[Script Info]
Title: 
[Events]
Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
Dialogue: 0,0:00:00.00,0:00:19.87,Default,,0000,0000,0000,,{\i1}36c3 preroll music{\i0}
Dialogue: 0,0:00:19.87,0:00:22.92,Default,,0000,0000,0000,,Herald: Ok, I have to say, I'm always\Ndeeply impressed about how much we already
Dialogue: 0,0:00:22.92,0:00:31.43,Default,,0000,0000,0000,,learned about space, about the universe\Nand about our place in the universe,
Dialogue: 0,0:00:31.43,0:00:37.23,Default,,0000,0000,0000,,our solar system. But the next speakers\Nwill explain us how we can use
Dialogue: 0,0:00:37.23,0:00:44.32,Default,,0000,0000,0000,,computational methods to simulate the\Nuniverse and actually grow planets. The
Dialogue: 0,0:00:44.32,0:00:49.53,Default,,0000,0000,0000,,speakers will be Anna Penzlin (miosta).\NShe is PHC student in computational
Dialogue: 0,0:00:49.53,0:00:55.23,Default,,0000,0000,0000,,astrophysics in Tübingen and Carolin\NKimmich (caro). She is a physics master's
Dialogue: 0,0:00:55.23,0:01:02.63,Default,,0000,0000,0000,,student at Heidelberg University. And the\Ntalk is entitled "Grow Your Own Planets
Dialogue: 0,0:01:02.63,0:01:07.69,Default,,0000,0000,0000,,How Simulations Help us understand the\Nuniverse." Thank you!
Dialogue: 0,0:01:07.69,0:01:14.74,Default,,0000,0000,0000,,{\i1}applause{\i0}
Dialogue: 0,0:01:14.74,0:01:24.72,Default,,0000,0000,0000,,caro: So hi, everyone. It's a cool\Nanimation right? And the really cool thing
Dialogue: 0,0:01:24.72,0:01:28.73,Default,,0000,0000,0000,,is that there's actually physics going on\Nthere. So this object could really be out
Dialogue: 0,0:01:28.73,0:01:35.38,Default,,0000,0000,0000,,there in space but was created on a\Ncomputer. So this is how a star is
Dialogue: 0,0:01:35.38,0:01:41.68,Default,,0000,0000,0000,,forming, how our solar system could have\Nlooked like in the beginning. Thank you
Dialogue: 0,0:01:41.68,0:01:47.44,Default,,0000,0000,0000,,for being here and that you're interested\Nin how we make such an animation. Anna and
Dialogue: 0,0:01:47.44,0:01:54.06,Default,,0000,0000,0000,,I are researchers in astrophysics. And\Nwe're concentrating on how planets form
Dialogue: 0,0:01:54.06,0:01:58.79,Default,,0000,0000,0000,,and evolve. She's doing her PHD and in\NTübingen and I'm doing my masters in
Dialogue: 0,0:01:58.79,0:02:04.03,Default,,0000,0000,0000,,Heidelberg. And in this talk, we want to\Nshow you a little bit of physics and how
Dialogue: 0,0:02:04.03,0:02:13.39,Default,,0000,0000,0000,,we can translate that in such a way that a\Ncomputer can calculate it. So, let's ask a
Dialogue: 0,0:02:13.39,0:02:19.42,Default,,0000,0000,0000,,question first. What is the universe or\Nwhat's in the universe? The most part of
Dialogue: 0,0:02:19.42,0:02:23.65,Default,,0000,0000,0000,,the universe is something we don't\Nunderstand, yet. It's dark matter and dark
Dialogue: 0,0:02:23.65,0:02:28.68,Default,,0000,0000,0000,,energy and we don't know what it is, yet.\NAnd that's everything we cannot see in
Dialogue: 0,0:02:28.68,0:02:35.12,Default,,0000,0000,0000,,this picture here. What we can see are\Nstars and galaxies, and that's what we
Dialogue: 0,0:02:35.12,0:02:39.98,Default,,0000,0000,0000,,want to concentrate on in this talk. But\Nif we can see it, why would we want to
Dialogue: 0,0:02:39.98,0:02:48.59,Default,,0000,0000,0000,,watch a computer? Well, everything in\Nastronomy takes a long time. So each of
Dialogue: 0,0:02:48.59,0:02:54.30,Default,,0000,0000,0000,,these tiny specs you see here are galaxies\Njust like ours. This is how the Milkyway
Dialogue: 0,0:02:54.30,0:02:59.56,Default,,0000,0000,0000,,looks like. And we are living in this tiny\Nspot here. And as you all know, our earth
Dialogue: 0,0:02:59.56,0:03:04.08,Default,,0000,0000,0000,,takes one year to orbit around the sun.\NNow, think about how long it takes for the
Dialogue: 0,0:03:04.08,0:03:10.46,Default,,0000,0000,0000,,sun to orbit around the center of the\Ngalaxy. It's four hundred million years.
Dialogue: 0,0:03:10.46,0:03:16.16,Default,,0000,0000,0000,,And even the star formation is 10 million\Nyears. We cannot wait 10 million years to
Dialogue: 0,0:03:16.16,0:03:23.52,Default,,0000,0000,0000,,watch how a star is forming, right? That's\Nwhy we need computational methods or
Dialogue: 0,0:03:23.52,0:03:29.73,Default,,0000,0000,0000,,simulations on a computer to understand\Nthese processes. So, when we watch to the
Dialogue: 0,0:03:29.73,0:03:35.97,Default,,0000,0000,0000,,night sky, what do we see? Of course we\Nsee stars and those beautiful nebulas.
Dialogue: 0,0:03:35.97,0:03:42.32,Default,,0000,0000,0000,,They are a gas and dust. And all of these\Nimages are taken with Hubble Space
Dialogue: 0,0:03:42.32,0:03:51.06,Default,,0000,0000,0000,,Telescope. Oh, so there's one image that\Ndoes belong in there. But it looks very
Dialogue: 0,0:03:51.06,0:03:56.81,Default,,0000,0000,0000,,similar, right? This gives us the idea\Nthat we can describe the gases in the
Dialogue: 0,0:03:56.81,0:04:04.84,Default,,0000,0000,0000,,universe as a fluid. It's really\Ncomplicated to describe the gas in every
Dialogue: 0,0:04:04.84,0:04:10.07,Default,,0000,0000,0000,,single particle. So, we cannot track every\Nsingle molecule in the gas that moves
Dialogue: 0,0:04:10.07,0:04:15.49,Default,,0000,0000,0000,,around. It's way easier to describe it as\Na fluid. So remember that for later, we
Dialogue: 0,0:04:15.49,0:04:22.31,Default,,0000,0000,0000,,will need that. But first, let's have a\Nlook how stars form. A star forms from a
Dialogue: 0,0:04:22.31,0:04:29.50,Default,,0000,0000,0000,,giant cloud of dust and gas. Everything\Nmoves in that cloud. So, eventually more
Dialogue: 0,0:04:29.50,0:04:38.70,Default,,0000,0000,0000,,dense regions occur and they get even\Ndenser. And these clams can eventually
Dialogue: 0,0:04:38.70,0:04:47.16,Default,,0000,0000,0000,,collapse to one star. So, this is how a\Nstar forms. They collapse due to their own
Dialogue: 0,0:04:47.16,0:04:53.81,Default,,0000,0000,0000,,gravity. And in this process, a disc\Nforms. And in this disc, planets can form.
Dialogue: 0,0:04:53.81,0:04:59.71,Default,,0000,0000,0000,,So why a disc? As I said, everything moves\Naround in the cloud. So it's likely that
Dialogue: 0,0:04:59.71,0:05:05.12,Default,,0000,0000,0000,,the cloud has a little bit of an initial\Nrotation. As it collapses, this rotation
Dialogue: 0,0:05:05.12,0:05:11.84,Default,,0000,0000,0000,,gets larger and faster. And now you can\Nthink of making a pizza. So when you make
Dialogue: 0,0:05:11.84,0:05:16.87,Default,,0000,0000,0000,,a pizza and spin your dough on your\Nfinger, you get a flat disc like a star,
Dialogue: 0,0:05:16.87,0:05:25.11,Default,,0000,0000,0000,,like a disc around a star. That's the same\Nprocess, actually. In this disc, we have
Dialogue: 0,0:05:25.11,0:05:31.33,Default,,0000,0000,0000,,dust and gas. From this dust in the disc\Nthe planet can form. But how do we get
Dialogue: 0,0:05:31.33,0:05:40.76,Default,,0000,0000,0000,,from tiny little dust particles to a big\Nplanet? Well, it somehow has to grow and
Dialogue: 0,0:05:40.76,0:05:46.32,Default,,0000,0000,0000,,grow even further and compact until we\Nhave rocks. And even grow further until we
Dialogue: 0,0:05:46.32,0:05:53.39,Default,,0000,0000,0000,,reach planets. How does it grow? Well,\Nthat dust grows we know that. At least
Dialogue: 0,0:05:53.39,0:06:00.72,Default,,0000,0000,0000,,that's what I observed when I took those\Nimages in my flat. Well, so dust can grow
Dialogue: 0,0:06:00.72,0:06:06.68,Default,,0000,0000,0000,,and grow even further and compact. But\Nwhen you take two rocks, we're now at this
Dialogue: 0,0:06:06.68,0:06:11.73,Default,,0000,0000,0000,,in this stage. When you take two rocks and\Nthrow them together, you don't expect them
Dialogue: 0,0:06:11.73,0:06:21.06,Default,,0000,0000,0000,,to stick, right? You expect them to crash\Nand crack into a thousand pieces. So,
Dialogue: 0,0:06:21.06,0:06:28.30,Default,,0000,0000,0000,,we're standing on the proof that planets\Nexist. How does this happen? And it's not
Dialogue: 0,0:06:28.30,0:06:34.70,Default,,0000,0000,0000,,quite solved yet in research. So, this is\Na process that is really hard to observe
Dialogue: 0,0:06:34.70,0:06:39.38,Default,,0000,0000,0000,,because planets are very, very tiny\Ncompared to stars. And even stars are only
Dialogue: 0,0:06:39.38,0:06:45.38,Default,,0000,0000,0000,,small dots in the night sky. Also, as I\Nsaid, planets form in a disc. And it's
Dialogue: 0,0:06:45.38,0:06:52.39,Default,,0000,0000,0000,,hard to look inside the disc. So this is\Nwhy we need computation to understand a
Dialogue: 0,0:06:52.39,0:06:58.67,Default,,0000,0000,0000,,process that how planets form and other\Nastronomical processes. So let's have a
Dialogue: 0,0:06:58.67,0:07:09.53,Default,,0000,0000,0000,,look at how this simulated on a computer.\Nmiosta: OK. So, somehow we have seen
Dialogue: 0,0:07:09.53,0:07:15.95,Default,,0000,0000,0000,,nature. It's beautiful and it's just like\Na tank of water and a bubbly fluid we
Dialogue: 0,0:07:15.95,0:07:21.00,Default,,0000,0000,0000,,already have. So, now we have this bubbly\Nfluid and here in the middle demonstrated.
Dialogue: 0,0:07:21.00,0:07:25.82,Default,,0000,0000,0000,,But now we have to teach our computer to\Ndeal with the bubbly fluid. And that's way
Dialogue: 0,0:07:25.82,0:07:31.60,Default,,0000,0000,0000,,too much single molecules to simulate\Nthem, as we already said. So there are two
Dialogue: 0,0:07:31.60,0:07:37.76,Default,,0000,0000,0000,,ways to discretize it in a way that we\Njust look at smaller pieces. One is the
Dialogue: 0,0:07:37.76,0:07:47.08,Default,,0000,0000,0000,,Lagrangian description, just like taking\Nsmall bubbles or balls of material that
Dialogue: 0,0:07:47.08,0:07:52.37,Default,,0000,0000,0000,,have a fixed mass. They have a certain\Nvelocity that varies between each particle
Dialogue: 0,0:07:52.37,0:07:57.18,Default,,0000,0000,0000,,and they have, of course, a momentum\Nbecause they have a velocity and a mass.
Dialogue: 0,0:07:57.18,0:08:01.63,Default,,0000,0000,0000,,And we've created a number of those\Nparticles and then just see how they move
Dialogue: 0,0:08:01.63,0:08:08.26,Default,,0000,0000,0000,,around and how they collide with each\Nother. That would be one way. And that was
Dialogue: 0,0:08:08.26,0:08:12.64,Default,,0000,0000,0000,,described last year in a very good talk. I\Ncan highly recommend to hear this talk if
Dialogue: 0,0:08:12.64,0:08:18.10,Default,,0000,0000,0000,,you're interested in this method. However,\Nthere's a second way to also describe
Dialogue: 0,0:08:18.10,0:08:23.17,Default,,0000,0000,0000,,this. Not just going with the flow of the\Nparticles, but we are a bit lazy, we just
Dialogue: 0,0:08:23.17,0:08:30.04,Default,,0000,0000,0000,,box it. So we create a grid. And as you\Nsee down here in this grid, you have the
Dialogue: 0,0:08:30.04,0:08:38.79,Default,,0000,0000,0000,,certain filling level, a bit of a slope.\NSo, what's the trend there? And then we
Dialogue: 0,0:08:38.79,0:08:44.91,Default,,0000,0000,0000,,just look for each box, what flows in what\Nflows out through the surfaces of this
Dialogue: 0,0:08:44.91,0:08:51.28,Default,,0000,0000,0000,,box. And then we have a volume or a mass\Nfilled within this box. And this is how we
Dialogue: 0,0:08:51.28,0:08:57.23,Default,,0000,0000,0000,,discretize what is going on in the disc.\NAnd actually, since we are usually in the
Dialogue: 0,0:08:57.23,0:09:04.22,Default,,0000,0000,0000,,system of a disc, we do not do it in this\Nnice box way like this. But we use boxes
Dialogue: 0,0:09:04.22,0:09:09.71,Default,,0000,0000,0000,,like those because they are already almost\Nlike a disc and we just keep exactly the
Dialogue: 0,0:09:09.71,0:09:14.78,Default,,0000,0000,0000,,same boxes all the time and you just\Nmeasure what goes through the surface in
Dialogue: 0,0:09:14.78,0:09:22.89,Default,,0000,0000,0000,,these boxes. So, this is how these two\Nmethods look like if you compute with both
Dialogue: 0,0:09:22.89,0:09:30.03,Default,,0000,0000,0000,,of them. So, one was done by me. I'm\Nusually using this boxing method and the
Dialogue: 0,0:09:30.03,0:09:35.96,Default,,0000,0000,0000,,other was done by my colleague. You see\Nthis like when you look at them, at the
Dialogue: 0,0:09:35.96,0:09:40.49,Default,,0000,0000,0000,,colors, at the structure here, you have\Nthe slope inwards, you have the same slope
Dialogue: 0,0:09:40.49,0:09:46.60,Default,,0000,0000,0000,,inwards here. You have even this silly\Nstructure here. The same here. But what
Dialogue: 0,0:09:46.60,0:09:52.64,Default,,0000,0000,0000,,you notice is you have this enlarge dots\Nthat are really the mass particles we saw
Dialogue: 0,0:09:52.64,0:09:57.98,Default,,0000,0000,0000,,before, these bubbles. And here you have\Nthis inner cutout. This is because when
Dialogue: 0,0:09:57.98,0:10:05.45,Default,,0000,0000,0000,,you create this grid, you have the very\Nregion at the inner part of the disc where
Dialogue: 0,0:10:05.45,0:10:11.41,Default,,0000,0000,0000,,the boxes become tiny and tinier. And\Nwell, we can't compute that. So, we have
Dialogue: 0,0:10:11.41,0:10:19.13,Default,,0000,0000,0000,,to cut out at some point in inner part So, here\Nwhen you go to low densities, these
Dialogue: 0,0:10:19.13,0:10:24.58,Default,,0000,0000,0000,,bubbles blow up and distribute their mass\Nover a larger area. So, it's not very
Dialogue: 0,0:10:24.58,0:10:30.84,Default,,0000,0000,0000,,accurate for these areas. And here we have\Nthe problem we can't calculate the inner
Dialogue: 0,0:10:30.84,0:10:38.59,Default,,0000,0000,0000,,area. So both methods have their pros and\Ncons. And are valid. But now, for most we
Dialogue: 0,0:10:38.59,0:10:51.46,Default,,0000,0000,0000,,will focus on this one. Just so we have\Nthis nice stream features. So, again,
Dialogue: 0,0:10:51.46,0:11:00.15,Default,,0000,0000,0000,,going back to the boxes, we have to\Nmeasure the flow between the boxes. This
Dialogue: 0,0:11:00.15,0:11:06.58,Default,,0000,0000,0000,,flow, in physics we call it flux, and we\Nhave a density row one, density row too.
Dialogue: 0,0:11:06.58,0:11:12.28,Default,,0000,0000,0000,,And the flux is the description of what\Nmass moves through the surface here from
Dialogue: 0,0:11:12.28,0:11:22.10,Default,,0000,0000,0000,,one box to the next. So, if we write this\Nin math terms, it looks like this. This
Dialogue: 0,0:11:22.10,0:11:36.56,Default,,0000,0000,0000,,says the time derivative of the density,\Nmeaning the change over time. So how much
Dialogue: 0,0:11:36.56,0:11:43.54,Default,,0000,0000,0000,,faster or slower, the velocity would be a\Nchange in time. And then this weird
Dialogue: 0,0:11:43.54,0:11:50.44,Default,,0000,0000,0000,,triangle symbol it's called nabla is a\Npositional derivative. So, it's like a
Dialogue: 0,0:11:50.44,0:12:00.62,Default,,0000,0000,0000,,slope. So, how do we change our position,\Nactually. So, if we change, look at the
Dialogue: 0,0:12:00.62,0:12:10.20,Default,,0000,0000,0000,,density over time, it should correlate to\Nwhat inflow we have over position. That is
Dialogue: 0,0:12:10.20,0:12:16.10,Default,,0000,0000,0000,,what that says. So and then we have in\Nphysics a few principles that we have
Dialogue: 0,0:12:16.10,0:12:21.92,Default,,0000,0000,0000,,always to obey because that is just almost\Ncommon sense. One of them is, well, if we
Dialogue: 0,0:12:21.92,0:12:29.76,Default,,0000,0000,0000,,have mass in a box. Well, like this, the\Nmass should not go anywhere unless someone
Dialogue: 0,0:12:29.76,0:12:35.20,Default,,0000,0000,0000,,takes it out. So, if we have a closed box\Nand mass in that box, nothing should
Dialogue: 0,0:12:35.20,0:12:42.62,Default,,0000,0000,0000,,disappear magically. It should all stay in\Nthis box. So, even if these particles jump
Dialogue: 0,0:12:42.62,0:12:48.05,Default,,0000,0000,0000,,around in our box with a certain velocity,\Nit's the same number of particles in the
Dialogue: 0,0:12:48.05,0:12:57.87,Default,,0000,0000,0000,,end. That's again, the same equation just\Ntold in math. So, a second very
Dialogue: 0,0:12:57.87,0:13:04.22,Default,,0000,0000,0000,,rudimentary principle is if we have energy\Nin it, in a completely closed box. So, for
Dialogue: 0,0:13:04.22,0:13:10.48,Default,,0000,0000,0000,,example, this nice chemicals here and we\Nhave a certain temperature. So, in this
Dialogue: 0,0:13:10.48,0:13:18.15,Default,,0000,0000,0000,,case, our temperature is low, maybe like\Noutside of around zero degree Celsius. And
Dialogue: 0,0:13:18.15,0:13:24.07,Default,,0000,0000,0000,,then we have this nice chemicals down here\Nand at some point they react very heavily.
Dialogue: 0,0:13:24.07,0:13:30.77,Default,,0000,0000,0000,,We suddenly end up with much less chemical\Nenergy and a lot more thermal energy. But
Dialogue: 0,0:13:30.77,0:13:36.94,Default,,0000,0000,0000,,overall, the complete energy summed up\Nhere, like the thermal and the chemical
Dialogue: 0,0:13:36.94,0:13:47.29,Default,,0000,0000,0000,,energy, also the energy of the movement\Nand the energy of potential added up to
Dialogue: 0,0:13:47.29,0:13:53.72,Default,,0000,0000,0000,,this variable "U". That should not change\Nover time if you sum up everything.
Dialogue: 0,0:13:53.72,0:13:59.50,Default,,0000,0000,0000,,Because our energy is conserved within our\Nclothed box. And then the third thing is I
Dialogue: 0,0:13:59.50,0:14:09.86,Default,,0000,0000,0000,,think you all know this. If you have like\Na small mass with a certain velocity, a
Dialogue: 0,0:14:09.86,0:14:14.18,Default,,0000,0000,0000,,very high velocity in this case and it\Nbumps into someone very large, what
Dialogue: 0,0:14:14.18,0:14:21.33,Default,,0000,0000,0000,,happens? Well, you get a very small\Nvelocity in this large body and the
Dialogue: 0,0:14:21.33,0:14:28.26,Default,,0000,0000,0000,,smaller mass stops. And the principle here\Nis that momentum is conserved, meaning
Dialogue: 0,0:14:28.26,0:14:36.68,Default,,0000,0000,0000,,that the velocity times the mass of one\Nobject is the same as then later for the
Dialogue: 0,0:14:36.68,0:14:42.70,Default,,0000,0000,0000,,other one. But since it's larger, this\Nproduct has to be the same. That doesn't
Dialogue: 0,0:14:42.70,0:14:49.38,Default,,0000,0000,0000,,change. And we have also in our\Nsimulations to obey these rules and we
Dialogue: 0,0:14:49.38,0:14:54.73,Default,,0000,0000,0000,,have to code that in so that we have\Nphysics in them. So you say, ok, this is
Dialogue: 0,0:14:54.73,0:14:59.45,Default,,0000,0000,0000,,really simple, these rules, right? But\Nactually, well, it's not quite as simple.
Dialogue: 0,0:14:59.45,0:15:03.88,Default,,0000,0000,0000,,So, this is the Navier-Stokes equation, a\Nvery complicated equation is not
Dialogue: 0,0:15:03.88,0:15:10.55,Default,,0000,0000,0000,,completely solved. And we have here all\Nthat is marked red are derivatives. Here
Dialogue: 0,0:15:10.55,0:15:16.23,Default,,0000,0000,0000,,we have our conservation law that was the\Nnice and simple part. But now we have to
Dialogue: 0,0:15:16.23,0:15:25.70,Default,,0000,0000,0000,,take other physical things into accounting\Nfor pressure, accounting for viscosity,
Dialogue: 0,0:15:25.70,0:15:33.37,Default,,0000,0000,0000,,for compression. So squeezing. And like\Nhow sticky is our fluid? And also gravity.
Dialogue: 0,0:15:33.37,0:15:38.79,Default,,0000,0000,0000,,So, we have a lot of additional factors,\Nadditional physics we also have to get in
Dialogue: 0,0:15:38.79,0:15:45.47,Default,,0000,0000,0000,,somehow. And all of these also depend\Nsomehow on the change of position or the
Dialogue: 0,0:15:45.47,0:15:51.85,Default,,0000,0000,0000,,change of time. And these derivatives\Naren't really nice for our computers
Dialogue: 0,0:15:51.85,0:15:57.37,Default,,0000,0000,0000,,because they well, they don't understand\Nthis triangle. So, we need to find a way
Dialogue: 0,0:15:57.37,0:16:03.92,Default,,0000,0000,0000,,to write an algorithm so that it can\Nsomehow relate with these math formula in
Dialogue: 0,0:16:03.92,0:16:14.92,Default,,0000,0000,0000,,a way that the computer likes. And one of\Nthe way to do this is, well, the simplest
Dialogue: 0,0:16:14.92,0:16:24.69,Default,,0000,0000,0000,,solution actually is just we say, OK, we\Nhave now this nasty derivatives and we
Dialogue: 0,0:16:24.69,0:16:32.16,Default,,0000,0000,0000,,want to get rid of them. So, if we look\Njust at one box now and we say that in
Dialogue: 0,0:16:32.16,0:16:42.17,Default,,0000,0000,0000,,this box, the new value for the density in\Nthis box would be the previous density,
Dialogue: 0,0:16:42.17,0:16:49.59,Default,,0000,0000,0000,,plus the flux in and out times the time\Nstepover which we measure this flux,
Dialogue: 0,0:16:49.59,0:16:58.26,Default,,0000,0000,0000,,right? So, and we have to somehow get to\Nthis flux and we just say, OK, this flux
Dialogue: 0,0:16:58.26,0:17:06.22,Default,,0000,0000,0000,,now is if we start here and the slope of\Nthis curve, the trends so to say, where
Dialogue: 0,0:17:06.22,0:17:10.55,Default,,0000,0000,0000,,this curve is going right now, it would\Nlook like this. So, in our next step, time
Dialogue: 0,0:17:10.55,0:17:19.14,Default,,0000,0000,0000,,step, we would have a density down here.\NAnd well, then we do this again. We again
Dialogue: 0,0:17:19.14,0:17:25.62,Default,,0000,0000,0000,,look at this point, where's the trend\Ngoing, where's the line going? And then we
Dialogue: 0,0:17:25.62,0:17:36.54,Default,,0000,0000,0000,,end up here. Same here. So, again, we just\Ntry to find this flax and this is the
Dialogue: 0,0:17:36.54,0:17:43.13,Default,,0000,0000,0000,,trend at this position in time. So, this\Ngoes up here. And then if we are here now,
Dialogue: 0,0:17:43.13,0:17:48.40,Default,,0000,0000,0000,,look at this point, it should go up here.\NSo this is what our next trend would be.
Dialogue: 0,0:17:48.40,0:17:55.27,Default,,0000,0000,0000,,And we do this over all the times. And\Nthis is how our simulation then would
Dialogue: 0,0:17:55.27,0:18:02.93,Default,,0000,0000,0000,,calculate the density for one box over a\Ndifferent time steps. So, that kind of
Dialogue: 0,0:18:02.93,0:18:09.25,Default,,0000,0000,0000,,works. So, the blue curve is the\Nanalytical one, the red curve, well it
Dialogue: 0,0:18:09.25,0:18:17.74,Default,,0000,0000,0000,,kind of similar, it works. But can we do\Nbetter? It's not perfect, yet, right? So,
Dialogue: 0,0:18:17.74,0:18:23.26,Default,,0000,0000,0000,,what we can do is we refine this a bit,\Ntaking a few more steps, making it a bit
Dialogue: 0,0:18:23.26,0:18:31.00,Default,,0000,0000,0000,,more computationally heavy, but trying to\Nget a better resolution. So, first we
Dialogue: 0,0:18:31.00,0:18:36.31,Default,,0000,0000,0000,,start with the same thing as before. We go\Nto this point, find the trend in this
Dialogue: 0,0:18:36.31,0:18:43.69,Default,,0000,0000,0000,,point. That point like the line would go\Nin this direction from this point. And
Dialogue: 0,0:18:43.69,0:18:51.53,Default,,0000,0000,0000,,then we go just half a step now. Sorry!\NAnd now we look at this half a step to
Dialogue: 0,0:18:51.53,0:18:57.65,Default,,0000,0000,0000,,this point now. And again, the same\Nsaying, OK, where's the trend going now?
Dialogue: 0,0:18:57.65,0:19:07.54,Default,,0000,0000,0000,,And then we take where this point would go\Nand added to this trend. So that would be
Dialogue: 0,0:19:07.54,0:19:14.18,Default,,0000,0000,0000,,that. The average of this trend, of this\Nexact point and this trend, this dark
Dialogue: 0,0:19:14.18,0:19:19.36,Default,,0000,0000,0000,,orange curve. And then we go back to the\Nbeginning with this trend now and say this
Dialogue: 0,0:19:19.36,0:19:24.26,Default,,0000,0000,0000,,is a better trend than the one we had\Nbefore. We now use that and go again and
Dialogue: 0,0:19:24.26,0:19:34.70,Default,,0000,0000,0000,,search the point for half a time step. And\Nthen again, we do the same thing. Now we
Dialogue: 0,0:19:34.70,0:19:42.46,Default,,0000,0000,0000,,again try to find actually the trend and\Naverage it with the arrow before. So it's
Dialogue: 0,0:19:42.46,0:19:46.32,Default,,0000,0000,0000,,not exactly the trend. It's a bit below\Nthe trend because we averaged it with the
Dialogue: 0,0:19:46.32,0:19:51.88,Default,,0000,0000,0000,,arrow before. And now we take this\Naveraging trend from the beginning to the
Dialogue: 0,0:19:51.88,0:19:57.08,Default,,0000,0000,0000,,top like this. Okay. This is already quite\Ngood, but we can still do a little bit
Dialogue: 0,0:19:57.08,0:20:02.57,Default,,0000,0000,0000,,better if we averaged with our ending\Npoint. So, we go here, look, where is the
Dialogue: 0,0:20:02.57,0:20:10.74,Default,,0000,0000,0000,,trend going that would go quite up like\Nthis and we average this and this together
Dialogue: 0,0:20:10.74,0:20:15.11,Default,,0000,0000,0000,,and then we end up with a line like this.\NThis is so much better than what we had
Dialogue: 0,0:20:15.11,0:20:22.92,Default,,0000,0000,0000,,before. It's a bit more complicated, to be\Nfair. But actually it's almost on the
Dialogue: 0,0:20:22.92,0:20:29.06,Default,,0000,0000,0000,,line. So, this is what we wanted. So, if\Nyou compare both of them, we have here our
Dialogue: 0,0:20:29.06,0:20:34.69,Default,,0000,0000,0000,,analytical curve. So, over time in one\Nbox, this is how the densities should
Dialogue: 0,0:20:34.69,0:20:39.91,Default,,0000,0000,0000,,increase. And now with it both of the\Nnumerical method, the difference looks
Dialogue: 0,0:20:39.91,0:20:46.05,Default,,0000,0000,0000,,like this. So, if we have smaller and\Nsmaller time steps, even the Euler gets
Dialogue: 0,0:20:46.05,0:20:55.75,Default,,0000,0000,0000,,closer and closer to the curve. But\Nactually the Runge Kutta this four step process
Dialogue: 0,0:20:55.75,0:21:00.62,Default,,0000,0000,0000,,works much better and much faster.\NHowever, it's a bit more computationally
Dialogue: 0,0:21:00.62,0:21:08.37,Default,,0000,0000,0000,,and difficult.\Ncaro: When we simulate objects in
Dialogue: 0,0:21:08.37,0:21:15.04,Default,,0000,0000,0000,,astronomy, we always want to compare that\Nto objects that are really out there. So,
Dialogue: 0,0:21:15.04,0:21:20.49,Default,,0000,0000,0000,,this is a giant telescope, well consisting\Nof a lot of small telescopes. But they can
Dialogue: 0,0:21:20.49,0:21:27.01,Default,,0000,0000,0000,,be connected and used as a giant telescope\Nand it takes photos of dust in the sky.
Dialogue: 0,0:21:27.01,0:21:34.16,Default,,0000,0000,0000,,And this is used to take images of discs\Naround stars. And these discs look like
Dialogue: 0,0:21:34.16,0:21:41.05,Default,,0000,0000,0000,,this. So, these images were taken last\Nyear and they are really cool. Before we
Dialogue: 0,0:21:41.05,0:21:46.12,Default,,0000,0000,0000,,had those images, we only had images with\Nless resolution. So, they were just
Dialogue: 0,0:21:46.12,0:21:52.12,Default,,0000,0000,0000,,blurred blobs. And we could say, yeah,\Nthat might be a disc. But now we really
Dialogue: 0,0:21:52.12,0:21:58.66,Default,,0000,0000,0000,,see the discs and we see rings here, thin\Nrings and we see thicker rings over here.
Dialogue: 0,0:21:58.66,0:22:05.59,Default,,0000,0000,0000,,And even some spiraly structures here. And\Nalso some features that are not really
Dialogue: 0,0:22:05.59,0:22:11.99,Default,,0000,0000,0000,,radial symmetric like this arc here. And\Nit's not completely solved how these
Dialogue: 0,0:22:11.99,0:22:24.26,Default,,0000,0000,0000,,structures formed. And to find that out a\Ncolleague of mine took this little object
Dialogue: 0,0:22:24.26,0:22:30.80,Default,,0000,0000,0000,,with the asymmetry here. And so, this is\Nimage we just saw. And this is his
Dialogue: 0,0:22:30.80,0:22:37.59,Default,,0000,0000,0000,,simulation. So, this is how the disc\Nlooked like in the beginning, probably.
Dialogue: 0,0:22:37.59,0:22:43.98,Default,,0000,0000,0000,,And we put in three planets and let the\Nsimulation run. And so, what we see here
Dialogue: 0,0:22:43.98,0:22:52.03,Default,,0000,0000,0000,,is that the star is cut out as Anna said.\NSo, the grid cells in the inner part are
Dialogue: 0,0:22:52.03,0:22:56.69,Default,,0000,0000,0000,,very, very small. And it would take a long\Ntime to compute them all. So, that's why
Dialogue: 0,0:22:56.69,0:23:06.78,Default,,0000,0000,0000,,we're leaving out that spot in the middle.\NAnd what we see here is three planets
Dialogue: 0,0:23:06.78,0:23:16.31,Default,,0000,0000,0000,,interacting with the material in the disc.\NAnd we can see that these planets can make
Dialogue: 0,0:23:16.31,0:23:24.44,Default,,0000,0000,0000,,this thing here appear so that in the end\Nwe have something looking very similar to
Dialogue: 0,0:23:24.44,0:23:30.70,Default,,0000,0000,0000,,what we want to have or what we really\Nobserve. So, we can say three planets
Dialogue: 0,0:23:30.70,0:23:37.38,Default,,0000,0000,0000,,could explain how these structures formed\Nin this disc. It's a little bit
Dialogue: 0,0:23:37.38,0:23:42.41,Default,,0000,0000,0000,,elliptical, you see that. That's because\Nit's tilted from our side of line. It
Dialogue: 0,0:23:42.41,0:23:47.43,Default,,0000,0000,0000,,would be round if you watched at it face\Non. But it's a little bit tilted. That's
Dialogue: 0,0:23:47.43,0:23:55.27,Default,,0000,0000,0000,,why it looks elliptical.\Nmiosta: So, we already saw we can put
Dialogue: 0,0:23:55.27,0:24:02.08,Default,,0000,0000,0000,,planets in the gas and then we create\Nstructures. One very exciting thing that
Dialogue: 0,0:24:02.08,0:24:08.74,Default,,0000,0000,0000,,we found in the last year - or two years\Nago it started but then we found more - is
Dialogue: 0,0:24:08.74,0:24:15.69,Default,,0000,0000,0000,,this system PDS 70. In this system, for\Nthe very first time, we found a planet
Dialogue: 0,0:24:15.69,0:24:24.25,Default,,0000,0000,0000,,that was still embedded completely within\Nthe disc. So, the gas and dust. Usually,
Dialogue: 0,0:24:24.25,0:24:32.26,Default,,0000,0000,0000,,because the gas and dust is the main thing\Nthat creates this signal of some radiation
Dialogue: 0,0:24:32.26,0:24:37.75,Default,,0000,0000,0000,,because of heat. We only observe that and\Nthen we can't observe the planet embedded.
Dialogue: 0,0:24:37.75,0:24:41.63,Default,,0000,0000,0000,,But in this case, the planet was large\Nenough. And in the right position that we
Dialogue: 0,0:24:41.63,0:24:48.94,Default,,0000,0000,0000,,actually were able to observe some\Nsignature of accretion on this planet that
Dialogue: 0,0:24:48.94,0:24:57.44,Default,,0000,0000,0000,,was brighter than the rest of the disc.\NAnd then later, just this year, just a few
Dialogue: 0,0:24:57.44,0:25:03.74,Default,,0000,0000,0000,,months ago, we actually found out well,\Nthis is not the only object here. This is
Dialogue: 0,0:25:03.74,0:25:10.85,Default,,0000,0000,0000,,very clearly a planet. But actually,\Nlike this spot here is also something. So,
Dialogue: 0,0:25:10.85,0:25:17.30,Default,,0000,0000,0000,,we can see it in different grains. Every\Npicture here is a different set of grains
Dialogue: 0,0:25:17.30,0:25:26.95,Default,,0000,0000,0000,,observed. And we can see \Nthis in five different kinds of
Dialogue: 0,0:25:26.95,0:25:32.80,Default,,0000,0000,0000,,observations. So, there is a planet here.\NAnd then there is also something we don't
Dialogue: 0,0:25:32.80,0:25:37.71,Default,,0000,0000,0000,,know what it is yet, but its point like\Nand actually creates the feature that we
Dialogue: 0,0:25:37.71,0:25:43.24,Default,,0000,0000,0000,,reproduce in different kinds of\Nobservational bands or different kinds of
Dialogue: 0,0:25:43.24,0:25:52.07,Default,,0000,0000,0000,,signals here. This is very interesting.\NFor the first time, we actually see a
Dialogue: 0,0:25:52.07,0:25:58.03,Default,,0000,0000,0000,,planet forming right now within the disc.\NAnd so a colleague of mine also is very
Dialogue: 0,0:25:58.03,0:26:04.93,Default,,0000,0000,0000,,interested in the system and started to\Nsimulate how do two planets in a disc
Dialogue: 0,0:26:04.93,0:26:13.15,Default,,0000,0000,0000,,change the dynamics of a disc? So here we\Nhave, of course, this disc is again tilted
Dialogue: 0,0:26:13.15,0:26:20.23,Default,,0000,0000,0000,,because it's not phase on, it's like 45\Ndegrees tilted, not like this, but like
Dialogue: 0,0:26:20.23,0:26:27.29,Default,,0000,0000,0000,,this. And so he had it face on. This is\Nwhat a simulation looks like. So, there
Dialogue: 0,0:26:27.29,0:26:33.88,Default,,0000,0000,0000,,are two planets: these blobs here, again,\Nas in this simulation. Here we have a
Dialogue: 0,0:26:33.88,0:26:39.29,Default,,0000,0000,0000,,close up. You can actually see this little\Nboxes are actually our simulation boxes in
Dialogue: 0,0:26:39.29,0:26:47.43,Default,,0000,0000,0000,,which we have our own densities. And then\Nhe just looked at how the planets would
Dialogue: 0,0:26:47.43,0:26:52.62,Default,,0000,0000,0000,,change the structure and the gas and also\Nhow the gas would interact with the
Dialogue: 0,0:26:52.62,0:26:59.25,Default,,0000,0000,0000,,planets, shifting them around. And it's\Ninteresting. So, the planets tend to clear
Dialogue: 0,0:26:59.25,0:27:05.26,Default,,0000,0000,0000,,out an area, open a gap, and within the\Ndisk, that block has a lot of gas around
Dialogue: 0,0:27:05.26,0:27:11.04,Default,,0000,0000,0000,,here. So, you have the brighter ring here\Nagain and then clearing out more and more.
Dialogue: 0,0:27:11.04,0:27:23.39,Default,,0000,0000,0000,,And at some point in the simulation you\Nsaw they get a bit jumpy. So it's very nice.
Dialogue: 0,0:27:23.39,0:27:29.57,Default,,0000,0000,0000,,You also see that planets induce in the\Nwhole disc some kind of features like
Dialogue: 0,0:27:29.57,0:27:36.74,Default,,0000,0000,0000,,spiral features. And so a single planet\Nwill change the symmetry and the
Dialogue: 0,0:27:36.74,0:27:40.99,Default,,0000,0000,0000,,appearance of a whole disc.\Ncaro: So, the reason why the planet is
Dialogue: 0,0:27:40.99,0:27:46.49,Default,,0000,0000,0000,,staying at this point is because we're\Nrotating with the planet. So it's actually
Dialogue: 0,0:27:46.49,0:27:53.50,Default,,0000,0000,0000,,going around the disc, but the like camera\Nis rotating with the planet. So, it's
Dialogue: 0,0:27:53.50,0:28:00.36,Default,,0000,0000,0000,,staying at that fixed place we put it in.\Nmiosta: Exactly. But there's more because
Dialogue: 0,0:28:00.36,0:28:04.60,Default,,0000,0000,0000,,as I already said, in the Navier-Stokes\Nequation, we have a lot of different kinds
Dialogue: 0,0:28:04.60,0:28:08.97,Default,,0000,0000,0000,,of physics that we all have to include in\Nour simulations. One of the things, of
Dialogue: 0,0:28:08.97,0:28:15.15,Default,,0000,0000,0000,,course, is we maybe don't have just a star\Nand a disc. We have planets in there and
Dialogue: 0,0:28:15.15,0:28:20.60,Default,,0000,0000,0000,,maybe two stars in there. And all of these\Nlarger bodies have also an interaction
Dialogue: 0,0:28:20.60,0:28:27.36,Default,,0000,0000,0000,,between each other. So, if we have the\Nstar, every planet will have an
Dialogue: 0,0:28:27.36,0:28:32.60,Default,,0000,0000,0000,,interaction with the star, of course. But\Nthen also the planets between each other,
Dialogue: 0,0:28:32.60,0:28:40.38,Default,,0000,0000,0000,,they have also an interaction, right? So,\Nin the end, you have to take into account
Dialogue: 0,0:28:40.38,0:28:48.82,Default,,0000,0000,0000,,all of these interactions. And then also\Nwe have accretion just looking like this.
Dialogue: 0,0:28:48.82,0:28:59.35,Default,,0000,0000,0000,,So, accretion means that the gas is bound\Nby some objects. It can be the disc, the
Dialogue: 0,0:28:59.35,0:29:07.01,Default,,0000,0000,0000,,planet or the star that takes up the mass,\Nthe dust or the gas and bounce it to this
Dialogue: 0,0:29:07.01,0:29:14.84,Default,,0000,0000,0000,,object. And then it's lost to the disc or\Nthe other structures because it's
Dialogue: 0,0:29:14.84,0:29:22.31,Default,,0000,0000,0000,,completely bound to that. So, the\Nprinciple of this would be the simulation
Dialogue: 0,0:29:22.31,0:29:29.28,Default,,0000,0000,0000,,I did last year and published, we have\Nhere a binary star. So, these two dots are
Dialogue: 0,0:29:29.28,0:29:38.81,Default,,0000,0000,0000,,stars. I kind of kept them in the same\Nspot. But every picture will be one orbit
Dialogue: 0,0:29:38.81,0:29:42.76,Default,,0000,0000,0000,,of this binary, but since we have\Ninteractions, you actually see them
Dialogue: 0,0:29:42.76,0:29:48.54,Default,,0000,0000,0000,,rotating because of the interactions, with\Neach other. And then also we have here a
Dialogue: 0,0:29:48.54,0:29:52.67,Default,,0000,0000,0000,,planet and here a planet. And the\Ninteresting thing was that these two
Dialogue: 0,0:29:52.67,0:30:00.36,Default,,0000,0000,0000,,planets interact in such a way that they\Nend up on exactly the same orbit. So, one
Dialogue: 0,0:30:00.36,0:30:06.18,Default,,0000,0000,0000,,star's further out, the orange one, and then\Nvery fast they go in. And they end up on
Dialogue: 0,0:30:06.18,0:30:28.42,Default,,0000,0000,0000,,exactly the same orbit. If it now play nicely. \NSo, another thing is with the accretion here,
Dialogue: 0,0:30:28.42,0:30:36.60,Default,,0000,0000,0000,,we actually see clouds from above dropping\Ndown onto the new forming star here. So,
Dialogue: 0,0:30:36.60,0:30:44.41,Default,,0000,0000,0000,,all of this, what you see here would be\Ngas, hydrogen. And it's a very early phase
Dialogue: 0,0:30:44.41,0:30:49.50,Default,,0000,0000,0000,,so that disc is not completely flat. It\Nhas a lot of material. And then we
Dialogue: 0,0:30:49.50,0:30:55.78,Default,,0000,0000,0000,,actually have this infall from above\Ntowards the star and then the star keeps
Dialogue: 0,0:30:55.78,0:31:01.77,Default,,0000,0000,0000,,the mass. And we have to take this also\Ninto account in our simulations. Another
Dialogue: 0,0:31:01.77,0:31:07.22,Default,,0000,0000,0000,,thing we have to take into account up till\Nnow, we just cared about masses and
Dialogue: 0,0:31:07.22,0:31:12.93,Default,,0000,0000,0000,,densities. But of course what we actually\Nsee is that stars are kind of warm,
Dialogue: 0,0:31:12.93,0:31:21.76,Default,,0000,0000,0000,,hopefully. Otherwise, temperatures here\Nwould also not be nice. And different
Dialogue: 0,0:31:21.76,0:31:27.93,Default,,0000,0000,0000,,chemicals have different condensation\Npoints. And this is also true in a system.
Dialogue: 0,0:31:27.93,0:31:35.02,Default,,0000,0000,0000,,So, we start with the start temperature at\Nthe surface of the star. We have a
Dialogue: 0,0:31:35.02,0:31:41.48,Default,,0000,0000,0000,,temperature around 4.000 Kelvin. And then\Nwe go a bit into the disc. And there is a
Dialogue: 0,0:31:41.48,0:31:47.89,Default,,0000,0000,0000,,point where we for the first time reach a\Npoint where we have any material at all.
Dialogue: 0,0:31:47.89,0:31:52.17,Default,,0000,0000,0000,,Because it starts to condensate and we\Nactually have something solid like iron.
Dialogue: 0,0:31:52.17,0:31:58.18,Default,,0000,0000,0000,,For example, at a 1500 Kelvin. And then if\Nwe go further in, we reach a point where
Dialogue: 0,0:31:58.18,0:32:07.69,Default,,0000,0000,0000,,we have solid water and this is at 200\NKelvin. This is what we then would need
Dialogue: 0,0:32:07.69,0:32:12.59,Default,,0000,0000,0000,,actually to have a planet that also has\Nwater on it. Because if we don't get the
Dialogue: 0,0:32:12.59,0:32:18.89,Default,,0000,0000,0000,,water in the solid state, it will not fall\Nonto a terrestrial planet and be bound
Dialogue: 0,0:32:18.89,0:32:24.90,Default,,0000,0000,0000,,there, right? So, this is important for\Nour Earth, actually. And then if we go
Dialogue: 0,0:32:24.90,0:32:33.34,Default,,0000,0000,0000,,even further out, we have also other gases\Ncondensating to solids like CO2 or methane
Dialogue: 0,0:32:33.34,0:32:40.59,Default,,0000,0000,0000,,or things like that. And since we only get\Nwater on a planet when we have a
Dialogue: 0,0:32:40.59,0:32:48.01,Default,,0000,0000,0000,,temperature that is low enough so that the\Nwater actually forms is solid and it's
Dialogue: 0,0:32:48.01,0:32:54.27,Default,,0000,0000,0000,,important for us to think about where that\Nis in our forming disc. Where do we start?
Dialogue: 0,0:32:54.27,0:32:59.77,Default,,0000,0000,0000,,We have a planet like Earth that could\Nhave some water, right? But it's not just
Dialogue: 0,0:32:59.77,0:33:07.57,Default,,0000,0000,0000,,the simple picture, where we have all these\Nnice ring structures, where we have a clear
Dialogue: 0,0:33:07.57,0:33:13.62,Default,,0000,0000,0000,,line. Actually, it gets more complicated\Nbecause we have pressure and shocks. And
Dialogue: 0,0:33:13.62,0:33:19.54,Default,,0000,0000,0000,,thermodynamics is a lot like pogo dancing,\Nright? You crash into each other. And it's
Dialogue: 0,0:33:19.54,0:33:25.63,Default,,0000,0000,0000,,all about collisions. So, the gas\Ntemperature is determined by the speed of
Dialogue: 0,0:33:25.63,0:33:31.30,Default,,0000,0000,0000,,your gas molecules. Like you bouncing and\Ncrashing into each other, exchanging
Dialogue: 0,0:33:31.30,0:33:39.34,Default,,0000,0000,0000,,momentum. So, there's two ways to heat up\Nsuch dance. First thing is you get a large
Dialogue: 0,0:33:39.34,0:33:45.94,Default,,0000,0000,0000,,amount of velocity from the outside like a\Nhuge kick, a shock into your system. A
Dialogue: 0,0:33:45.94,0:33:51.52,Default,,0000,0000,0000,,second way would be if we have a higher\Npressure, like more molecules, then also
Dialogue: 0,0:33:51.52,0:33:55.91,Default,,0000,0000,0000,,you of course have more collisions and\Nthen a higher temperature. So, if you
Dialogue: 0,0:33:55.91,0:34:02.53,Default,,0000,0000,0000,,change - because you have a planet now in\Nthe system - the pressure at some point,
Dialogue: 0,0:34:02.53,0:34:08.70,Default,,0000,0000,0000,,you actually get a higher temperature. So,\Nthat is not then we lose this nice line
Dialogue: 0,0:34:08.70,0:34:19.14,Default,,0000,0000,0000,,because suddenly we have different\Npressures at different locations. And a
Dialogue: 0,0:34:19.14,0:34:24.70,Default,,0000,0000,0000,,colleague of mine also simulated this. \NSo, this is the initial condition we
Dialogue: 0,0:34:24.70,0:34:28.86,Default,,0000,0000,0000,,just assumed: OK, if we have no\Ndisturbance whatsoever, we have our nice
Dialogue: 0,0:34:28.86,0:34:36.89,Default,,0000,0000,0000,,planet here at 1au. So, same distance as\Nearth to the sun. Here, too. But here we
Dialogue: 0,0:34:36.89,0:34:46.67,Default,,0000,0000,0000,,assume that less heat gets transferred\Nfrom the surface of the disc. And here we
Dialogue: 0,0:34:46.67,0:34:52.03,Default,,0000,0000,0000,,have the planet far out like Jupiter or\Nsomething. And now we actually let this
Dialogue: 0,0:34:52.03,0:34:59.59,Default,,0000,0000,0000,,planet change the structure of the disc.\NAnd what happens is - we found these spirals
Dialogue: 0,0:34:59.59,0:35:05.80,Default,,0000,0000,0000,,and within these spirals, we change\Npressure. And with this actually, if you
Dialogue: 0,0:35:05.80,0:35:11.59,Default,,0000,0000,0000,,see this orange, everywhere where it's\Norange it's hotter than the iceline. So,
Dialogue: 0,0:35:11.59,0:35:17.02,Default,,0000,0000,0000,,we don't have water where it's orange. And\Nwhere it's blue we can have water. And the
Dialogue: 0,0:35:17.02,0:35:22.35,Default,,0000,0000,0000,,interesting thing is, even if we put a\Nplanet out here like Jupiter, we still
Dialogue: 0,0:35:22.35,0:35:32.57,Default,,0000,0000,0000,,form these regions in the inner system\Nwhere we have less water.
Dialogue: 0,0:35:32.57,0:35:38.02,Default,,0000,0000,0000,,caro: One problem in astrophysical\Nsimulations is that we don't always know
Dialogue: 0,0:35:38.02,0:35:47.94,Default,,0000,0000,0000,,how to shape our boxes or how small these\Nboxes have to be. So, we use a trick to
Dialogue: 0,0:35:47.94,0:35:54.67,Default,,0000,0000,0000,,reshape the boxes as we need them. It's\Ncalled adaptive mesh. And this is a
Dialogue: 0,0:35:54.67,0:35:58.89,Default,,0000,0000,0000,,simulation of the red fluid flowing in\Nthis direction and the blue fluid in the
Dialogue: 0,0:35:58.89,0:36:06.58,Default,,0000,0000,0000,,other one. So, at the boundary, the two\Nfluid shear and they mix up somehow and we
Dialogue: 0,0:36:06.58,0:36:12.99,Default,,0000,0000,0000,,don't know how in advance. So, we start a\Nsimulation and as the simulation starts,
Dialogue: 0,0:36:12.99,0:36:19.64,Default,,0000,0000,0000,,we reshape those boxes here. So, in the\Nmiddle we don't need much. We reshape
Dialogue: 0,0:36:19.64,0:36:25.40,Default,,0000,0000,0000,,because it's not that complicated here.\NIt's just the flow. But at the boundary we
Dialogue: 0,0:36:25.40,0:36:35.06,Default,,0000,0000,0000,,see those mixing up of the two fluids. And\Nso, we reshape the cells as we need them.
Dialogue: 0,0:36:35.06,0:36:44.76,Default,,0000,0000,0000,,This is done in a program, in an\Nastrophysical program called AREPO. We
Dialogue: 0,0:36:44.76,0:36:52.75,Default,,0000,0000,0000,,will later show you some more programs to\Nuse for simulations. But another
Dialogue: 0,0:36:52.75,0:36:59.02,Default,,0000,0000,0000,,simulation I want to show you first is\Nalso done with AREPO and it's a simulation
Dialogue: 0,0:36:59.02,0:37:04.71,Default,,0000,0000,0000,,of the universe. So, from here to here,\Nit's very big. It's 30 million light
Dialogue: 0,0:37:04.71,0:37:12.21,Default,,0000,0000,0000,,years. So each of these dots you see here\Nis the size of a galaxy or even more. And
Dialogue: 0,0:37:12.21,0:37:17.84,Default,,0000,0000,0000,,here you can actually see that at some\Nregions it's very empty. So, we're
Dialogue: 0,0:37:17.84,0:37:23.42,Default,,0000,0000,0000,,rotating around this universe, this\Nsimulated universe here. And these regions
Dialogue: 0,0:37:23.42,0:37:28.99,Default,,0000,0000,0000,,here are empty. And we don't need a lot of\Nboxes there. The big boxes are enough
Dialogue: 0,0:37:28.99,0:37:35.01,Default,,0000,0000,0000,,here. But in this dense regions where we\Nhave a lot of material, we need smaller
Dialogue: 0,0:37:35.01,0:37:42.38,Default,,0000,0000,0000,,boxes. And this method I showed you where\Nwe reshape the boxes as we need them is
Dialogue: 0,0:37:42.38,0:37:53.42,Default,,0000,0000,0000,,used for this simulation.\Nmiosta: So, actually, you see the
Dialogue: 0,0:37:53.42,0:37:56.34,Default,,0000,0000,0000,,beginning of the universe there.\Ncaro: Yes!
Dialogue: 0,0:37:56.34,0:38:01.00,Default,,0000,0000,0000,,miosta: Basically, the initial mass\Ncollapsing to the first galaxies and first
Dialogue: 0,0:38:01.00,0:38:07.03,Default,,0000,0000,0000,,supernovae starting. Very beautiful\Nsimulation.
Dialogue: 0,0:38:07.03,0:38:19.82,Default,,0000,0000,0000,,caro: So, there are different programs, as\NI already mentioned, in astrophysics.
Dialogue: 0,0:38:19.82,0:38:24.97,Default,,0000,0000,0000,,Three of them, those three are all open\Nsource, so you can download them and use
Dialogue: 0,0:38:24.97,0:38:31.09,Default,,0000,0000,0000,,them on your own machine, if you like. But\Nthere are more, a lot more. Some of them
Dialogue: 0,0:38:31.09,0:38:38.63,Default,,0000,0000,0000,,are open source, some of them are not.\NSometimes it's hard to get them. In the
Dialogue: 0,0:38:38.63,0:38:43.70,Default,,0000,0000,0000,,following, we will present the tool\NFARGO3D and PLUTO in a detailed version or
Dialogue: 0,0:38:43.70,0:38:53.16,Default,,0000,0000,0000,,a more detailed vision than AREPO \Nbecause we usually use those two for our
Dialogue: 0,0:38:53.16,0:38:58.38,Default,,0000,0000,0000,,simulations. What I want to show you with\Nthis slide is that depending on what you
Dialogue: 0,0:38:58.38,0:39:04.52,Default,,0000,0000,0000,,want to simulate, you need to choose a\Ndifferent program. And one thing is that
Dialogue: 0,0:39:04.52,0:39:10.25,Default,,0000,0000,0000,,in astrophysics we sometimes call the\Nwhole program code. So, if I use the word
Dialogue: 0,0:39:10.25,0:39:19.17,Default,,0000,0000,0000,,code. Sorry about that. I mean, the whole\Nprogram. So, let's have a look at FARGO3D.
Dialogue: 0,0:39:19.17,0:39:27.87,Default,,0000,0000,0000,,It's a hydro dynamics code and what you\Nsee here is an input parameter file. There
Dialogue: 0,0:39:27.87,0:39:35.18,Default,,0000,0000,0000,,you define how the disc looks like. How\Nmuch mass does it have? How big is it? And
Dialogue: 0,0:39:35.18,0:39:43.14,Default,,0000,0000,0000,,what planet? So, here at Jupiter, do you\Nsee that? Jupiter is put in. And we also
Dialogue: 0,0:39:43.14,0:39:51.28,Default,,0000,0000,0000,,define how big our boxes are. This\Nprogram is written in C, which is quite
Dialogue: 0,0:39:51.28,0:39:57.50,Default,,0000,0000,0000,,nice because a lot of astrophysical\Nprograms are still written in Fortran. So,
Dialogue: 0,0:39:57.50,0:40:05.60,Default,,0000,0000,0000,,this is good for me because I don't know\Nany Fortran. We can run this and what's
Dialogue: 0,0:40:05.60,0:40:11.01,Default,,0000,0000,0000,,typical for FARGO3D. So that's a compilation\Nactually on my computer. So, I don't need
Dialogue: 0,0:40:11.01,0:40:18.84,Default,,0000,0000,0000,,a fancy computer. I just did it on my\Nsmall laptop and now we run it. Now,
Dialogue: 0,0:40:18.84,0:40:26.13,Default,,0000,0000,0000,,typical for FARGO3D, as you will see are lot\Nof dots. So, here it will print out a lot
Dialogue: 0,0:40:26.13,0:40:33.81,Default,,0000,0000,0000,,of dots and it will create at certain\Ntimes some outputs. And these outputs are
Dialogue: 0,0:40:33.81,0:40:38.30,Default,,0000,0000,0000,,huge files containing numbers. So, if you\Nlook at them they are not really
Dialogue: 0,0:40:38.30,0:40:44.29,Default,,0000,0000,0000,,interesting. They just are a numbers in\Nsomething like a text file. So, a big part
Dialogue: 0,0:40:44.29,0:40:50.43,Default,,0000,0000,0000,,of astrophysics is also to visualize the\Ndata. Not only to create it but also to
Dialogue: 0,0:40:50.43,0:40:57.08,Default,,0000,0000,0000,,make images so that we can make movies out\Nof them. For that, I prefer to use Python
Dialogue: 0,0:40:57.08,0:41:01.60,Default,,0000,0000,0000,,but there are a lot of tools Python\NMatplotlib, but there are a lot of
Dialogue: 0,0:41:01.60,0:41:09.29,Default,,0000,0000,0000,,different tools to visualize the data. So,\Nthis is actually that output. That first
Dialogue: 0,0:41:09.29,0:41:16.35,Default,,0000,0000,0000,,one we just saw. The Jupiter planet in the\Ndisc that I defined in this parameter file
Dialogue: 0,0:41:16.35,0:41:23.28,Default,,0000,0000,0000,,and it's already started to do some\Nspirals. And if I would have let it
Dialogue: 0,0:41:23.28,0:41:33.68,Default,,0000,0000,0000,,run further than the spirals were more\Nprominent. And yeah, now we have a planet
Dialogue: 0,0:41:33.68,0:41:45.23,Default,,0000,0000,0000,,here on our computer.\Nmiosta: OK, so we also have PLUTO. PLUTO
Dialogue: 0,0:41:45.23,0:41:53.59,Default,,0000,0000,0000,,somehow has a bit more setup files. So,\Nwhat I need is three files here. Looks a
Dialogue: 0,0:41:53.59,0:41:59.32,Default,,0000,0000,0000,,bit complicated to break it down. This\Nfile defines my grid and initial values.
Dialogue: 0,0:41:59.32,0:42:04.77,Default,,0000,0000,0000,,And this simulation time here we input\Nactually what physics do we want to need?
Dialogue: 0,0:42:04.77,0:42:13.02,Default,,0000,0000,0000,,What is our coordinate system? So, do we\Nwant to have a disc or just like spherical
Dialogue: 0,0:42:13.02,0:42:20.66,Default,,0000,0000,0000,,boxes or like squared boxes? And how is\Nthe time defined? And here we then
Dialogue: 0,0:42:20.66,0:42:26.72,Default,,0000,0000,0000,,actually write a bit of code to say, OK,\Nnow how do I want a gravitational
Dialogue: 0,0:42:26.72,0:42:34.58,Default,,0000,0000,0000,,potential? So, what's the source of\Ngravity or what will happen at the inner
Dialogue: 0,0:42:34.58,0:42:39.89,Default,,0000,0000,0000,,region where we have this dark spot? We\Nhave somehow to define what happens if gas
Dialogue: 0,0:42:39.89,0:42:45.09,Default,,0000,0000,0000,,reaches this boundary. Is it just falling\Nin? Is it bouncing back or something? Or
Dialogue: 0,0:42:45.09,0:42:50.89,Default,,0000,0000,0000,,is it looping through the one end to the\Nnext? This is also something we then just
Dialogue: 0,0:42:50.89,0:43:01.53,Default,,0000,0000,0000,,have to code in. And if we then make it\Nand let run, it looks like this. So,
Dialogue: 0,0:43:01.53,0:43:08.59,Default,,0000,0000,0000,,again, our nice thing we hopefully put in\Nor wanted to put in: the time steps, what
Dialogue: 0,0:43:08.59,0:43:14.89,Default,,0000,0000,0000,,our boundaries were, parameters of\Nphysics. Hopefully, the right ones and
Dialogue: 0,0:43:14.89,0:43:21.40,Default,,0000,0000,0000,,then nicely we start with our time steps\Nand then we see this. It's hooray! It
Dialogue: 0,0:43:21.40,0:43:27.24,Default,,0000,0000,0000,,worked actually. Because it's actually not\Nquite simple usually to set up a running
Dialogue: 0,0:43:27.24,0:43:32.00,Default,,0000,0000,0000,,program. A running problem, because you\Nhave to really think about what should be
Dialogue: 0,0:43:32.00,0:43:38.17,Default,,0000,0000,0000,,the physics. What's the scale of your\Nproblem? What's the timescale of your
Dialogue: 0,0:43:38.17,0:43:44.99,Default,,0000,0000,0000,,problem? And specify this in a good way.\NBut in principle, this is how it works.
Dialogue: 0,0:43:44.99,0:43:49.32,Default,,0000,0000,0000,,There are few test problems if you\Nactually want to play around with this to
Dialogue: 0,0:43:49.32,0:43:56.39,Default,,0000,0000,0000,,make it easy for the beginning. And this\Nis how we do simulations. So, as I already
Dialogue: 0,0:43:56.39,0:44:02.32,Default,,0000,0000,0000,,set, we can just start them on our\Nlaptops. So, here this is my laptop. I
Dialogue: 0,0:44:02.32,0:44:07.86,Default,,0000,0000,0000,,just type a dot slash FARGO3D and that\Nshould run, right? And then I just wait
Dialogue: 0,0:44:07.86,0:44:16.45,Default,,0000,0000,0000,,for ten years to finish the simulations of\N500 timesteps or outputs. Well, that's not the best
Dialogue: 0,0:44:16.45,0:44:27.66,Default,,0000,0000,0000,,idea. So, we need more power. And both of\Nus, for example, are using a cluster for
Dialogue: 0,0:44:27.66,0:44:36.88,Default,,0000,0000,0000,,Baden-Württemberg and that takes down our\Ncomputation time by a lot. Usually, like a
Dialogue: 0,0:44:36.88,0:44:45.05,Default,,0000,0000,0000,,factor of maybe 20, which is a lot. So, I\Nwould need on my computer maybe a year and
Dialogue: 0,0:44:45.05,0:44:53.04,Default,,0000,0000,0000,,then I just need maybe 5 hours, a few days\Nor a week on this cluster, which is
Dialogue: 0,0:44:53.04,0:44:56.38,Default,,0000,0000,0000,,usually the simulation time about a week\Nfor me.
Dialogue: 0,0:44:56.38,0:45:04.44,Default,,0000,0000,0000,,caro: So, what you see here is that we use\NGPUs, yes. But we do not or mostly not use
Dialogue: 0,0:45:04.44,0:45:09.63,Default,,0000,0000,0000,,them for gaming. We use them for actually\Nactual science. Yeah, would be nice to
Dialogue: 0,0:45:09.63,0:45:20.61,Default,,0000,0000,0000,,play on that, right? That just said!\Nmiosta: So, back to our Earth, actually.
Dialogue: 0,0:45:20.61,0:45:27.67,Default,,0000,0000,0000,,So, can we now? We wanted to grow our own\Nplanet. We can do that, yes of course. Can
Dialogue: 0,0:45:27.67,0:45:31.60,Default,,0000,0000,0000,,we grow Earth? Well, Earth is a very\Nspecial planet. We have a very nice
Dialogue: 0,0:45:31.60,0:45:37.72,Default,,0000,0000,0000,,temperature here, right? And we have not a\Ncrushing atmosphere like Jupiter, like a
Dialogue: 0,0:45:37.72,0:45:43.44,Default,,0000,0000,0000,,huge planet that we could not live under.\NWe have a magnetic field that shields us
Dialogue: 0,0:45:43.44,0:45:53.76,Default,,0000,0000,0000,,from the radiation from space and we have\Nwater. But just enough water so that we
Dialogue: 0,0:45:53.76,0:46:00.17,Default,,0000,0000,0000,,still have land on this planet where we\Ncan live on. So, even if we fine tune
Dialogue: 0,0:46:00.17,0:46:05.23,Default,,0000,0000,0000,,simulations, the probability that we\Nactually hit Earth and have all the
Dialogue: 0,0:46:05.23,0:46:12.80,Default,,0000,0000,0000,,parameters right is actually tiny. It's\Nnot that easy to simulate an Earth. And
Dialogue: 0,0:46:12.80,0:46:17.32,Default,,0000,0000,0000,,there are a lot of open questions, too.\NHow did we actually manage to get just
Dialogue: 0,0:46:17.32,0:46:24.24,Default,,0000,0000,0000,,this sip of water on our surface? How did\Nwe manage to collide enough mass or
Dialogue: 0,0:46:24.24,0:46:30.06,Default,,0000,0000,0000,,aggregate enough mass to form this\Nterrestrial planet without Jupiter is
Dialogue: 0,0:46:30.06,0:46:35.74,Default,,0000,0000,0000,,sweeping up all the mass in our system?\NHow could we be stable in this orbit when
Dialogue: 0,0:46:35.74,0:46:42.66,Default,,0000,0000,0000,,there are seven other planets swirling\Naround and interacting with us? All of
Dialogue: 0,0:46:42.66,0:46:48.66,Default,,0000,0000,0000,,this is open in our field of research\Nactually, and not completely understood.
Dialogue: 0,0:46:48.66,0:46:54.62,Default,,0000,0000,0000,,This is the reason why we still need to \Ndo astrophysics and even in all our
Dialogue: 0,0:46:54.62,0:47:01.01,Default,,0000,0000,0000,,simulations there is no planet B. And the\Nearth is quite unique and perfect for
Dialogue: 0,0:47:01.01,0:47:06.57,Default,,0000,0000,0000,,human life. So, please take care of the\NEarth and take care of yourself and of all
Dialogue: 0,0:47:06.57,0:47:12.27,Default,,0000,0000,0000,,the others people on the Congress. And\Nthank you for listening and thank you to
Dialogue: 0,0:47:12.27,0:47:20.38,Default,,0000,0000,0000,,everyone who helped us make this possible.\NAnd to the people who actually coded our
Dialogue: 0,0:47:20.38,0:47:24.21,Default,,0000,0000,0000,,programs with which we simulate. \NThank you!
Dialogue: 0,0:47:24.21,0:47:37.37,Default,,0000,0000,0000,,{\i1}applause{\i0}
Dialogue: 0,0:47:37.37,0:47:42.32,Default,,0000,0000,0000,,Herald: Thank you for the beautiful talk\Nand for the message at the end, the paper
Dialogue: 0,0:47:42.32,0:47:47.97,Default,,0000,0000,0000,,is open for discussion, so if you guys\Nhave any questions, please come to the
Dialogue: 0,0:47:47.97,0:47:57.16,Default,,0000,0000,0000,,microphones. I'm asking my Signal Angel?\NNo questions right now. But microphone two
Dialogue: 0,0:47:57.16,0:48:00.16,Default,,0000,0000,0000,,please!\NMic2: Oh, yeah. Thank you very much.
Dialogue: 0,0:48:00.16,0:48:05.69,Default,,0000,0000,0000,,Really beautiful talk. I can agree. I have\Ntwo questions. The first is short. You are
Dialogue: 0,0:48:05.69,0:48:10.98,Default,,0000,0000,0000,,using Navier-Stokes equation, but you have\Non the one hand, you have the dust disc
Dialogue: 0,0:48:10.98,0:48:14.94,Default,,0000,0000,0000,,and on the other hand, you have solid\Nplanets in it. And so are you using the
Dialogue: 0,0:48:14.94,0:48:18.62,Default,,0000,0000,0000,,same description for both \Nor is it a hybrid?
Dialogue: 0,0:48:18.62,0:48:23.55,Default,,0000,0000,0000,,miosta: It very much depends. This is one\Nof the things I showed you that for PLUTO,
Dialogue: 0,0:48:23.55,0:48:31.30,Default,,0000,0000,0000,,we write this C file that specifies some\Nthings and about every physicist has
Dialogue: 0,0:48:31.30,0:48:39.09,Default,,0000,0000,0000,,somewhat his or her own version of things.\NSo, some usually the planets, if they are
Dialogue: 0,0:48:39.09,0:48:47.03,Default,,0000,0000,0000,,large, they will be put in as a gravity\Nsource. And possibly one that can accrete
Dialogue: 0,0:48:47.03,0:48:54.09,Default,,0000,0000,0000,,and pebbles are usually then put in a\Ndifferent way. However, also pebbles are
Dialogue: 0,0:48:54.09,0:48:57.54,Default,,0000,0000,0000,,at the moment a bit complicated. There are\Nspecial groups specializing in
Dialogue: 0,0:48:57.54,0:49:04.08,Default,,0000,0000,0000,,understanding pebbles because as we said\Nin the beginning, when they collide,
Dialogue: 0,0:49:04.08,0:49:10.45,Default,,0000,0000,0000,,usually they should be destroyed. If you\Nhit two rocks very together, they don't
Dialogue: 0,0:49:10.45,0:49:14.87,Default,,0000,0000,0000,,stick. If you hit them hard together, they\Nsplatter around and we don't end up with an bigger object
Dialogue: 0,0:49:14.87,0:49:23.39,Default,,0000,0000,0000,,caro: Just to explain pebbles are small\Nrocks or like big sand stones or something
Dialogue: 0,0:49:23.39,0:49:28.71,Default,,0000,0000,0000,,like that. Yeah. So bigger rocks, \Nbut not very big, yet.
Dialogue: 0,0:49:28.71,0:49:33.19,Default,,0000,0000,0000,,miosta: Yes!\Ncaro: It depends on which code you use.
Dialogue: 0,0:49:33.19,0:49:38.37,Default,,0000,0000,0000,,Mic2: Thank you. Very short, maybe one.\NDo you also need to include relativistic
Dialogue: 0,0:49:38.37,0:49:46.52,Default,,0000,0000,0000,,effects. Or is that completely out?\Nmiosta: It's a good question. Mostly if
Dialogue: 0,0:49:46.52,0:49:54.68,Default,,0000,0000,0000,,you have a solar type system, you're in\Nthe arrange where this is not necessary.
Dialogue: 0,0:49:54.68,0:50:00.01,Default,,0000,0000,0000,,For example, with the binaries, if they\Ngot very close together, then at the inner
Dialogue: 0,0:50:00.01,0:50:05.20,Default,,0000,0000,0000,,part of the disc, that is something we\Ncould consider. And actually, I know for
Dialogue: 0,0:50:05.20,0:50:10.50,Default,,0000,0000,0000,,PLUTO, it has modules to include\Nrelativistic physics, too, yes!
Dialogue: 0,0:50:10.50,0:50:14.00,Default,,0000,0000,0000,,Mic2: Thank you!\NHerald: OK, we have quite some questions,
Dialogue: 0,0:50:14.00,0:50:19.70,Default,,0000,0000,0000,,so keep them short. Number one, please!\NMic1: Thank you. Yeah. Thank you very
Dialogue: 0,0:50:19.70,0:50:24.49,Default,,0000,0000,0000,,much for your interesting talk. And I\Nthink you had it on your very first slides
Dialogue: 0,0:50:24.49,0:50:31.78,Default,,0000,0000,0000,,that about 70 percent of the universe\Nconsists of dark matter and energy. Is that
Dialogue: 0,0:50:31.78,0:50:37.00,Default,,0000,0000,0000,,somehow considered in your \Nsimulations or how do you handle this?
Dialogue: 0,0:50:37.00,0:50:43.02,Default,,0000,0000,0000,,caro: Well in the simulations we make, we\Nare doing planets and discs around stars.
Dialogue: 0,0:50:43.02,0:50:47.44,Default,,0000,0000,0000,,It's not considered there. In the\Nsimulation we showed you about the
Dialogue: 0,0:50:47.44,0:50:52.67,Default,,0000,0000,0000,,universe at the beginning, the blueish\Nthings were all dark matter. So, that was
Dialogue: 0,0:50:52.67,0:50:56.26,Default,,0000,0000,0000,,included in there.\NMic1: OK, thank you.
Dialogue: 0,0:50:56.26,0:51:00.51,Default,,0000,0000,0000,,Herald: OK. Microphone 3.\NMic3: Hi, thanks. Sorry, I think you
Dialogue: 0,0:51:00.51,0:51:05.74,Default,,0000,0000,0000,,talked about three different programs. I\Nthink PLUTO, FARGO3D and a third one. So,
Dialogue: 0,0:51:05.74,0:51:09.62,Default,,0000,0000,0000,,for a complete beginner: which program\Nwould you suggest is like you more use
Dialogue: 0,0:51:09.62,0:51:12.57,Default,,0000,0000,0000,,like if you want to learn more? \NWhich one is user friendly or good?
Dialogue: 0,0:51:12.57,0:51:18.59,Default,,0000,0000,0000,,miosta: I would suggest FARGO3D first. It's\Nkind of user friendly, has a somewhat good
Dialogue: 0,0:51:18.59,0:51:26.24,Default,,0000,0000,0000,,support and they are always also very\Nthankful for actual comments and additions
Dialogue: 0,0:51:26.24,0:51:32.03,Default,,0000,0000,0000,,if people actually are engaged in trying\Nto improve on that. Because we are
Dialogue: 0,0:51:32.03,0:51:37.33,Default,,0000,0000,0000,,physicists. We're not perfect programmers\Nand we're also happy to learn more. So
Dialogue: 0,0:51:37.33,0:51:42.72,Default,,0000,0000,0000,,yeah, FARGO3D I would suggest, it has some\Neasy ways of testing some systems and
Dialogue: 0,0:51:42.72,0:51:45.44,Default,,0000,0000,0000,,getting something done.\Ncaro: And it also has a very good
Dialogue: 0,0:51:45.44,0:51:53.57,Default,,0000,0000,0000,,documentation and also a manual "How to\Nmake the first steps on the Internet". So,
Dialogue: 0,0:51:53.57,0:51:56.98,Default,,0000,0000,0000,,you can look that up.\NMic3: Awesome. Thank you.
Dialogue: 0,0:51:56.98,0:52:00.15,Default,,0000,0000,0000,,Herald: Let's get one question from\Noutside, from my Signal Angel.
Dialogue: 0,0:52:00.15,0:52:05.60,Default,,0000,0000,0000,,Signal Angel: Thank you for your talk.\NThere's one question from IRC: How do you
Dialogue: 0,0:52:05.60,0:52:09.51,Default,,0000,0000,0000,,know your model is good when you can only\Nobserve snapshots?
Dialogue: 0,0:52:09.51,0:52:17.77,Default,,0000,0000,0000,,caro: Oh, that's a good question. As we\Nsaid, we're in theoretical astrophysics.
Dialogue: 0,0:52:17.77,0:52:25.17,Default,,0000,0000,0000,,So, there are theoretical models and these\Nmodels cannot include everything. So,
Dialogue: 0,0:52:25.17,0:52:32.61,Default,,0000,0000,0000,,every single process, it's not possible\Nbecause then we would calculate for years.
Dialogue: 0,0:52:32.61,0:52:37.48,Default,,0000,0000,0000,,Yeah, to know if a model is \Ngood you have to…
Dialogue: 0,0:52:37.48,0:52:46.43,Default,,0000,0000,0000,,miosta: Usually, you have a hypothesis or\Nan observation that you somehow want to
Dialogue: 0,0:52:46.43,0:52:54.06,Default,,0000,0000,0000,,understand. With most of the necessary\Nphysics at this stage to reproduce this
Dialogue: 0,0:52:54.06,0:53:01.66,Default,,0000,0000,0000,,image. So, also from the observation we\Nhave to take into the account what our
Dialogue: 0,0:53:01.66,0:53:07.65,Default,,0000,0000,0000,,parameters kind of should be, how dense\Nthis end of the simulation should be and
Dialogue: 0,0:53:07.65,0:53:13.15,Default,,0000,0000,0000,,things like this. So, by comparing two\Nobservations, that's the best measure we
Dialogue: 0,0:53:13.15,0:53:21.79,Default,,0000,0000,0000,,can get. If we kind of agree. Of course,\Nif we do something completely wrong, then
Dialogue: 0,0:53:21.79,0:53:26.60,Default,,0000,0000,0000,,it will just blow up or we will get a\Nhorribly high density. So, this is how we
Dialogue: 0,0:53:26.60,0:53:34.27,Default,,0000,0000,0000,,know. Physics will just go crazy if we do\Ntoo large mistakes. Otherwise, we would
Dialogue: 0,0:53:34.27,0:53:39.33,Default,,0000,0000,0000,,try to compare two observations that it\Nactually is sensible what we did.
Dialogue: 0,0:53:39.33,0:53:44.44,Default,,0000,0000,0000,,caro: Yeah, that's one of the most\Ncomplicated tasks to include just enough
Dialogue: 0,0:53:44.44,0:53:52.40,Default,,0000,0000,0000,,physics that the system is represented in\Na good enough way. But not too much that
Dialogue: 0,0:53:52.40,0:53:57.40,Default,,0000,0000,0000,,our simulation would blow up in time.\NHerald: Number two, please.
Dialogue: 0,0:53:57.40,0:54:03.21,Default,,0000,0000,0000,,Mic2: I've got a question about the\Nadaptive grids. How does the computer
Dialogue: 0,0:54:03.21,0:54:10.80,Default,,0000,0000,0000,,decide how to adapt the grid? Because the\Ndata where's the high density comes after
Dialogue: 0,0:54:10.80,0:54:17.66,Default,,0000,0000,0000,,making the grid...\Nmiosta: Yes, this is actually quite an
Dialogue: 0,0:54:17.66,0:54:25.47,Default,,0000,0000,0000,,interesting and also not quite easy to\Nanswer question. Let me try to give a
Dialogue: 0,0:54:25.47,0:54:34.30,Default,,0000,0000,0000,,breakdown nutshell answer here. \NThe thing is, you measure and evaluate the
Dialogue: 0,0:54:34.30,0:54:39.38,Default,,0000,0000,0000,,velocities. Or in the flux, you also\Nevaluate the velocity. And if the velocity
Dialogue: 0,0:54:39.38,0:54:44.84,Default,,0000,0000,0000,,goes high, you know there's a lot\Nhappening. So, we need a smaller grid then
Dialogue: 0,0:54:44.84,0:54:50.42,Default,,0000,0000,0000,,there. So, we try to create more grid\Ncells where we have a higher velocity. In
Dialogue: 0,0:54:50.42,0:54:55.05,Default,,0000,0000,0000,,a nutshell, this is of course in an\Nalgorithm a bit harder to actually
Dialogue: 0,0:54:55.05,0:55:00.00,Default,,0000,0000,0000,,achieve. But this is the idea. We measured\Nthe velocities at each point. And then if
Dialogue: 0,0:55:00.00,0:55:03.51,Default,,0000,0000,0000,,we measure a high velocity, \Nwe change to a smaller grid.
Dialogue: 0,0:55:03.51,0:55:08.64,Default,,0000,0000,0000,,Mic2: So, you can predict where the mass\Nwill go and whether densities are getting high.
Dialogue: 0,0:55:08.64,0:55:12.60,Default,,0000,0000,0000,,miosta: Exactly. Step by step so to say.
Dialogue: 0,0:55:12.60,0:55:15.89,Default,,0000,0000,0000,,Mic2: Thanks\NHerald: We stay with Microphone 2.
Dialogue: 0,0:55:15.89,0:55:20.64,Default,,0000,0000,0000,,Mic2: Okay. I've got a bit of a classical\Nquestion. So, I guess a lot relies on your
Dialogue: 0,0:55:20.64,0:55:25.20,Default,,0000,0000,0000,,initial conditions and I have two\Nquestions related to that. So first, I
Dialogue: 0,0:55:25.20,0:55:30.67,Default,,0000,0000,0000,,guess they are inspired by observations.\NWhat are the uncertainties that you have?
Dialogue: 0,0:55:30.67,0:55:33.85,Default,,0000,0000,0000,,And B, then what is the impact if you\Nchange your initial conditions like the
Dialogue: 0,0:55:33.85,0:55:41.17,Default,,0000,0000,0000,,density in the disc?\Nmiosta: Yeah, right now my main research
Dialogue: 0,0:55:41.17,0:55:46.11,Default,,0000,0000,0000,,is actually figuring out a sensible\Ninitial conditions or parameters for a
Dialogue: 0,0:55:46.11,0:55:53.22,Default,,0000,0000,0000,,disc. If you just let it have an initial\Nset of conditions and a sensible set of
Dialogue: 0,0:55:53.22,0:56:00.42,Default,,0000,0000,0000,,parameters and let it run very long, you\Nexpect a system hopefully to convert to
Dialogue: 0,0:56:00.42,0:56:05.13,Default,,0000,0000,0000,,the state that it should be in. But your\Nparameters are of course very important.
Dialogue: 0,0:56:05.13,0:56:12.24,Default,,0000,0000,0000,,And here we go back to what we can\Nactually understand from observations. And
Dialogue: 0,0:56:12.24,0:56:17.88,Default,,0000,0000,0000,,what we need for example is the density,\Nfor example. And that is something we try
Dialogue: 0,0:56:17.88,0:56:24.90,Default,,0000,0000,0000,,to estimate from the light we see in these\Ndiscs that you saw in this nice grid with
Dialogue: 0,0:56:24.90,0:56:31.11,Default,,0000,0000,0000,,all these discs we estimate OK, what's the\Naverage light there? What should then be
Dialogue: 0,0:56:31.11,0:56:37.79,Default,,0000,0000,0000,,the average densities of dust \Nand gas in comparable disks.
Dialogue: 0,0:56:37.79,0:56:42.89,Default,,0000,0000,0000,,Mic2: Okay, thanks.\NHerald: Okay, one more at number two.
Dialogue: 0,0:56:42.89,0:56:50.15,Default,,0000,0000,0000,,Mic2: Yes. Thank you for the talk. When\Nyou increase the detail on the grid and
Dialogue: 0,0:56:50.15,0:56:59.55,Default,,0000,0000,0000,,you learn more. When you want to compute\Nthe gravitational force in one cell, you
Dialogue: 0,0:56:59.55,0:57:05.09,Default,,0000,0000,0000,,have to somehow hold masses from the all\Nthe other cells. So, the complexity of the
Dialogue: 0,0:57:05.09,0:57:07.09,Default,,0000,0000,0000,,calculus grows.\Nmiosta: Yes
Dialogue: 0,0:57:07.09,0:57:13.82,Default,,0000,0000,0000,,Mic2: Quadratically, at the square of the... \Nhow do you solve that? With more CPUs?
Dialogue: 0,0:57:13.82,0:57:20.93,Default,,0000,0000,0000,,caro: Well, that would be one way to do\Nthat. But there are ways to simplify if
Dialogue: 0,0:57:20.93,0:57:26.29,Default,,0000,0000,0000,,you have a lot of particles in one\Ndirection and they are far away from the
Dialogue: 0,0:57:26.29,0:57:34.40,Default,,0000,0000,0000,,object you're looking at. So, yeah. So, if\Nyou have several balls here and one ball
Dialogue: 0,0:57:34.40,0:57:41.71,Default,,0000,0000,0000,,here, then you can include all these balls\Nor you can think of them as one ball. So,
Dialogue: 0,0:57:41.71,0:57:48.77,Default,,0000,0000,0000,,it depends on how you look at it. So, how\Nyou define how many particles you can take
Dialogue: 0,0:57:48.77,0:58:02.23,Default,,0000,0000,0000,,together is when you look at the angle of\Nthis... many particles we'll have from the
Dialogue: 0,0:58:02.23,0:58:08.04,Default,,0000,0000,0000,,seen from the object you're looking at.\NAnd you can define a critical angle. And
Dialogue: 0,0:58:08.04,0:58:14.23,Default,,0000,0000,0000,,if an object gets smaller or if lot of\Nobjects get smaller than this angle, you
Dialogue: 0,0:58:14.23,0:58:20.20,Default,,0000,0000,0000,,can just say, OK, that's one object. So,\Nthat's a way to simplify this method. And
Dialogue: 0,0:58:20.20,0:58:23.39,Default,,0000,0000,0000,,there are some, yeah, \NI think that's the main idea.
Dialogue: 0,0:58:23.39,0:58:30.92,Default,,0000,0000,0000,,Herald: Okay, we have another one.\NMic2: Do you have a strategy to check if
Dialogue: 0,0:58:30.92,0:58:35.89,Default,,0000,0000,0000,,the simulation will give a valuable\Nsolution or does it happen a lot that you
Dialogue: 0,0:58:35.89,0:58:42.06,Default,,0000,0000,0000,,wait one week for the calculation and find\Nout OK it's total trash or it crashed in
Dialogue: 0,0:58:42.06,0:58:45.40,Default,,0000,0000,0000,,the time.\Ncaro: So, that also depends on the program
Dialogue: 0,0:58:45.40,0:58:53.24,Default,,0000,0000,0000,,you're using. So, in FARGO3D, it gives \Nthese outputs after a certain amount of
Dialogue: 0,0:58:53.24,0:58:58.98,Default,,0000,0000,0000,,calculation steps and you can already look\Nat those outputs before the simulation is
Dialogue: 0,0:58:58.98,0:59:05.21,Default,,0000,0000,0000,,finished. So, that would be a way to\Ncontrol if it's really working. Yeah, but
Dialogue: 0,0:59:05.21,0:59:11.53,Default,,0000,0000,0000,,I think...\Nmiosta: It's the same for PLUTO. So, there
Dialogue: 0,0:59:11.53,0:59:18.04,Default,,0000,0000,0000,,is a difference between timesteps and\Nactually output steps. So and you could
Dialogue: 0,0:59:18.04,0:59:23.49,Default,,0000,0000,0000,,define your output steps not and as the\Nwhole simulation, but you can look at each
Dialogue: 0,0:59:23.49,0:59:31.15,Default,,0000,0000,0000,,output step as soon as it's produced. So,\NI usually get like 500 outputs, but I
Dialogue: 0,0:59:31.15,0:59:36.63,Default,,0000,0000,0000,,already can look at the first and second after \Nmaybe half an hour or something like that.
Dialogue: 0,0:59:36.63,0:59:39.95,Default,,0000,0000,0000,,caro: Yeah, but it also happens that you
Dialogue: 0,0:59:39.95,0:59:44.06,Default,,0000,0000,0000,,start a simulation and wait, and wait, and\Nwait and then see you put something wrong
Dialogue: 0,0:59:44.06,0:59:48.76,Default,,0000,0000,0000,,in there and well then you have to do it\Nagain. So, this happens as well.
Dialogue: 0,0:59:48.76,0:59:53.07,Default,,0000,0000,0000,,Mic2: Thanks.\NHerald: Okay. One final question.
Dialogue: 0,0:59:53.07,1:00:02.24,Default,,0000,0000,0000,,Mic2: Yeah, OK. Is there a program in\Nwhich you can calculate it backwards? So
Dialogue: 0,1:00:02.24,1:00:07.18,Default,,0000,0000,0000,,that you don't have the starting\Nconditions but the ending conditions
Dialogue: 0,1:00:07.18,1:00:15.22,Default,,0000,0000,0000,,and you can calculate how it had started?\Nmiosta: Not for hydrodynamic. If you go to
Dialogue: 0,1:00:15.22,1:00:22.24,Default,,0000,0000,0000,,n-body, there is a way to go backwards in\Ntime. But for hydrodynamics, the thing is
Dialogue: 0,1:00:22.24,1:00:31.58,Default,,0000,0000,0000,,that you have turbulent and almost like\Nchaotic conditions. So, you cannot really
Dialogue: 0,1:00:31.58,1:00:38.81,Default,,0000,0000,0000,,turn them back in time. With n-body you \Ncan it because actually it's kind of... Well,
Dialogue: 0,1:00:38.81,1:00:44.89,Default,,0000,0000,0000,,it's not analytically solved, but it's\Nmuch closer than like turbulences,
Dialogue: 0,1:00:44.89,1:00:50.47,Default,,0000,0000,0000,,streams, spirals and all the \Nthings you saw in the simulations.
Dialogue: 0,1:00:50.47,1:00:57.56,Default,,0000,0000,0000,,Herald: OK, I guess that brings us to the\Nend of the talk and of the session. Thank
Dialogue: 0,1:00:57.56,1:01:03.27,Default,,0000,0000,0000,,you for the discussion and of course,\Nthank you guys for the presentation.
Dialogue: 0,1:01:03.27,1:01:16.73,Default,,0000,0000,0000,,{\i1}36c3 postroll music{\i0}
Dialogue: 0,1:01:16.73,1:01:30.00,Default,,0000,0000,0000,,Subtitles created by c3subtitles.de\Nin the year 2021. Join, and help us!