Ep. 468: Simulations for Science and Fun

Astronomy, Science | 0 comments

Astronomers depend on simulations to study the Universe. From relatively straightforward orbital simulations to vast simulations that try to recreate the large scale structure of the Universe from the Big Bang. Today we’re going to talk about some of those simulations, as well as tools you can use simulate the Universe.
We usually record Astronomy Cast every Friday at 3:00 pm EST / 12:00 pm PST / 20:00 PM UTC. You can watch us live on here on AstronomyCast.com, or the AstronomyCast YouTube page.
If you would like to support Astronomy Cast, please visit our page at Patreon here – https://www.patreon.com/astronomycast. We greatly appreciate your support!
If you would like to join the Weekly Space Hangout Crew, visit their site here and sign up. They’re a great team who can help you join our online discussions!

Download the show [MP3] | Jump to Shownotes | Jump to Transcript

This episode is sponsored by: LovePop

Show Notes

Illustris simulation
Computer simulations of universe
Projects from MIT – Large-scale Galaxy Formation, Dwarf Galaxy Formation, Large-scale Magnetic Fields, Dark Matter Detection and more
Universe Sandbox
Kerbal Space Program for orbital physics and propulsion
Weather simulation games
Extreme Weather Simulator
The Yarkovsky effect
WorldWide Telescope
Space Engine
Transit Finder


Transcription services provided by: GMR Transcription

Fraser: Astronomy Cast, Episode 468: Simulations for Fun and Science. Welcome to Astronomy Cast, your weekly fact-based journey through the cosmos, where we help you understand not only what we know, but how we know what we know. My name is Fraser Cain. I’m the Publisher of Universe Today, and with me is Dr. Pamela Gay, the Director of Technology and Citizen Science at the Astronomical Society of the Pacific, and the Director of CosmoQuest. Hey, Pamela, how are you doing?
Pamela: I’m doing well. How are you doing?
Fraser: Doing great. So, before we get any deeper into this episode, I just want to give a quick shout-out to what is going to become sort of an unfolding story, and that is to our friends at Oceanside Photo and Telescope, who just sent the two of us a brand new 70 millimeter refracting –
Pamela: Apo refractor.
Fraser: Refracting telescopes, with cool CCD sensors to do astrophotography, and we are going to sort of learn – upgrade our game – in visual and astrophotography, and bring you guys along with it. And, a huge, huge thank you to our friends at Oceanside Photo and Telescope. Podcast listeners, you can’t see it, but Pamela just made a little heart sign. So, big thanks to OPT, who I think they’re like the biggest telescope retailer in the world, I think. So, they’re a great company. They’ve been great to work with for years – decades, we’ve been talking to them.
Pamela: And, they’re an utterly honest company, which is why I’ve recommended them so many times. They have more than once prevented me from buying something more expensive than I needed to, and been like, “No, no. You actually want this other thing that costs less.” And, that earns my coming back over and over forever. And, so I have.
Fraser: So, they haven’t sponsored the show. They just put a powerful telescope in our hands, and, you know, we wanted to give them a big thank you. Of course, it’s going to turn into more work because now we’re going to be going, and having to, like, learn how to do this, and make videos about it.
Pamela: You’re learning. I’m going back to my roots.
Fraser: Yeah, that’s true. That’s true. I’m learning. Alright. Let’s get on with this week’s show. So, astronomers depend on simulations to study the universe, from relatively straight-forward orbital simulations to vast simulations that try to recreate the large-scale structure of the universe from the big bang onward.
Today we’re going to talk about some of those simulations, as well as tools you can use to simulate the universe for fun and science. Alright, Pamela, so let’s talk about some universe simulations. Should we talk about the history first about this idea where you, as a professional astronomer, can make observations, but then you have to predict? You have to try to figure out if the theories match the observations that you’ve made. So, how do simulations – from the perspective of an astronomer – how do those two kind of come together?
Pamela: Well, it goes all the way back to the beginning, and initially simulations were simply lots and lot of pieces of paper strung together.
Fraser: Organized by computers. In other words, human beings who –
Pamela: Right. So, it goes back to the idea that folks like Shander Sakar went through and figured out what are all of the equations that allow us to balance out the internal dynamics of a star, and once we have all of those equations, we have the ability to go through layer by layer by layer, and simulate a star. And, this is actually, probably, one of the first big simulations that almost every astronomy major is required to write – where you’re taking the equations. You’re balancing the star. You’re figuring out at what level do you have radiation transfer? At what layer does convection kick in? And, it kind of is miserable to do that by hand.
So, we’ve been looking to find ways to simulate this in computers since before computers had anything more advanced than punch cards. And, beyond just simulating stars, it’s the, “How do we go through and figure out orbital percubations to figure out what happens to comets over time; to figure out, given the fullness of time, how do the orbits of the planets change? Yes, you can go through, and you can repeat the calculations over and over and over on paper. Kepler did this. But, if you like yourself, then you figure out how to program a computer.
Fraser: Right, and let a computer do the heavy lifting. The one made of silicone; not necessarily a human being.
Pamela: Very true.
Fraser: I think one of the things that’s really important about this is this idea that if you have a theory, and that theory makes predictions, but you have lots of historical observations that you’ve made through the past – like, climate simulations and things like that. Or, you have climate data, and then you develop a theory and a simulation. You’ve programmed a computer that does it, and then it runs the math, and tries to mimic the past data, and then make predictions about the future, right?
Pamela: And, it’s not just going from past data to match current, to match future. We do that with climate, but with galaxy simulations half the fun is saying, “Hey, there’s this crazy, distorted set of galaxies in the process of colliding, so look at the Mys, look at the antenna. Let’s see if we can, within the software, collide galaxies and figure out what the initial conditions where that led to these amazing, splayed out spatterings of star and gas, and how do we get the dark matter content correct and interacting correctly so that the collision starts as early as we see it starting on the sky?
It’s from going through and trying to reproduce these in progress collisions that we’re able to build up a complete understanding of how collisions occur.
Fraser: And, the most powerful computers in the world are used for this. Like, simulating super nova explosions – they use some of the most powerful computers to just try to understand what’s going on in there.
Pamela: And, globular clusters is actually on of the places where all of that started. Back in graduate school for me – so, back in the early 2000s, late 1990s – a group of scientists trying to understand all of the orbital dynamics inside of globular clusters had to develop the first ever petaflop super computer in order to get all of the different interactions going correctly. And, this has grown into being the Dragon Globular Cluster simulations being done at the Max Planck institute. So, these folks are still going along, still figuring out how to evolve a million gravitationally interacting stars, and try to understand how is it that you have new binaries forming? How is it that you have these triple star interactions that fling a star outward?
And, one of the coolest results is, again, over the fullness of time, if you watch a globular cluster longer than humanity has existed, they actually pulsate like a heartbeat if the simulations have it correct.
Fraser: And, of course, we can’t wait longer than human beings have existed, so having a simulation on a petaflop computer is the best that we can do.
Pamela: Yes.
Fraser: And, some of the ones that I find most fascinating – I’ve done a bit of writing about this – is just the simulations of the entire universe, where they start from shortly after the cosmic microwave background radiation is released to the first stars; to the first proto-galaxies; to the first galaxies; to the cosmic web at the largest scale that we have today. And, the level that the simulation can now predict – what the universe looks like – is just amazing.
Pamela: And, what’s so cool about these is watching how they’ve changed over time; trying to understand how you go from and essentially smooth distribution of matter in the dark ages of the universe to having the completely lumpy, bumpy Swiss cheese of modern day cosmology requires getting all sorts of different things interacting – from figuring out how does different dark matter interact or not interact? How does the temperature – the velocity of the particles right after the cosmic microwave background is released – affect things? You have to factor in all of these different effects, and it’s difficult.
So, we talk about how many particles there are. We talk about – is it gas? Is it all of these different ways of approximating things? And, we’ve gone from basically 1000 x 1000 particle cube to million by a million, to ever, ever larger simulations, spanning more fine-grained periods of time. So, it’s no longer this seed represents what will become a galaxy to this seed represents a star cluster, to this represents a star, and we can see the universe in our simulations turn on, light up, collapse down, and evolve into what we see.
Fraser: Yeah, the people of the podcast didn’t see it, but I was putting up videos from this illustrious simulation showing what those simulations look like, and it’s so impressive just to see how well, and how detailed, and how accurate these simulations are.
Pamela: This episode of Astronomy Cast is brought to you by Love Pop. Unlock special pricing for five or more cards, and get free shipping on any order by going to LovePop.com/astronomy. So, I’m kind of struggling with exactly what to say because I’m giving a bunch of these cards as part of my Christmas gifts this year, and if I say too much, people who are listening, who are also personal friends will totally know what I’m getting them. So, I have to be vague and say that these are the cards that are more than meets the eye.
You pull them out of the envelope, and it’s like, “Wow. This is a really thick card.” And, then you open them up, and pop, three dimensional paper sculpture pops out to great you. They come in a whole bunch of different options, different color combinations, different seasonal choices. They have Christmas cards. They have random cards. They have all of the awesome you could want in paper form. So, go to LovePop.com/astronomy.
Take a look at stuff – some of which I’m buying for Christmas gifts, and pick out your favorite cards. And, like I said, unlock special pricing for five or more cards, and get free shipping on any order by going to LovePop.com/astronomy. Thanks.
Fraser: Another simulation that I think is pretty great is people have simulated, sort of, what the future is going to look like. Let’s take the future of The Milkyway and its interactions with the Andromeda Galaxy because they’re going to be colliding in a few billion years.
Pamela: And, here again, it starts to be just how big are the dark matter halos between our two galaxies because it’s those dark matter halos that will hit first, and start triggering the interactions, and then once the interactions are triggered, how fast does dust and gas plunge into the center of our respective galaxies and turn on those black holes. And, we won’t be here. Our planet will be a crispy critter, but it’s awesome to think about, essentially, what will this fireworks display look like?
Fraser: Absolutely. Now, let’s talk about some stuff that people can use on their own if they want to kind of get into this and be able to start playing around. What are some tools they can use to actually run some simulations?
Pamela: Well, I think the one that people play with the most that in some ways is perhaps the most satisfying is Universe Sandbox. It’s available for Steam. It’s in the humble bundle, which is a way of buying things and also giving back; paying it forward through your purchase. And, the Universe Sandbox allows you to play with gravity; to drop in new masses. It runs on all of the major platforms – Windows, Mac, Linux. And, it’s just a nice, easy going, kind of pretty simulation that lets you play with the universe, and who doesn’t want to play with the universe.
Fraser: They’re adding stuff to it all of the time, and sort of a recent Alpha that I was playing around with – they had added tidal heating efforts and Roche limits, and things like that. So, you could have the moon be going around the earth, and you could move it closer, and closer, and closer until the tidal force had crossed the Roche limit and it tore the moon apart into this ring, and then the chunks of this moon ring come raining down on earth. For some of our videos, we’ve done things like taken the sun and turned it into smaller stars, and had the stars orbit around each other. And, they’re adding new stuff all of the time to the game.
They’ve added – whenever something new is discovered – I haven’t checked, but I’m sure they’ve added the interstellar asteroid Omuahmuah into it already. It’s the kind of thing that they do. So, as soon as new objects – new planetary objects are figured out; new stars are figured out; new galaxies are discovered – they’ll drop them into the simulation. Spacecraft, and then you can do stuff.
There’s a great simulation you can do where you put two baseballs a couple of meters apart from each other, but with nothing else in the universe, and you can just watch the gravity of them bring them together over the course of a couple of months, and they bonk into each other because their gravity is pulling them together.
Pamela: And, if you’re more interested in just exploring our own reality without tearing things apart, which we all know is fabulous – if you’d rather stick to the actual universe, there’s Celestia, Solarium, and Worldwide Telescope – all of which have different data layers you can turn on and off, all of which allow you to essentially zoom around into different perspectives within the universe, and get in, and really see what’s out there.
But, what I found really cool is a lot of the physics that goes into creating things like Space Engine, Universe Sandbox – all of these different, “Let’s go in and mess with the solar system,” simulations – the same physics we need to do for that is the same physics that goes into video games like Portal. And, it turns out that they’re hiring people who have degrees in physics and astrophysics to go work at Pixar and figure out how to get the hair correct, and the animations. To go work at EA and figure out how to get it so that you actually fall through the portals just right, and then tweak reality a little bit so you don’t accidentally become a harmonic oscillator between two layers because your velocity balanced out too perfectly.
Fraser: We talked about Universe Sandbox, and they have someone on staff who is a physicist and astrophysicist by training, and her job is to provide the calculations, and to make sure that the way that they’re simulating this stuff matches reality to the best our pathetic computers can handle it.
Pamela: And, for a long time on of the leads over at Pixar was Chris Ford, who worked on their simulations for their animations, and he’s another PhD physicist who took all of this knowledge and turned it into figuring out how to get clothing to move right; how to get hair to move right. And, all of these problems are just different versions of the same physics that happens in outer space. It’s just written at a different scale.
Fraser: Yeah. We did a whole episode about some of the space video games that we like to play. We always talk about the Kerbal Space Program as this great simulation. It has some flaws. It doesn’t handle La Grange points correctly. The amount – It’s a lot easier to get into space from Kerbal that it is the in the real world, although people have made mods that make it more – matching reality. I say this every time we talk about this game, I have learned more about the way orbital mechanics works – the way space flight works – from playing that game for just a few hours than all of the years of reporting on space.
Because it’s one thing to, “Oh, yeah. It had a centar upper stage, and released the satellite into a geosynchronous transfer orbit,” etc., etc., etc. That’s one thing. But, to actually make your spacecraft get into a geosynchronous transfer orbit, and realize how you needed to build that second stage to be able to do it is just absolutely vital. So, it’s amazing to see this coming back around that the simulations make for fun games, and learning to play these fun games are making people who are more prepared as scientists to be able to then go into the sciences, and help push the science further. It’s a really beautiful kind of connection.
Pamela: And, what’s cool is it starts to allow you to really answer questions like, “What would happen if a black hole slowly started to invade our solar system?” And, science fiction has often taken on these kinds of questions, and have over-simplified them; have made it seem like everything would go to hell in a hand-basket. But, the reality is you can get a tiny black hole fairly far in before we start feeling its effects because the effects add up over time, and it takes time to see, “Oh, my planet is no longer where it should be.”
Fraser: So, I’d like to talk a bit about where, maybe, simulations can go a little – not awry – but, it can cause controversy. Climate change is one of these examples where a tremendous amount of resources are being put in to try to develop simulations for what the climate is going to be doing over time, and you have all of these past observations, so now people are creating simulations to try to predict where the climate is going to be going next, and yet the models that they’re developing don’t necessarily match. What cause this uncertainty?
Pamela: Well, first of all, you have to make approximations if you ever want your software to finish running. And, that is a fundamental problem. It used to be that when we were simulating stars, we essentially did a one day simulation where we did a single cut from the core of a star out to its surface along a straight line, and ignored everything going on outside of that line through the star. From there, we moved on to doing two dimensional simulations where you start to see more of how the convection affects things. You start to get more effect from pocket of material, but the balloon you have in your two-day slice is going to be very different from the sphere of material you have moving in a 3D reality.
So, as our simulations become more and more able to handle the real world as our processors get faster, and our algorithms get better – more because our processors get faster – we’re able to get a more realistic view on things, but it’s still not perfect. It’s still not accurate because we don’t know exactly how to model convection. This is like one of the most ridiculous things that the way oil burbles and move in your frying pan; the way your lava lamp works – the detailed math of that, we’re still trying to figure out how to do that precisely. And, stars are a lot more complicated than lava lamps.
Fraser: Yeah.
Pamela: So, we have to make all of these approximations. And, with our planet it gets even harder because cities, for instance, affect weather because they have a different heat capacity – a different way of storing heat in their asphalt, and their cement. How cities grow affects the weather as it passes over those cities. So, you go back in time, and you figure out what model matched what happened in the 50s? Okay, that’s fine, accept now everything has changed. So, we’re dealing with this urban sprawl, cities changing, forest coverage, plant coverage, land capacity, amount of water in the land. All of these variables are hard.
Fraser: And, I think that we – We talked about this in a previous episode about how actually short-range weather forecasting has gotten surprisingly accurate. Like, if the weather says that next week, on Thursday it’s going to be raining, it used to be that that was ridiculous. And, now you kind of can rely on it. Don’t plan your vacation next week because it could be raining. So, that is the power of these super computers that have been simulating weather in the short-term and the medium-term, but in the long-term, you just have so many variables that it can just take things into other directions.
So, you just sort of put together all of the different simulations all at the same time, and try to sort of find out what is the average that it’s all telling you.
Pamela: And, this is part of why we can’t, for instance, say what the orbit of an asteroid is going to be over more than a few thousand or a few tens of thousands of years. The simple variations in soil color combined with rotation and interaction with sunlight is going to change that asteroids orbit in ways we’re still figuring out, and can’t fully simulate. So, all of these things introduce uncertainty. They introduce error. And, in our climate models, we’re getting better, but we’re not getting perfect. And, it turns out that the world is falling to bits in a hand-basket faster than we had anticipated, in part because, well, we’re affecting our environment faster than we anticipated.
And, then there are also all sorts of things we’re just discovering like heat sources under the arctic ice shield that we didn’t know about, and those are – Sometimes you don’t know you’re unknowns, and that makes it hard.
Fraser: Right. Classic Donald Rumsfeld advice. Sage words, he says, ironically. But, this prediction about the asteroid movements. I just want to go back and talk about that again because it’s one of those situations where the universe feels – for a certain extent – once you understand the math, the whole thing works like clockwork. If you get Newton’s gravitational equations down, you should be able to predict the movements of all of these bodies for hundreds, thousands, millions of years. But, the reality is that even our most powerful computers can only simulate the movements of the planets within these solar systems to a certain point, and then it all just kind of goes, “I don’t know.”
Pamela: And, there are so many silly little things that can be of huge impact. So, for instance, the wrong coronal mass ejection hitting a near passing asteroid at just the right albito ratio might create a minor deflection of the object, that over a thousand years, puts it in a completely different place.
Fraser: Or, YORP.
Pamela: YORP?
Fraser: You know, the – I don’t really know all of the parts of it – but, the fact that asteroids, as they rotate, and they emit radiation and it causes a tiny thrust. I’m sure someone listening to this – Sondy Springer is probably going, “YORP is,” and, I’ll look it up while you talk about YORP, but the implications that it has on those simulations, right?
Pamela: So, yes. And, this is one of those things where we don’t necessarily know where on an asteroids surface it’s going to suddenly decide, “Hey, I have volatiles. I shall turn them into a jet.” And, like I said, it’s that fine-grained interaction between different soil colors can make all the difference in the world. This is why painting asteroids allows us to move them. – Not that we’ve done that yet, but we have in simulations.
Fraser: Exactly. We’ve done it in simulations. I’m sure, you know, it will be easy to do. But, the point being that these asteroids are tumbling in ways that are really hard to predict, and the asteroids interact with each other and with other bodies in the solar system in ways which are really hard to predict. And, your simulation runs out of steam beyond a certain point, trying to look into the future. So, what are the limits of simulation. How far can we go, and what can we not simulate?
Pamela: The inside of a black hole. So, if you want to know what we can’t do, that’s one of those things we can’t do. We can simulate what happened prior to the release of the cosmic microwave background, but we can’t yet test those simulations, necessarily. So, there are lots of random holes where we kind of get stuck, but yeah, it’s so hard to say what we can’t do because someone is going to prove me wrong with an unpublished dissertation as soon as I say anything other than black hole.
Fraser: Other than black holes, but you feel pretty confident –
Pamela: I feel totally confident.
Fraser: You know, this is funny. – Time for a brief rabbit hole – which is that even gravitational waves can’t show us what’s going on inside the event horizon of a black hole.
Pamela: Yeah. It’s totally true.
Fraser: It is forever beyond our reach.
Pamela: And, that’s okay.
Fraser: Yarkovsky-O’Keefe-Radzievskii-Paddack effect. YORP. It’s named after the four researchers who developed parts of it. So, that’s what YORP means.
Pamela: Okay.
Fraser: But, let’s give some people some recommendations for some software to buy, or some gifts to give people to get them simulating. So, what would you – Let’s give some people some recommendations. What would you suggest they go out and buy?
Pamela: So, as a starting point, Universe Sandbox, and Kerbal Space – to start with those two. If you want to start with free, another good one to explore existing spacecraft is NASA Eyes – so eyes.nasa.gov, and to explore the actual universe is Worldwide Telescope, and then Celestia or Solaria.
Fraser: And, there’s another game called Space Engine, which I know a lot of people want me to check out, which I haven’t actually tried, but I’m sure people would be like, yeah, Space Engine is cool. The other thing is Stellarium, which is great to simulate what the sky is going to look like in your area, and then you can drive that forward to be able to see what it’s going to look like in the future. And, there’s one more thing.
If you go to Transit Finder – I think it’s www.transit-finder.com – But, anyway, do a search for transit finder, and it will simulate when the International Space Station is going to pass in front of the moon, or the sun within a certain range of your house, and the exact date and time that you can do. So, you can then go at that exact time, go to that location, look up, and you will see the International Space Station fly right in front of the moon perfectly, and it just seemed like a wizard, which is awesome. I have one coming up on the 4th, I think of December, that’s going to be about 30 kilometers from my house.
Pamela: Are you going to try to photograph that?
Fraser: I may try, yeah. I may try to go. I don’t know if I’m going to try to photograph it, although –
Pamela: Video it. That would be easier.
Fraser: Or, just go out and enjoy it. Cool. Well, thanks, Pamela.
Pamela: It’s been my pleasure.
Male Speaker: Thank you for listening to Astronomy Cast, a non-profit resource provided by Astroshpere New Media Association, Fraser Cain, and Dr. Pamela Gay. You can find show notes and transcripts for every episode at AstronomyCast.com. You can email us at info@astronomycast.com; tweet us at Astronomy Cast; like us on Facebook; or, circle us on Google Plus. We record our show live on YouTube every Friday at 1:30 p.m. Pacific, 4:30 p.m. Eastern, or 20:30 GMT. If you miss the live event, you can always catch up over at Cosmoquest.org, or on our YouTube page.
To subscribe to the show onto your pod-catching software at astronomycast.com/podcast.xml, or subscribe directly from iTunes. If you would like to listen to the full, unedited episode, including the live viewers questions and answers, you can subscribe to astronomycast.com/feed/fullraw. Our music is provide by Travis Searl, and the show is edited by Chad Weber. This episode of Astronomy Cast was made possible thanks to donations by people like you. Please give by going to patreon.com/astronomycast.
[End of Audio]
Duration: 31 minutes

Download the show [MP3] | Jump to Shownotes | Jump to Transcript


Submit a Comment

Your email address will not be published.