a very short scale uh of of uh of the inversion problem so there is much more there are other data seismic data well it's sort of earthquake data that we can tap into to try to to infer the source of the tsunami and that uh the recently well like the last 10 20 years uh the real-time gp gns s well uh uh uh gns s data uh uh it become available and that measures the permanent deformation right away which is sort of closer to the tsunami generation phenomenon than the the you know the seismic way because seismic way was just shaking to infer the permanent deformation from shaking is fairly difficult proposition you have to integrate that and it's it's not it's not easy especially in real time that measurements the measurements the the gps gms you know there are several other systems there that that that provide real-time uh estimates of the deformation uh is is closer to is estimating the tsunami where the problem with those is that they are all on land and uh if you have a lot of them on land like for the case of of the japan tsunami there's a lot of real-time gps stations there uh uh gns s station well mostly gps then the time uh and you can infer uh the source that's even far away in the in the depth of the ocean and you see it here is the uh the results of the inversion i'm not going to talk about that very very much here either i'm pretty sure you'll have a latch on that separately because it's a fairly complex phenomenon complex phenomenon but what potentially it can give us and it's still potentially uh it's not operational yet is a very fast estimate of the same t solution that i was talking about yeah this this point source if nothing else it will at least give you the the good estimate of magnitude which is very important in fact even the magnitude when i said you know the same dissolution does give you the magnitude but it comes later what's what what what sensing data gives you first and at fairly fast is the magnitude and location and even that can be a very critical data for tsunami warming but even that simple three three parameters uh will will be very uncertain in the first minutes after the earthquake especially the large earthquakes like like japan tsunami unfortunately only large earthquakes generate tsunamis so that's here's the dilemma you know we need the large earthquakes is the most difficult to assess but that's the one that we need to assess with the data with the with the data simulation um but that's not the only problem yeah with the with the seismic data in the in the assessment data and and the uh earthquake source uh assessment data here's the plot uh that i use very often as well from my colleague from russia slava guzikov he's an expert in a historical database of tsunamis he's looking at historical data and then slicing and dicing it in many different ways so here's he's trying to see how how good of a predictor is the tsunami magnitude uh for the i mean not necessarily for the earthquake magnitude and then mw is probably the the the most uh correct and the most sophisticated earthquake magnitude so how good of a predictor of the tsunami force is that so he plots i mean there are many ways to do that so he here he plotted uh the earthquake magnitude mw on the horizontal axis and on the vertical axis he plotted you know tsunami magnitude it's some measure of tsunami intensity uh the the integral measure of intensity so what you see here if you say take just one magnitude you fix the magnitude eight which is sort of the borderline magnitude when you see the magnitude eight earthquake somewhere in the in the world oceans you you start to be concerned about tsunamis now the magnitude eight can generate what this this tsunami intensity mt is a little bit of strange number it has even negative numbers and negative numbers mean very small tsunami it can be detected but but it's very small so magnitude eight earthquake can generate intensity minus or tsunami magnitude minus two which is pretty much non-visible tsunami can be detected but probably not comprehended all the way to the to the magnitude four tsunami which is catastrophic tsunami so you see the dilemma that if you use just the earthquake magnitude which is the first available parameter uh for uh for our inversion uh for predicting tsunami strength tsunami power tsunami energy you have a very difficult task in your hand you know it can be all the you know it can be no tsunami or it can be catastrophic tsunami so the problem is of course that you know the earthquake is is one phenomenon and tsunami is related but a different phenomenon and and even with all this sophistication you know the final fault solution uh is is a great stuff if it comes in time it it may be it great to use and it is great to use the problem is that there's a physical limit there the the we still we need to get this all this seismic data uh into our system and do the inversion and that takes time you know the the seismic seismic wave propagates fast but it's a limited uh uh uh speed it's about you know thousand two thousand kilometers per second it still takes minutes and minutes for all the stations to register the earthquakes and then and it does take all the stations in the world to do a good assessment for the uh for the cmt solution and for the final fault solution in consequence so it takes time and and then well it takes time it's it may still be inaccurate especially in the first few minutes and that was case in point in japan's money i'll i'll talk a little bit more about that um vassily i'm very sorry to disturb you further so vassily do you hear me yes sorry it is alec yeah i i have a question would you like now to have a break or in five minutes uh what what time actually very good very good point to take a break yes uh yes and actually i have also a question which is a scientific related to this uh slava's graph uh he used the all earthquakes or earthquakes with the trust faults oh that's i mean it's a mechanism because it's if it is the old earthquake then it's definitely tsunami from magnitude eight will generate a very small tsunami that's my question it's here is all earthquakes or earthquakes with the mechanism of trust faulting is all earthquakes and that's a it's a good question because yes i you know there are there are sort of uh two two answers to this you know first yes i want i mean this is to definitely to make the point that the earthquake may not just the magnitude may not be good victor of tsunami uh tsunami density but there is a there's a there's a second point here also the um thrust falls definitely if you if you said you know separate this wall it will narrow the uncertainty there's no question about it so you will uh it will still be fairly uncertain parameters but it is also interestingly we started to find out that that even sort of what we what we used to know we'll to think of a benign mechanism like a strike sleep can generate fairly deadly tsunamis and and that was the case in indonesia in 2018 when it's you know almost pure strike sleep fault for different reasons generated very deadly tsunami you know over 2000 people died uh in a very special setting of course so while you're absolutely right you can you know this this plot is to make the point and it's a little sort of it's not exaggerated but it's uh but it's put it's a lot a lot of historical earthquakes in a stronger earthquakes you know not all of them it's not not all 3000 earthquake but all strong earthquakes but it also uh um you know if you if you narrow it down you you cannot probably eliminate all the earthquakes you know just but but thrust falls because they still pause sexually and danger there but yes uh what what is my suggestion just during the break is uh i think it's your students or maybe it's slava students could color them these dots into red uh you know black and green and or or or or blue and so on meaning that you know the strike sleep faults it's a normal fault but normal i don't know the generator not but anyway this is a and it is interesting to see how clusters this earthquake and in which area might be this trust will be on the top and others down and so on but it's quite interesting actually yeah yes okay i am very sorry i would i wouldn't like to keep you and others and uh let's now make a break until 18 uh 55 it's okay sorry 17 a 55 i mean it's something like six minutes it's