08:05:12 Welcome, everyone. Good morning.
08:05:16 Thank you for joining us for our KITP fundamentals of gases halos program today.
08:05:24 I'm Cameron Hummels, one of the core co organizers and the moderator for this week's program.
08:05:31 And this week, we will be focusing on the theme, what rules do non thermal components play in defining the circle galactic medium. As you can see we have a different theme for each week so pick out the ones that you're interested in and make sure to be
08:05:45 here for the maybe all of them, maybe you're excited about all of them, I certainly am, I will be here for all eight weeks.
08:05:54 Just a few announcements, as we roll into our first keynote of the week.
08:06:01 First of all, you probably already know this, but just to make sure everyone does. We have a full list of the program details including detailed schedule recordings of all of almost all of the events, we leave off recordings for the speed collaboration
08:06:17 on Mondays, and for the discussion on Fridays to allow people to speak freely without it without concerns that their, their comments will be recorded in perpetuity.
08:06:28 We just added due to feedback from the community, we've added transcripts to the various talks that will be included on this as well as the KTP website, along with slides and, and, and recordings.
08:06:45 And so this is really the central place that you can go to get information about the program, other than just contacting people directly on slack.
08:06:55 Something in response to the surveys that we put out, we you know we put out a different feedback survey every week because this is really an experiment and we're trying to get improvements on how to do this, this properly and address all of the issues
08:07:09 involved. So one of the pieces of feedback we received the last two weeks, is that there aren't a lot of opportunities for junior scientists who are on the job market to gain exposure for their work and really sell themselves to potential employers.
08:07:25 It's difficult within the to our constraints of each of our days events are already kind of jam packed with with activities and because we don't have the associated conference with the program.
08:07:37 This is challenging so we've created the halo 21 on the market Slack channel where people can submit 10 to 15 minute videos for junior scientists who are either on the market right now, or will soon be on the job market.
08:07:53 To help, you know, gain exposure for their work and and get there, get their name out there, so go if you're interested, either as a, as a junior scientist or you are a potential employer right now, or will be soon.
08:08:07 I encourage you to check out that channel, we will be moving those videos just as we have been for the halo 21 new results videos we will be putting those on our YouTube site, which you can look up, it's also linked in this channel.
08:08:25 And, and then yeah, check out what some of the cool. The cool activities and the cool research that Junior scientists are doing.
08:08:32 As I mentioned, this is feedback that we got from our survey so also.
08:08:38 Each week, we'll have a different survey and and yeah please, please take two minutes and fill that out and give us feedback on what we're doing right and what we're doing wrong, and hopefully we can make this better for everyone.
08:08:49 So today's activities we have a keynote from Dr Pang, oh from UCSB for about an hour or so, and then we'll follow it up with a discussion for the remaining hour.
08:09:09 A panel, made up of both Pang, as well as other experts in the field of MHD and turbulence and non thermal processes so Blakeslee Burkhardt Chad bustard arena boots ski and Evans can have Keiko will be joining us.
08:09:18 I'll be moderating and I'll be primarily getting questions and discussion topics from the Slack channel Halo 21 dash week three dash non thermal, so each week we have a different channel, where we try and have a lot of the discussions and comments, particularly
08:09:34 during these keynotes occur as opposed to within the zoom, because the zoom essentially all the comments go away, and there's no way to thread comments into multiple discussions, but slack allows that to occur so please during the presentation if you
08:09:50 have a question, even just a clarification question or a comment on something that ping says, type it into Halo 21 week three non thermal slack chat and other people can respond, maybe address that question, right then.
08:10:09 So that, that, that you can get on and still have clarification on what's going on.
08:10:13 Alright so, a brief introduction to ping ping is a professor of theoretical astrophysics astrophysics at UC Santa Barbara.
08:10:24 He has a huge breadth of research topics over the last 30 years, covering everything from galactic dynamics to stellar feedback to turbulence to cosmic rays magnetic fields thermal instability hydrodynamic, the whole gamut.
08:10:43 And not only that, but he's extremely energizing and enthusiastic both in his presentations, as well as just like having a conversation with him, I always look forward to running into paying at conferences and.
08:10:58 And yeah, hopefully, you'll, you'll get the same idea after, after this talk. So, thank you for joining us ping Do you want to you want to share your screen.
08:11:09 Thank you, Cameron, I share the love.
08:11:14 Can you Is it is it full screen can you see it full screen, it's windowed.
08:11:20 Even now.
08:11:23 There, now it is. Okay, great. Okay, so I don't know about the first part but I will try to entertain you. So I'll try to keep, keep faith with a second.
08:11:34 So whoops, so okay so non thermal processes this week, I'm so used to be that you know whenever you want it to annoy the speaker, you could always put up your hand and an Astro talk and say, What about magnetic fields.
08:11:49 but in the CGM we actually have three Boogie men right we have magnetic fields like before. And there's magnetic fields, you know without them we would not be around right would get zapped by the sun.
08:12:02 So thank goodness for them. There's also cosmic rays, and there you see Victor has going up in a balloon discovering cosmic rays he discovered that radioactivity increased as you went away from the earth, instead of the other way around.
08:12:17 And then there's also the perennial problem of turbulence right this famous unsolved famously difficult problem. So, these are you know the doc horseman that we have to deal with in the CGM.
08:12:31 So you know why should you care about them. So, I would argue on two grounds. And the first is just equal vote, right in the isn.
08:12:43 All of these components are comparable in energy density to the gas thermal pressure and when we write on hydro stimulation pure hydro, a lot of the times we're just focusing on gas when there's all this other stuff around.
08:13:00 And by the way in the ISS, this question of why they're all equal. I've asked a different is them theorists and, uh, you know, you get different answers ranging from, it's trivial to.
08:13:12 That's not important to, it's a coincidence. So I'm curious to see what people think about that in galaxy clusters, which are another Halo that we've talked about.
08:13:24 cluster cluster cluster ICM folks are well aware that even though they are maybe even one to 10% of the energy density there, they make a huge difference.
08:13:40 So people there are very conscious of, including these in, you know, they're considerations of what's going on.
08:13:47 And the CGM our constraints are really quite poor, so we have to consider a very wide range of possibilities.
08:13:58 And the other thing I would argue is that it's just fun. You know, there's a lot of very rich physics, when you consider that and you know if you can't fit things with standard hydro, the chances is that one of these three things is playing a role.
08:14:15 Okay. And, you know, our fearless leader this week Cameron said please make sure you appeal to the younger folks in your top. So, I thought a little bit about that and I have some semi philosophical rambling, which may or may not be useful, but you know
08:14:32 this this quote from Einstein. He says I have little patience for scientists who take a board of would look for its thinnest part and drill a great number of holes when the drilling is easy.
08:14:45 And that's entirely my approach to research, I'm afraid.
08:14:50 So I want you know, I'm going to try and give some review, but I'm also going to try to talk about some very simple minded, a numerical experiments, which may not seem to connect with this grand goal of understanding how galaxies form and how they work.
08:15:09 Um, but I want to offer a defense of simple mindedness. So, you know, when I was in grad school, um, you know, the, what are the great things about our field is that you know you're surrounded by really smart, talented people.
08:15:24 But that's also one of the dangers of our field right so when I was in grad school, I had what is I think we now call imposter syndrome. Other back then be I don't think we had a name for it so you know you wake up, you feel.
08:15:37 I don't know what's going on. Oh my god, I'm so dumb. It's horrible. And I'll never forget the day when you know I went to tea. And there's this famous Professor bow down for Kinski he was, he was sitting there and he was holding court and talking about
08:16:01 some in comprehensible thing. And then suddenly he jumped up and draw on the blackboard this diagram. And it's interesting, and difficult. And he said, you guys.
08:16:06 Too many of you are living here, you're working on really hard things, trying to show that you can climb up Mount Everest barefoot. And at the end, there's five people in the world who care.
08:16:21 And, you know, there are other parts of this diagram you can do easy and boring problems but then there's no job, uh, this is great if you are Einstein.
08:16:31 But really, where you want to be is here. And I would argue, if you listen to the things that we are talking about in this conference, you know, I'm going to talk about okay you take a box of gas, you stir it, and you let it cool, or you take to fluids
08:16:45 and you send the sound waves through them what happens right this is not string theory here right we're talking about very simple minded things. And I would argue that's actually a good sign.
08:16:57 It's a sign of a field that you can get around you can, you know, you can call your mom.
08:17:04 Tell her what you're doing, and she'll get it right away. And so there's a lot of low hanging fruit, and when things start getting complicated, then I don't know you're venturing over here anyway okay that's my cheek pop philosophy.
08:17:18 Okay, on to the science. So, magnetic fields, so why should you care about magnetic fields. So if you're a charged particle, you will experience a Lorenz force from magnetic field, and you're going to spiral around it.
08:17:32 Right, so if you're a proton. That results in an isotopic momentum transport, you have an isotopic viscosity.
08:17:41 If your electron, you know, that's an isotopic heat transport, so and isotopic conduction of your cosmic ray you know you're, you also have this power on this few lines and that has many implications interesting implications for cosmically transport right
08:17:55 so Alan I think on Thursday, we'll talk a lot about pressure and I saw it up, and things like that, but you know it's something always to keep in mind.
08:18:05 The other thing about magnetic fields, is that it creates an HD forces right so there's magnetic pressure when it feels don't like to squeeze together resist it.
08:18:16 And magnetic tension, where felines don't like to bed. They also resist that.
08:18:23 Now in terms of how magnetic fields grow, you know that's a whole other complicated can of worms.
08:18:29 Ellen is an expert in that.
08:18:31 But in, in terms of, you know, maybe the one thing you should keep in mind is that turbulence can grow magnetic field so either bogeyman can grow magnetic fields by something called the turbulent Dynamo, and a simple kind of rule of thumb is that you
08:18:46 can grow quite quickly when the field is very weak you can grow exponentially. Eventually you go linearly and you saturate very roughly, when the kinetic energy density is of order the magnetic energy density so you get equity partition.
08:19:03 And the reason is because at that point make no tension, sort of doesn't let you bend the few lines as easily as here.
08:19:10 So this might be, you know, a rough, you know hand waving explanation for why you have a competition between magnetic fields and turbulence.
08:19:20 Now the magnetic field in the CGM is very uncertain. Um, it was great to have that speak collaboration So Brian told me about this very interesting result, where people see coherent magnetic fields of to micro Gauss in magnesium.
08:19:38 magnesium to absorbers.
08:19:40 And there are also FRB constraints.
08:19:44 You know from 30 rotation where basically you get something like, you know, a micro Gauss or a bit less you know in halos, you know, these are actually substantial numbers, these are comparable to is and values, but it's far out in the CGM right where
08:20:02 think the gas pressure has dropped by orders of magnitude. So one of my hobby horses, is that, you know, it's conceivable that the CGM could be magnetically dominated or it is a very important component and we cannot ignore it, right.
08:20:19 So, for instance, in these beautiful simulations by Fraker, uh, you know, you can see here that in these by conical outflows, you have extremely low, what plasma physics is called beta right the ratio of thermal to magnetic energy.
08:20:38 So, here, you're completely magnetically dominated. And so it's important thing to keep in mind. Now, you know, theory is very it's very unclear at this point you know the fire simulations, get a, they don't, they they're always totally dominated the
08:20:55 Halo, but it's still early days.
08:20:59 So, you know, what are some implications of magnetic fields in the CGM.
08:21:05 So they completely change clock morphology, so this is a recent set of simulations by Michael Jennings and you only where you know they show okay if you have hydro, then you make these clouds that we've all been talking about it if you have MHD, and even
08:21:22 though the magnetic energy density is 110 thousands of the thermal one. Look, you have filaments the mythology is completely different.
08:21:33 And the other thing is magnetic pressure support right something very near and dear to say, Justice heart that you know this, this, This means that you don't have to enforce them more pressure balance between hot and cold phases.
08:21:48 And these are cloud crushing simulations that that Max has, where you can see in the clouds, the clouds themselves which condensed out of the coal, gas, they're magnetically dominated.
08:22:02 Another important effect is that, you know, this breakup of cold gas by by hydro instabilities for instance the Kelvin Helmholtz instability, can be suppressed by magnetic fields.
08:22:21 The main if something is moving through an Ambien medium with magnetic fields you draped around the cloud. And typically you reach magnetic energy densities, which are comparable to the ram pressure right and then at that point, magnetic tension is strong
08:22:33 enough to damp instabilities. So here's an early illustration partners in Frommer.
08:22:41 And another point is that magnetic drag increases momentum coupling of hot and cold phases basically you have a total between the two phases, so that if you try to move one of them, the medic fields hat help transmit momentum to the other face and so
08:22:59 when you push it on one of them. And so as you increase the magnetic field strength, this distance shortens.
08:23:21 So, there's a huge amount to talk about, you know, a lot of people in the audience and the panel of experts so I'm, you know, I'm afraid I'm not going to be because of all of it.
08:23:31 One thing I will say is that, you know, we think that we all admit turbulence is very hot. Right. And also, we all admit that plasma physics and cosmic rays are also very difficult and so we think fine you know if there are things we don't understand.
08:23:50 That's good. But, you know, MHD forces we think our policy, you know, like come on, like, let's be serious here we know what they are they're just two of them.
08:24:01 I just want to, to, to say that sometimes they behave in ways that at least, at least I find a bit sorry, the sun counter intuitive and you still have to run the simulation.
08:24:14 So here's an example from searching G, who's now at Caltech, and I think he's, he's giving a talk on Wednesday.
08:24:22 But I'm back in 2018 he looked at the influence of magnetic fields on thermal instability.
08:24:30 And the something that he talked about this this ratio of the cooling time to the freefall time.
08:24:37 And what searching found was, was that magnetic fields can have a significant effect. This is a cold fraction of gas as a function of this ratio.
08:24:47 And, you know, as you increase the main field, even a small magnetic field which is like 3% of the gas pressure can still have a very big effect.
08:25:00 And you know, at the time, you know, searching was was not the the famous Caltech fellow that he is now. And Mike and I were like, oh man, there's a bug in your code like go look at it.
08:25:13 We were of course he was right. And the reason is that, You know, what you care about our perturbed forces right so gas pressure is big, but it's, you know, gas great pressure gradients are posing something else that's really big which is gravity, the
08:25:31 difference of two big numbers is a small number. And that's where a small magnetic field can also change a change the system.
08:25:40 And so okay fine. Great. So magnetic fields suppress you know in Fall they suppress buoyancy. So it makes sense that, you know, the freefall time becomes less important.
08:25:53 And it's just magnetic tension. But then the other head scratcher was like look. Okay, that's what happens if you have fields that are horizontal. Now let's turn the fields around the guy should just be able to fall freely like before.
08:26:08 And we should get the hydro answer no you actually get the same density perturbations, even though the mythology of the call gas is completely different.
08:26:19 So again we were like, Oh man. What's this, but then in retrospect, it's also quite obvious right so if you look at trace of particles. They are also prevented from falling.
08:26:32 And the reason is because it's like you're surrounded the magnetic field or like confining pipes. If you try to fall, you make a high pressure region, and that suppresses your info and the strong the main magnetic field, the stronger the confining wall.
08:26:47 And so you can work it out it suppresses information in exactly the same way.
08:26:53 So maybe these are obvious to you but at least you know at the time it was not obvious to us.
08:27:02 And, and, for instance, here's another head scratcher.
08:27:05 This is a simulation of thermal instability by young Phaedra.
08:27:26 I was on the panel on Thursday. And I'm looking at this, like what is this, these are these there are these sort of tadpoles racing away. And I'm, I don't know.
08:27:32 So, at least to me it's a little bit of a head scratcher.
08:27:39 with solid tongues but you know if any of you have ideas, I'd be interested. Okay.
08:27:41 So, um, you know, I've got, I worked out I got about 15 minutes per component. So let's move on to turbulence and turbulence is a famous Lehigh problem.
08:27:54 You know, Verna Heisenberg said that he thought God understood relativity, but not turbulence.
08:28:01 So, uh, what is it, right. So, I think in the first week or second week, Joel brakeman reminded us that we astronomers just think of turbulence is anything that's not a thermal right so so like, you know, all the bolt motions that we see in spectra we
08:28:21 call turbulence. But, strictly speaking, you know turbulence involves and energy cascade usually from large to small skills but not always.
08:28:29 And it's due to this nonlinear time in the hydro equations, right this bad thing is responsible for a lot of head scratching, you know, trillions of hours of CPU time, you know, all that stuff.
08:28:46 If we didn't have this, then we would be like in quantum mechanics, or lot of eNm you know you have a linear system you can write down basis functions you superposition.
08:28:55 You know, it's so much better.
08:28:57 But no, and the sort of standard.
08:29:03 But no, and the sort of standard uh you know the probably the most important thing to know about turbulence is coma girl of turbulence, where basically you say there are these things called Eddie's, and they go from large to small scales there's a cascade
08:29:28 from large to small scales. After an ad turn over time. One of these things goes nonlinear and breaks up into Dodger Eddie's and this keeps on going. And there's this very nice ditty that sort of summarizes this, there is no dissipation until you reach
08:29:35 sort of molecular scales viscous scales. So in between. There's a constant energy flux from large to small scales, which choreograph wrote down in this way, and that it's very easy to apply.
08:29:52 And, um, you know, it's the sort of standard go to every time you try to estimate something.
08:29:57 you try to estimate something. But there is more to life to the to life and karma graph, it's very good to remember.
08:30:04 Right. Instead of a cascade like this for instance if you push, if you push on something, especially a bit of a shock right you can directly transfer power from large scales to small some things, some people sometimes called burgers turbulence right if
08:30:18 you have a magnetic field that changes the character of turbulence there three independent cascades then.
08:30:26 If you have stratified turbulence, then turbulence in a gravitational field.
08:30:32 Then vertical motions will be suppressed by buoyancy and the turbulence, again, becomes an issue tropic, and you can watch great YouTube videos on this by our fearless leader on the presentation channel.
08:30:48 That's the king turbulence. Okay, lots of stuff. Now, why should you care about turbulence.
08:30:53 It can provide pressure support again. But it's pressure support with a time bomb right because turbulence you always have to drive it because it's always you know cascading and decay.
08:31:04 So it's sort of different say from my network support.
08:31:07 It can also diffuse things and crucially it diffusers things a lot faster than say, you know, standard thermal random motions right if you rely on a thumb emotions to diffuse your sugar, you would, you know you'd never going to get your sugar hit.
08:31:25 So you need you know turbulence can diffuse metals and can diffuse entropy.
08:31:31 So it's very important for that.
08:31:34 It tangles magnetic fields. It very much changes that geometry, because the man feels just move with the fluid, and by the turbulent Dynamo that I mentioned earlier, it also amplifies them.
08:31:50 It heats the gas so there's turbulent dissipation so at the end of that cascade all the energy in the turbulent cascade is eventually converted to random motions in the, in the, in thermal particles right so heats the gas.
08:32:07 And a lot of these ingredients are put together.
08:32:11 When you consider the multi phase structure of of the CGM right consider hot and cold phases, so I will, you know, there's so much to talk about all this stuff, and you know people on the panel I've done work on all of this, I'm sure you'll hear about
08:32:27 this from them. I'm going to focus on this aspect.
08:32:31 Okay and just one quick reminder is that you know what we call turbulence you know classically turbulence is something that has a very high Reynolds number.
08:32:43 You know rounds number is a is a dimensional is measure of viscosity, right. If you have a high Reynolds number then viscosity is not terribly important.
08:32:55 Um, you know, in all simulations because of limited dynamic range, you, you, you all, you know you're very much affected by numerical viscosity. So for this purposes of a plot a straight line is good, and a curved line is bad, you know, a straight line
08:33:11 is very roughly, your inertial range where you are in this cascading region that coma graph talked about. And so you're affected by numerical viscosity, you know, 20 to 30 times above the grid scale you have, usually you have like less than a decade of
08:33:27 inertial range, and your standard say 512 cubes simulation, so it's just something to keep in mind, you know, like, especially if you try to apply skillings in your simulations.
08:33:39 Okay, so this is for drama and you know drama and said, This week we must talk about this. And so being an obedient soldier, here's a, here's some thoughts about turbulence and read of cooling.
08:33:55 So, I want to advertise this paper by Brent time was a graduate student at UCSB, which looks at the parallels between turbulence and really of cooling and turbulent combustion right so this is something that even, you know Zelda which appreciate it a
08:34:15 long time ago, you know, here you know you you take fuel and oxidizer, and you burn the fuel here you mix two phases, and you burn hot gas to make cold gas for condensation or vice versa.
08:34:30 And there's lots of scaling relations and stuff like that that you can understand in detail. Through this analogy. I'm, you know I'm not gonna have time to talk a lot about this and drama talked quite a bit about these mixing layers.
08:34:49 So I'm just going to highlight two things. One is, in regards to something that Frank Vanden Bosch raised and which worried us a lot, which is how important is numerical diffusion of heat between the phases in affecting your answer.
08:35:06 So we did a lot of resolution studies studies of, you know, do you have to put in conduction and resolve the field length, and all of that. And the short answer is that it's not very it's almost unimportant.
08:35:19 And it's the same reason why the value of numerical of molecular viscosity is not important. When you're mixing in cream in your coffee. You basically stir your cup and you get the thinner and thinner ribbons until likely diffusion takes place, but if
08:35:37 you change the level of like a diffusion, it doesn't really change things right if you change.
08:35:44 say if you're trying to heat a gas and you stir it. All that matters is the heat input you put in on the outer scale, eventually it will cascade down to some scale and disappear.
08:35:55 So, so our short answer is that you really just need to resolve the outer scales of turbulence.
08:36:02 The second thing which I thought was, was was interesting was that, um, you know when your cooling in a multi phase medium that effective cooling time is not the thermodynamic tooling time.
08:36:15 Right.
08:36:16 But it's also not just the mixing timescale right when you mix the gas can change entropy, right and so you might expect that to set a timescale.
08:36:27 Instead, it's the geometric mean of these two things. And I'm, you know that that we you can demonstrate that in the simulations, and that's what Brent did.
08:36:38 And, you know, it's actually very similar to a situation like in conduction, where the field length is actually geometric mean of the electron elastic mean free path.
08:36:50 And the cooling length of the electron, or in relative transfer suppose you have a photon bouncing around it undergoes some elastic scattering and undergoes some absorption, then you know if you do like the first few pages of right became Lightman.
08:37:09 The derive that the effective optical depth is actually the geometric need of the absorption optical depth and the scattering optical. So it's a very similar situation.
08:37:19 Okay. And I think the other thing that would be an interest, that's not in that paper is something that Brent is working on now is that you know you would like actually to have a, to be able to say put in some of this physics in larger scale simulations
08:37:36 where you cannot resolve these mixing layers right so if you, if you look at these mixing layers they're extraordinarily complicated, the mixing layers themselves are multi phase, and there's not a single position where a gas at a given temperature comes
08:37:51 from. It actually comes from a whole variety of different positions in the mixing layer.
08:37:59 But nonetheless, actually Brent has found that you can actually come up with a very simple one D analytic model that reproduces the results of the simulation quite well.
08:38:09 And you can come up with 1d temperature histograms, which is really a, you know what, say observers care about you, if you mass wait this, you get calling density ratios.
08:38:23 If you admission with this, then you get you know very you know ratios of emissions strength of different metal lines. So, so we're still working on this but I think, you know, we're excited about this and think that this is promising.
08:38:38 Whoops. And now, okay so there was a lot of discussion about, you know, fog and clouds last week, whoops, and, you know, Todd. Todd trip was very nice to us theorists right and he said, Oh observations need to catch up with theory.
08:39:03 You know, I think, I think he was, he was just being nice right really theorists are always chasing the observers, and it's no different in this case, you know, my mic when you thought about the fog was completely motivated by observations and the observation
08:39:14 he was motivated by was this cluster, which was this high redshift object that that Joe had now we talked about at UCSB, where he said, Look, there's this, there's this big object which we see both in absorption and emission from for ionization analysis
08:39:34 we see this dense gas.
08:39:36 With with coal bullets. But why is it covering this large area, okay it's covering this enormous area. When you can work out that given this density, and the monocle gas you see the volume filling fraction is tiny.
08:39:54 Right, so you, what are you going to do, you're going to put a very thin shell around the whole Halo.
08:40:00 And, you know the other problem that Joe mentioned was okay you've got these highly Supra thumb align with, you've got these low ionization lines which live at 10 kilometers per second.
08:40:14 And you know some of these optically thin and look, You know, hundred thousands of kilometers per second.
08:40:21 A anything that has a Mach number of 100 is not going to live, it's you know it's just not going to happen. So, um, you know, Mike, you know, the moment after Joe finishes talk.
08:40:36 Mike came to my office and he said, well it's obvious, unlike well maybe it's obvious to you it's not obvious to be, and he started talking about this fall, so it was you know it, I would say that that theorists, you know, are trying to invent ways to
08:40:50 get the fog, but in some systems, there is. I don't know how else to explain these two things right. Maybe I'll be interested to hear some thoughts on that.
08:41:02 So in that case, you you intersect many cloud outlets along the line of sight. And then these two observations make sense.
08:41:11 Oops.
08:41:13 Okay. So, but, you know, there's a lot of interest in sizes. And so, you know, I have three young boys, so everything's in terms of a battle. There is correct T rex Vs Dinosaurs, you know, Batman vs Superman.
08:41:30 So, Now let me give you a if I'm going to be a fight promoter might vs Max, okay there, it for those of you know you you know them. The both very tall, handsome exceptionally charismatic.
08:41:47 You know guys you know your your classic alpha males.
08:41:51 Even my wife would tell you I'm somewhat of a beta male. Okay so, so first you know what you talked to, whoops, how do I go back.
08:42:02 Oh, whoops.
08:42:05 Okay, so first you talk to Mike and Mike will tell you. Okay, look at look at look at these simulations. I had a big cloud. And now it's in these tiny pieces.
08:42:15 What happened.
08:42:17 There was cooling. That was strong thermal pressure gradients, and the cloud boss Sonic contact. and so it shot it down to tiny pieces until they could eventually reestablished on a contact.
08:42:32 And that happened at a scale CST pool of the coal, gas, and I was like, oh man, you're so right Mike, I'm gonna write a paper on this, and then you talked to Max.
08:42:45 Max will tell you.
08:42:47 Look, there's a minimum size for a cloud to survive right you blow something on it and mix it with a hot gas, and it's going to be destroyed in a cloud crushing time unless cooling can overwhelm the mixing.
08:43:01 There is a characteristic length scale for this, and it's essentially CST cool mix. Right. When the cooling time of the mixed gas is shorter than the cloud crushing time.
08:43:14 And you know you like max You're so right. We have to write a paper on this. Now there's a problem.
08:43:39 Which means that all these beautiful little cloudless will get destroyed.
08:43:34 Right. And this limescale is bigger than this one. So according to Mike, these things are just going to be destroyed by some old pressured radiance which develop.
08:43:41 Okay. Um, so, what to do, who is right.
08:43:47 And, um, you know, Max has been thinking about this more. And we think now that you know of course, like you know like in all things, you know, both are potentially right.
08:44:01 So, what's the difference so instead of doing this, right where you in both of those previous works you considered laminar flow right you blow on it with a win or you do a static setup and stuff instead you know most gas in the CGM lives in illegal hurricane
08:44:20 right that's all kinds of noise and perturbations going on. Right, so, uh, you know, this is something that a drum and fielding and Evan Schneider, you know, always rightly mentioned, so you would you know you would go to conferences and try to avoid
08:44:37 eye contact, but now I can reestablish eye contact.
08:44:41 So, so here is Max blowing, uh, you know, starring on a cloud right in a box, you know, with a mixture so annoying compressed of driving with trace of particles.
08:44:58 And there's many very nice features in this, right, so you know you can do a clump analysis, and then, you know, even after the mass has grown by quite a bit.
08:45:10 The mass is broadly distributed long range of skills.
08:45:14 The big cloud still survives, right, but not all the masses concentrated in it right there's still a lot of mass in smaller clouds.
08:45:26 And in particular, although the mass is quite evenly so this is a cumulative plot.
08:45:30 Even though the mass is broadly distributed.
08:45:33 The area is dominated by the small cloud. So this is, this is area covering fraction of a volume covering fraction normalize in such a way that for a sphere.
08:45:43 This is important one. Now if you look at a big clock, the big cloud, you know, sort of, stays roughly spherical.
08:45:51 But if you look at the entire box. The, the area covering fraction has grown enormously, meaning that and it will just continue to grow. Meaning that if you here's a random line of sight, through the box, you are much more likely to kiss the fog, then
08:46:19 are to kiss the cloud, even though the cloud contains a lot of the mass. Right, so, so that's something to keep in mind when you interpret observations. The other thing you know when we started this we thought, oh, what's going to happen is that the small
08:46:25 things will coagulate to make a big cloud, which can survive.
08:46:30 And you know, I think, in the past, some of you heard me mumbling about a peloton that turns out actually not to be the case. It's true for a laminar flow, but it's not true for turbulent flow, you need a big cloud in the beginning to survive.
08:46:46 If you just have an equivalent mass in small clouds, even though intuitively it should be similar, especially if they packed closely together, it doesn't work, they all die.
08:46:58 And the mass threshold for that big cloud is similar to the standard cloud crushing problem when you blow in a single big cloud with a wind.
08:47:08 And then, you know, the growth that whole growth of gas, you can kind of understand with analytic models, there's some interesting features when the small clouds dominate the area.
08:47:19 So, I'm, you know I'm really running out of time here, um, one thing that is important to keep in mind is that this thing is very you know with turbulence, it's very highly stochastic.
08:47:33 You can't just look at, you know, a handful of clouds in your simulation and draw conclusions you need to ensemble average. So here's an example from Max where there's exactly same initial conditions, Toby learns Same, same string parameters, just a different
08:47:51 random seed.
08:47:53 This one survives, and if you keep running the simulation the mass keeps growing and growing.
08:47:58 This one dies, and yet there's no more coal, gas, there is nothing different about the initial conditions for this.
08:48:08 And you know, it's because of the stochastic nature of turbulence. So you really need to have either very big box or two ensemble average. Right. So in this sense if you bring out, you know, it's more like thermodynamics, or quantum mechanics than classical
08:48:23 mechanics, you need to talk about probabilities.
08:48:28 And you know I think eventually you know Monte Carlo approach to, to understand this would could be fruitful right where you, you have steady accretion, you know, of course, of hot gas too cold, but also you have breakup, you know the opposite emerges
08:48:42 and you can imagine, you know, having a tree where you implement this some something like the merger trees in cosmology.
08:48:52 Okay, um, I am super out of time. So, so let me, I'm gonna have to skip the note version, it really pains me to do this but, you know, I'm you know the Hollywood version I guess of this.
08:49:07 This picture is that you have this big mothership. Okay, which launches the small droplets. The mist.
08:49:12 and these droplets are transient, they don't survive.
08:49:17 But the big ship, you know, because you know it's able to to it you know it, the cooling time is shorter than the breakup time, it will keep going, and it will keep making small droplets.
08:49:32 And so that's roughly, our picture of how things done now so I think they're there are both clouds and missed in the CGM and we must consider both of them.
08:49:45 They have both have interesting consequences.
08:49:48 Okay, so a cosmic rays in the last 10 minutes.
08:49:54 Okay, so why should you care about cosmic rays. So, they provide an ensemble pressure support.
08:50:02 You know, and here's a nice paper by Irina Buzzi was on the panel, which shows that if you have thermal instability in the Halo, then you can have cosmic ray pressure support it, filaments.
08:50:15 And one important thing about cosmic rays, is that actually that the pressure they provide is is more or less agnostic to the direction of the magnetic field.
08:50:25 So the cosmic rays pressure
08:50:32 forces are isotopic right the just you just have to calculate grad PC, not you don't have to do any projections along the magnetic field.
08:50:44 Um, then they also, you know, the of course, the sexiest application which many of you have written a lot of papers about is that the drive collected wins, right and you know that not subject to relative losses like thermal gas, and they can dominate
08:51:02 in the halo I think so GG will be talking about that on Wednesday. So, you know, and they can drive a cold when which is sort of more in line with what we see and observations and something which are also very much like which Josh wrote a paper with with
08:51:19 with Elena me was that they can also influence, because they provide pressure support and they provide heating, they can influence thermal interfaces, right.
08:51:33 So, and you know, we, there's a lot of literature on conduction friends, and it's worth also thinking about cosmic ray friends as well.
08:51:43 Okay, so, um, you know, Ellen has written two beautiful reviews about this, and she's going to talk on Thursday, so I'm not going to do too much of a cosmic physics.
08:51:55 Just a few very basic things, which is that cosmic rays, you know the light crossing dr relativistic particles, and the light crossing time of a galaxy is short.
08:52:12 But nonetheless, we have evidence from speculation and a bunch of other things that they live in our galaxy for a long time. And the reason is because they don't just fly straight out the random walk.
08:52:20 And we think the reason why they random walk, is that they scatter off magnetic irregularities in in the Halo, and they make these magnetic irregularities themselves something called self confinement.
08:52:34 It's also also possible for extrinsic turbulence to do it but there's a variety of complicated reasons why we think the streaming instability should dominate for the sort of gv cosmic rays that we care about.
08:52:50 So okay so they make kings in the magnetic field, and then they scatter off them, and the result is there's a short mean free path. And when you have a short mean free path, you can consider it to be a fluid, right and you can write down fluid equations.
08:53:08 And, you know, here's a, you know, that that some nomenclature we think that the stream.
08:53:16 Okay, I'm going to skip this. Basically, what you care about is they push on the gas right push on the gas that this force. And they also heat the gas.
08:53:29 Right. And, you know, this is intuitive right it's just like thermal pressure, you have a gradient over some pressure here be a gradient of cosmic refresher, and the heat the gas, this is just a velocity times a force right so that's a rate of work done.
08:53:41 work done. Okay, there you go you can hold your own in a cocktail party conversation. So, um, so.
08:53:50 Okay, so Jaan fe came up with a very clever way of handling cosmic re transport.
08:54:00 So that, which has, which is basically how people do it and read of transfer. And the bottom line is that now.
08:54:10 You know, everyone we can all do it in full generality. So, and you know you can apply it to something like shocks and it still works you know you reproduce analytics solutions.
08:54:24 Very well, which is something that before it's just too demanding that shocks such short line skills involved it's almost impossible to do with when you have cosmic ray streaming.
08:54:35 Okay, so I'm just gonna mention then, right, what now you have you have all these rich things that you can go after. If you can include both diffusers and streaming transport in your simulations.
08:54:50 What is interesting.
08:54:52 So, I will briefly mentioned two applications. One is the interaction of cosmic rays with turbulence.
08:54:59 And the other one is interaction with cosmic rays with sound waves.
08:55:04 So, uh, so for the first one way to cosmic rays come from right so the standard picture is a Fermi acceleration basically they scatter off, moving paddles, which accelerate them right so there's some sort of scattering right off magnetic irregularities,
08:55:25 and that increases their energy, right, um, and there's two effects. There's first order and second order for me acceleration. If you have a random configuration of scattering.
08:55:39 Right then, what you rely on what Fermi argued in his original argument was that you're going to end up with slightly more head on collisions than tail on collisions.
08:55:49 And so you'll have a net, a diffusion upward in energy, but it's only second order in the Blasio scatters competitive speed of light, so this is a very small number, uh, you know, then later on, it was, it was discovered okay if you have a converging
08:56:07 flow if you have two paddles moving towards one another, then you are first order and is a lot faster and this is what happens in shocks. And this is a you know this is currently what we think, actually happened so all the, you know, all the simulations
08:56:22 that you see they assume that cosmic rays are accelerated in shocks, which are set off by supernova.
08:56:32 Um, but, you know, I'm going to argue that we might we should think carefully about this assumption, it is worth revisiting. So this is worth by Chad Blue Star, who's who's also on the panel.
08:56:47 He's a key as a Kitt fellow and I'm very excited about this, I think, You know, potentially has a very interesting implications.
08:56:56 So, you know, the second order for me acceleration.
08:57:01 I'm, you know, I'm going to talk about it and this limits, where, you know the scattering is mean free path is very short so the fluid approximation is good.
08:57:13 And this is relevant to what we care about, right. So, we can write all simulations in this, and you know here what Chad did is he just filled a box with with cosmic rays.
08:57:26 And he started, we you know with the same string algorithm.
08:57:31 And, you know, watched what happened to the cosmic ray energy density.
08:57:37 And look, it rises exponentially on an eddy turn over time.
08:57:43 There's some hints of the saturation here. And, you know, for aficionados it actually blows past eco petition with kinetic energy density and appears to approach saturation maybe we've got to run this longer.
08:57:57 When the cosmic repression is comparable to the gas pressure.
08:58:01 So, so this is turbulence accelerating cosmic rays.
08:58:09 And, you know, we were very excited about this.
08:58:13 As with all things, it turns out the Russians were there first, some brilliant but very hard to digest analytic paper.
08:58:25 You know he to skin in 1988 derived this curve.
08:58:31 This is why chat simulations look like.
08:58:35 And what this shows is the, this is the cosmic ray diffusion coefficient normalize by starring parameters. And this is the time scale on which the cosmic ray energy density grows normalize to the at turn over time.
08:58:52 And basically, you know the fastest growth happens when the cosmic rate of diffusion coefficient is of order the turbulent diffuse activity which maybe coincidentally maybe not, is, is close to a number that people have been pushing especially those in
08:59:12 the fire collaboration.
08:59:15 You know definitely have thoughts about that but no time to go into it. The other thing is that this acceleration time skill is short.
08:59:23 So, uh, so this has very interesting implications. If you know if this holds up. It means that, you know, you have to consider what turbulence does your cosmic rays it could change the costume or a profile.
08:59:37 It changes when solutions.
08:59:40 And, you know, cosmic rays can be like a phoenix right you can revive them with turbulence. And, you know, we, the reason why we think saturation happens here is because this affects compatibility.
08:59:55 Right. You know when you, when you, when you increase the pressure the effective Mach number of your flow drops. And then the acceleration stops becoming effective, but, you know, this potentially is an interesting way of saying that there's a floor to
09:00:10 the cosmic ray pressure and applies to a number of interesting problems. For instance, in the, in the, in the lab data and the Fermi lack data.
09:00:23 There's this problem called the cosmic ray gradient problem where the profile of cosmic rays in the Milky Way. The clients more.
09:00:33 Very slowly away from sources, which we think are supernova and the disk.
09:00:37 Right, so I think this is still an outstanding problem as far as I know, so this could be an interesting way to solve it, you know if you have some process.
09:00:46 In the Halo, which basically injects energy into cosmic race.
09:00:56 Okay.
09:00:55 Now, I'm going to take three extra minutes right it's nine o'clock, but just a, you know, Cameron is a is looking asking that be okay very quickly. What happens to cosmic rays and sound waves.
09:01:07 Okay, so you have a to fluid saying okay you know we're not going to be able to talk about all this.
09:01:12 I'm very sorry Nevin.
09:01:18 But, you know, really, what happens is that, and this was something that these and you might recognize these people as well, predicted, which is that you will get an unstable Soundwave the cosmic rays will drive energy into the sound waves under the certain
09:01:30 conditions you have assemble harmonic oscillator, which is being driven by, you know driven and so it increases in amplitude.
09:01:41 And as you might expect eventually it's going to go nonlinear, and you're going to develop shocks.
09:01:48 Right.
09:01:49 And here is Naveen similar Naveen soon is a is a is a grad student at UCSB who's done really really nice work on this. And he found this crazy structure.
09:02:03 This is a staircase right there's some other Kitt workshop right now right like staircase 21.
09:02:10 But, um, and the other interesting thing is it's highly dynamic. And also, that there's strong fluctuations in gas pressure. Right.
09:02:24 Um, so, what is going on. So, we think we understand it is due to this thing called the bottleneck effect. Um, you know and and Ellen I think it's going to talk about this on Thursday, so I'm afraid I will have to skip this there's some cool things like
09:02:39 this distribution of the of the stair heights, you know, and the width of these tattoos, the array of pressure like distribution.
09:02:52 And you can you can understand all these things and, you know, it's very pretty. The main thing that you should care about you know is that you know that the time, you know in the end we think the time of average momentum energy transfer, are not different,
09:03:08 but the whole process is extremely dynamic. That's very short time scale transition density and velocity fluctuations right you get all the unit t density fluctuations.
09:03:21 And so, you know, it's something that's interesting, potentially visible with Mr B's.
09:03:28 And the other thing is that, you know, given these short these very rapid pressure fluctuations, assuming a fixed pressure is a bad approximation I mean look at this gas pressure.
09:03:40 Right. You do Fortran ization modeling on something like this is not a good idea. Of course you're going to ask what happens in hierarchy it's still there in 2d at least, And we're going to keep exploring this.
09:03:52 It's, it's kind of fun. So all right, Cameron, I'm sorry went over but I'm done.
09:03:59 Thank you very much paying excellent presentation.
09:04:04 There's been some really good discussion in the slack, as well as a few questions in the in the chat, but to give everyone a break for a couple of minutes and allow everybody to catch up on the discussion that's taken place in the slack.
09:04:20 Let's reconvene in five minutes at 910.
09:04:25 So people can grab a cup of coffee or use the bathroom or whatever, and we'll start with the panel. If you have questions or comments, go ahead and put them in the slack and, and I'll try and get to them as I start to reach out for moderating the discussion
09:04:40 that will take place and panelists, get ready to, to, to join us. Okay, so reconvene in five minutes 910.