08:05:10 Okay, good morning, good afternoon, Good evening.
08:05:15 I'm very happy to welcome everybody back to Thursday, have a week to multi phase week.
08:05:23 If you have been on our calls before you know that we have now merged with technology and we are using multiple platforms to engage with each other fully, and that includes simultaneously operating slack and zoom together and concert and beautiful harmony
08:05:47 so if you have not already, please do open up your slack window maybe on your phone while you're watching your computer screen maybe you've got two screens three screens.
08:05:59 Do it all but make sure you join the channel Halo 21 week two multi phase, type your questions and comments in there, thread them, so that we can monitor that channel, while we're engaging in a rich discussion that will immediately follow the keynote
08:06:16 presentation of today which I will introduce in just a moment.
08:07:11 meta. So Cameron Hummels updated uploaded a wonderful video on how to make a video. So if you have questions about that, you can go ahead and watch that video and then make your own video.
08:07:14 So yeah, don't, don't ignore the new results, where they're getting rave reviews from everybody who watches them so be prepared to have your mind blown.
08:07:25 Okay. And then finally I'm very happy to introduce today's speaker Dr Drummond fielding, who will speak for the next 15 minutes or so as soon as I'm done.
08:07:39 We'll take a five minute break. After that, where you can furiously type your questions and comments into the Slack channel, Halo 21 week two multi-phase, and we will then have a panel of panel, guiding our discussion where I will kind of moderate by
08:07:56 monitoring the slack feeding, some of the most interesting up voted questions that happen on that Slack channel to dr max rocky Dr. Evan Schneider dr project Sharma and Dr.
08:08:07 Jonathan Stern, all of whom will have great insight into the question of the week.
08:08:15 And just to give Drummond a little bit more of an introduction. He received his PhD from UC Berkeley and 2018. I think he is famous for his beautiful visualizations, one of which is featured here this plasma.
08:08:31 love it. I'm gonna make a T shirt with this on it, and his most recent paper, and 2020 is the multi phase gas and the fractal nature of radiative turbulent mixing layers.
08:08:43 And I really hope he shows some visualizations from that paper because it's an absolutely stunning piece of work. And so without further ado, I'm going to stop sharing and German will take it away.
08:08:57 Thank you very much, Jess. All right, I think you guys are all seeing this I'm going to start sharing.
08:09:04 I will say that's not my most recent paper I wrote one since then. But, which I won't talk about I'll talk a lot about the turbulent mixing layer stuff.
08:09:14 So first I guess I'll just say I'm really excited to be here and I was very excited to be asked to give this keynote which at first I thought was a really great honor.
08:09:25 Um, and then I realized it in many ways it was a curse because I'm the first year is giving a keynote and there's just so much I wanted to cover I didn't know what level to hit so we'll see, we'll see how this goes.
08:09:38 But I'm, I'm really excited to talk about sort of the way I think about the multi-phase CGM.
08:10:01 Yeah. And so I guess I'll say, one, one other thing I've been a little distracted lately putting this together as I think many of us have and it's really exciting to be distracted by good news for once And for me, not only did we have the inauguration
08:10:03 yesterday which was so exciting but I. A few days ago I proposed to my girlfriend, and we are now engaged. So I'm very excited to be distracted by that as well.
08:10:17 So I look forward to discussing that more over the coming weeks it's, it's, it's really fun although has taken my mind off the multi-phase CGM a little bit.
08:10:29 So I'll see what I can do.
08:10:31 Thank you for all the emojis and clapping and everything.
08:10:35 Okay, so here's the question right Why are gases halos often multi-phase well let's just get on the same page to begin with, um,
08:10:49 here's a little cartoon okay this is I just want to give a conceptual picture as a background and then we'll build on this so here's a nice little galaxy.
08:10:58 And there's this you know hot Halo surrounding it and this hot Halo intentionally is not perfectly homogeneous so there can be fluctuations and things within the hot Halo itself.
08:11:10 And we're here to talk about the multi phase CGM today right so let's add some multi phase there could be many cold discrete clumps that sort of follow a by conical outflow and maybe some streams or maybe the distinction between the phases isn't so clear
08:11:28 and it's a little more blurry like this, and or maybe the multi phase is in the form of big streams that are penetrating the hot Halo.
08:11:37 Or maybe I was playing with my iPad too much and there's worldly outflows, but whatever let's just, let's pick one okay so here here will be our multi phase CGM okay there's you know think by conical outflow maybe some streams, whatever, don't read too
08:11:53 much into it.
08:11:56 So what what do we have here in the CGM in particular, well, there's the volume filling phase okay the volume feeling hot phase that is roughly at the very low temperature or maybe as as Mark taught us in the first tutorial, there should be more like
08:12:12 the gravitational temperature.
08:12:14 And this is roughly a million Kelvin or so and this is of course a halo mass and redshift dependent statement, right. So, lower mass halos are going to have lower temperatures and therefore, you know the cooling will do will differ same higher Halo masses
08:12:32 will lead to much hotter Halo so something to keep in mind but let's just keep this as a baseline number.
08:12:39 Then of course, as I've drawn we have the dense phase and this dense phase. I'm going to call it cold, although maybe just some people who study the ISS.
08:12:49 This is important to call 10,000 Kelvin cold.
08:12:52 But when I want to talk about lower temperatures I'll call it maybe molecular atomic but.
08:12:57 And this this cold temperatures set by roughly being in thermal equilibrium with the UV background. Okay. And so that gives you a temperature of around 10,000 Kelvin and of course self shielding can drive you to lower temperature it can allow you to get
08:13:13 to lower temperatures.
08:13:15 Great. So those are the the two phases but of course this is a multi phase CGM so this is really you know there is a continuum of phases between here, temperatures between here and I want to highlight what all what I'll call the intermediate phase which
08:13:29 is sort of in between these two temperatures and the reason why this is so crucial in particular, is because at a temperature of around 10 100,000 Kelvin.
08:13:40 This is where radiative cooling becomes incredibly important.
08:13:45 And therefore, where a lot of the most exciting and and key physical processes take place.
08:13:52 Okay, so I now I want to outline what I view as the four primary sources of cold gas in the CGM.
08:14:05 So first, and nearest and dearest to my heart our galactic winds so you can have a galactic when directly launch cold and of course and hot gas out into the circle galactic medium.
08:14:20 Then you can also have satellite galaxies inside of the galaxies Of course they have their own galactic wins, and they can be stripped and therefore lead to code CGM material.
08:14:32 On top of that, you can consider accretion from the intergalactic medium.
08:14:37 And then lastly, we'll talk about the thermal instabilities in the halo itself and. And now for this third or the fourth point, I'm not going to get into that much detail on this.
08:14:51 We just heard a lot about this yesterday from critique Sharma there's already been a really rich discussion on this and Mark Voigt set up a little after party for this talk to talk about thermal instabilities even more.
08:15:05 I will of course talk about this, I'll say off the bat that these are not distinct groups, these are not distinct processes, they're all interconnected but let me go through them one at a time and then we'll talk about their interplay.
08:15:18 Okay so, Todd Thompson showed our Todd trip sorry Todd. Todd trip showed the observers favorite galactic wind system the you know what he called it the exploding cigar galaxy right well I think this is now becoming the theorists favorite example of a
08:15:39 galactic when this is just this absolutely stunning volume rendering of one of Evan Snyder's recent seagulls simulations, showing the density volume rendering flowing out of one of these galaxies.
08:15:55 And I'll talk more about this as I go I think it's a really useful touchstone, and I want to ask two primary questions of galactic wins. First, how much code gas is carried by the wins.
08:16:18 winds and then what happens to the code gas as the winds evolve. So let's start in the actual interstellar medium I'm going to show some, you know, just called me I'm good at visualizations maybe but the visualizations I'm going to show today are just
08:16:30 out of this world by by many other people. So here's a great one bye chain guru Kim he actually custom tailored this one for me.
08:16:39 And this is showing a small patch of the interstellar medium where they're actually resolving the star formation and supernova remnant processes within the interstellar medium, and then looking at the self consistently generated multi phase outflows that
08:16:55 are flying out of this, this patch of the ISS. And you see all kinds of really rich dynamics in one, you know thing incredibly important for us is what you can see right here is you have a volume feeling hot phase, and a cold, you know, tend to the fourth
08:17:16 Kelvin outflow co moving or nearly co moving with that hot phase and so looking at this is really useful for sort of informing what we expect the phase structure to be at the base of the winds.
08:17:32 And here I'm showing a way to look at this that I that I really like this is from chain goose recent papers on this.
08:17:40 And this shows the sound speed versus outflow velocity phase distribution.
08:17:47 So, the color shows how much the mass flux. In this phase distribution and, hot, hot stuff is up top, and fast stuff is to the right. And so you can see there is two, there are two main sort of pockets there's this hot fast moving phase right that's this
08:18:06 red stuff going really fast. And then there's this cool, you know, 10,000 Kelvin stuff that's that's moving you know maybe 30 kilometers per second or so.
08:18:19 So, if that's the baseline. That's telling us something very interesting because if you're only going roughly 30 kilometers per second, then the cold phase is never going to make it out into the circle galactic medium and your, your most of the wind that
08:18:34 the CGM sees will just be hot.
08:18:37 And that may well be the case and you would just have a sort of fountain flow of this stuff on you know the order of five to 10 k PC depending on your halo potential, and your gravitational potential from the galaxy.
08:18:49 But there is also another possibility that this cold stuff becomes accelerated by this hot stuff and they become roughly co moving and, you know, I may be striking a little bit too low of a level looking at the people who have their cameras on right now,
08:19:04 I mean everything I'm saying is really obvious so just take this as sort of baseline framework for the discussion that will have afterwards and I just want to briefly flash up some results from a student I've been working with barrage Pandya.
08:19:20 This is work that will be published any day now doing a very similar set of analyses on the fire simulation showing the same phase distribution in here we've shaded out the region.
08:19:33 That won't be able to make it out into the CGM and so you can see all of this cold gas in this wind is really in the shaded region where it won't have enough specific energy to make it out into the Halo.
08:19:48 And this is measured you know right at the the desk Halo interface. So let's now think about this a little more and I'll take some work from this, this great paper I think a really seminal paper by dawn Jang called entrainment in trouble.
08:20:02 And it's it's the question of how do you take a cold clump and accelerate it by a hot wind.
08:20:09 So if this code cloud has some density row cloud and the size our cloud and it's got a mass that's given by this the density times the volume.
08:20:18 And then you can ask, When does it become accelerated by this hot wind and the answer is when it is interacted with as much mass in the hot phase as it has itself.
08:20:31 And so you can sketch out this little cylinder which you know the density is the hot phase density the area is the, the cross sectional area is the same as the cloud and then the length is given by the velocity times this characteristic acceleration timescale.
08:20:47 So we solve for t Excel or other people call it T drag or whatever you want. And you get this very nice expression here that the timescale on which the cold cloud is expected to become co moving with the hot face whoops I made the mistake of looking at
08:21:06 this lack I'm going to ignore that for now.
08:21:09 So you can you can look at this and say, Okay, here's the acceleration time scale and it's given by the ratio of the densities, times the size of the cloud divided by the velocity, and I'll call this density ratio, chi the density contrast.
08:21:25 So this seems good.
08:21:28 However, as, as we've known for, you know, 30 plus years now.
08:21:32 This is not all that happens. Here's the simulation by Matthew Abruzzi Oh another student I've been working with showing this cloud crushing problem inaction.
08:21:42 And what you see is very rapidly the cloud is entirely destroyed and Matthew is about the millionth person who has studied this so I'm not going to even put the full list of simulations down here.
08:21:56 But the point is, is that the destruction timescale is roughly the cloud crushing timescale, which is shorter by a factor of chi to the one half, then the acceleration timescale.
08:22:12 So, here is our first big conundrum. Okay, what is this say well this says right that the destruction time is faster than the acceleration time so then you don't get any entrainment any code seeds in the wind will be shredded and mixed within just a few
08:22:28 killer parsecs you know if you cloud crushing times. And so you don't expect galactic winds to be a predominant source of code phase in the CGM.
08:22:40 However, I have this great picture of paying oh and Max cranky from 2018, when they were thinking about this.
08:22:49 And what they realized at this, this moment of insight.
08:22:54 You know this is like I believe this is actually exactly when they thought about it.
08:23:00 They realized that the, if the mixed gas cools more rapidly than the cloud is dispersed, then the cloud can survive.
08:23:09 And this is an incredibly simple and very compelling idea that to my mind has revolutionized the way we think about multi-phase gas and this, I mean, if you want to take one thing away This is it.
08:23:23 This is the important thing. Okay.
08:23:24 And really, you can just say it very simply mathematically, it's if the cooling time of the mixed phase, which is likely going to be the most rapidly cooling phase there is, is shorter than the destruction timescale, then your cloud will survive.
08:23:40 And they flip this around to say, if your cloud is bigger than this characteristic cloud size, then it can survive.
08:23:49 And so, I mean, this is fantastic and let's let's see one of these in detail. This is again from Matthew Abruzzi Oh, paper that should be out any day now and this is an otherwise identical simulation but now cooling is turned on, and it's more rapid than
08:24:20 And what happens instead of being dispersed right in the other simulation the cloud was gone by now, it's actually growing, and you're building this long structure of in the frame is being is following the cloud. So you're building this long cloud, that's much more massive than what you started out with. And you
08:24:26 you started out with. And you can see there's a huge list of people I think probably almost all of them are in attendance today, who have been studying this.
08:24:36 And, and, and to my mind this is just beautiful super exciting work. And I've been working on this problem myself and trying to really understand the mechanics of how this problem works by zooming in on just the turbulent radiative mixing layers or TRML
08:24:57 or terminal, or turmoil.
08:25:03 I mean, I'm really I'm pushing this this is a Greg Brian came up with one of the all time great dad jokes, saying that this was taming the turmoil, I love it so this is, this is the way I like to think about it.
08:25:14 So we've been zooming in on this on these turmoils and and looking at it. So here I'll just show a brief visualization of this showing the hot phase on top of the cold phase on the bottom.
08:25:29 And what you have is that the sheer flow leads to turbulence. That turbulence leads to mixing right mixing of the two different phases the mixing of the phases populates this intermediate temperature regime which cools extremely rapidly, and that cooling
08:25:46 then leads to an inflow you know this this blue color up here is an inflow from above, into the cold phase and so the cold phase is growing.
08:25:57 Due to this, cooling, which is induced by the mixing.
08:26:01 Okay so again it's it's it's a very simple picture exactly in line with the original picture by by Max and Tang, and we were able to come up with a nice simple description of the growth rate of the cold phase right this info velocity is how fast the cold
08:26:17 cold phase is growing, and its parameters just by you know the densities the size of your object, the turbulent velocities and the cooling time with this sort of strange one quarter power.
08:26:30 and a few months after I came out with that.
08:26:34 Brent tan wrote a very exciting paper drawing on the turbulent combustion literature right so turbulent combustion is very similar to turbulent cooling, just with a minus sign.
08:26:47 And so they had the great insight to look at this and in particular they looked at.
08:26:53 This is another little, I don't know the answer is in don't mix or Dom color mix number, whatever. Okay, so they were able to extend this to the full range of cooling rage and show that the the sort of power law behavior of the growth rate changes when
08:27:10 this this crucial number which is roughly given by this crosses above and below one.
08:27:17 So this is all really exciting you know in the pay this has just been really two years since people have been thinking about this and we've made really incredible strides I think and it's it is made a very compelling picture where you have this galactic
08:27:31 wind and you put these little seeds right like what chain goo, like we see in chain guru simulations. And those seeds can then grow as they are accelerated and then trained by this hot wind and can be a very important source of code gas in the CGM.
08:27:51 However, it's not quite so simple.
08:27:53 So, so here is what I'd say this is the figure is this one on the left, I've spent so much time looking at this figure, and I love it and this is again from that same simulation by Evan showing the temperature in slice.
08:28:11 And if you focus on this orange and blue these orange and blue lines here This shows the mass flux in this when as a function of radius wear orange is the cold Phase I flicked the color so Orange is the purple stuff, and blue is the orange stuff that's
08:28:26 maybe a little confusing, but bear with me so what what you see here is that beyond about to kill parsecs the code phase is actually losing mass to the hot phase.
08:28:39 Now it's not entirely disappearing like you would expect for just you know the no cooling case, and there is still cold gas flying out at you know 810 kilowatts x from the galaxy, but this is a really important point that I think is something act I'm
08:28:56 actively working on I know Max is working on this lots of people and I think this is really exciting. And it's the dichotomy between the intrinsic turbulence in the mixing layer right the sheer flow right on the skin of these clouds leads system turbulence,
08:29:11 but then you can also have turbulence in the background hot phase that can destroy the clouds and so.
08:29:17 So, this simple picture is a little more complicated although I think we're getting quite close to having all of the parts in place.
08:29:27 Okay, so that's all I want to say for galactic wins, and I'm going to move on now to talk about satellite galaxies. Okay. Um, so, Talk about a visualization.
08:29:43 I mean, boy, this is, this is from tangi 50 I think I assume doing made this I just found this on their website this is boy.
08:29:52 I love this you I could look at this forever I could really just give my whole talk looking at this, this is a visualization of a protein cluster forming.
08:30:00 And you can see what I'm really interested now, are these little satellite galaxies with these beautiful clumpy streams behind them all over the place, the more you look, the more you see them.
08:30:13 Okay. And so, the, this is an example of how ram pressure stripping is really important for fueling the cold phase in the CGM.
08:30:25 And in many ways, this is the exact same process as the cloud crushing process here's a simulation from Stephanie Thomas and that will be published shortly, showing a ram pressure stripped tail.
08:30:37 And if those tails can cool faster than their dispersed than just like clouds and wind, the tail can grow.
08:30:46 And now this is actually complicated, and you know maybe more exciting because you also now have to think about self gravity and potentially star formation in these tales So, so, you know, a lot of the baseline physics is the same although it is undoubtedly
08:31:03 more complicated, so ramparts are stripping though is not the only source of code gas from satellite galaxies. You can also look at just the winds from the satellites themselves.
08:31:16 This is a. I think a really beautiful beautiful paper BY ZACH Haith and last year, two years ago. Looking at the source of CGM material in the fire simulations, and what they show in brown here and these two different panels for two different times for
08:31:33 the same Halo is the satellite when the contribution to the CGM, and it's actually comparable to the actual wind contribution from the host galaxy.
08:31:46 So this shows that you know satellite winds themselves can be extremely important for populating the CGM of a central galaxy. And on top of that. There's also this blue stuff which is the direct IBM accretion and they can all you know it.
08:32:07 I advise you to look at tax paper for the details but they can be. Each of these can be very important contributions each of these three channels.
08:32:13 And so this leads me nicely to my next topic which is exactly that I GM accretion.
08:32:22 So I'll again turn to Dylan Nelson for an incredible visualization This is now five six years old from the original illustrious showing the density and temperature of a halo as you can see these cold dense streams are penetrating this hot Halo.
08:32:40 And there's also you know smooth accretion from beyond, and you get this, I don't know this really rich system there's of course satellites and streams and everything, there's a very clear accretion shock as well.
08:32:52 So there's, you know, there's all kinds of interesting dynamics and it's it's it's worth thinking about the outside of the hill so not just how gas gets out, or how gas is added in situ but also how gas comes in from from beyond.
08:33:08 And so, I want to ask you the question or rather near Mandelker asked the question in a really seminal series of papers with a great group of collaborators, as what happens as these filaments enter into the hot Halo.
08:33:25 And here, imagine this is a filament and this is a visualization of the density. So red and white is very high density black is very low and as this filament enters the Halo.
08:33:38 It mixes and is shredded into the background, medium, and I'm sure many of you already know where I'm going with this I'm sort of repeating myself there's this.
08:33:47 This process of shredding versus cooling and if cooling wins then then things can be stabilized in the cold phase can grow. So here's an otherwise identical simulation that near did just recently with strong cooling.
08:34:02 And you can see no longer is the cloud is the the film in shredded and disrupted and said, I mean it still becomes very turbulent but it remains very dense.
08:34:12 And so this is an incredibly attractive way of getting cold gas into the CGM you know directly from the IBM on its way into the Halo, and there's important observational differences that one could imagine from this you know this, IBM accretion would be
08:34:31 cold and dense but it would likely be lower medalist city then say a galactic wind which has been recently in rich So, so I think there's a lot of observational characteristics that might be different from these different pathways that I'd like to discuss
08:34:46 further in the discussion.
08:34:50 Okay, so I would, I would be remiss if I didn't also talk about building the hot Halo via IgM accretion right. So, as gas flows into the CGM they can come in and in filaments but it can also come in in a, in a smoother way and sort of interact with pre
08:35:09 existing material. And here's a set of simulations that I ran years and years ago now it's actually from this is a slide from a conference where I met many of you for years ago my very first conference in Italy.
08:35:26 I'm looking at Joe and Jess and Ben and a bunch of people who, who I hung out with there I think that's where I met Jonathan has become one of my, my closest collaborators.
08:35:35 So, so these simulations are, you know, fairly, fairly old at this point and and and they are, I would say, a rehashing of very classic results dating back to the 70s and 80s really crystallized by.
08:35:51 You've all in art and Avi shy in the early 2000s, you know, do Sean and Neil did great work on this with, you know, this is a rich topic so I'll just say, in short, as the gas comes in, accretion shocks against the existing material, if that accretion
08:36:10 shock he did gas cool slowly.
08:36:14 Right, and slow relative to the freefall time which is roughly the compression time and so that the heating time. And if it if it's longer than that then the hot outer CGM grows and there'll be all kinds of interesting dynamics, you can see clear signs
08:36:29 of thermal instability and feedback in the center of this halo which we'll talk about next.
08:36:34 But it's important to realize that this is not always the case. Right. This is where the real temperatures a million Kelvin. But what if the bureau temperature is 100,000 Kelvin.
08:36:46 Right. And this is a 10 times lower Halo mass, where the accretion Shockey the gas cools very rapidly. Then you have an incredibly different CGM where it's, you know, feedback and turbulence really rule the day and you have something that that looks looks
08:37:01 very different in the whole system will be out of pressure equilibrium they'll be sort of global thermal instabilities as opposed to local thermal instabilities and.
08:37:12 Well, let's let's actually get into that right now okay so let's let's get into thermal instability so check my time all right I got 17 minutes by my clock is going to be a little tighter than I thought.
08:37:25 I'm like, not halfway through so it'll be a little tighter but I promise I won't go over okay so thermal instabilities right.
08:37:37 Mark promised everyone that I would talk about this so I better talk about this. I want to make a distinction which is actually maybe less of a distinction than it should be that it's important to consider what the background is that you're talking about
08:37:47 that is terminally unstable. Okay, so you can have a background, that is in what's called a cooling flow and click flow, to my mind should be thought of as the basic property of the CGM you know it's accretion Shockey did from out beyond, and then it's
08:38:05 really just sitting there. And what is it going to do it's going to gradually cool in flow in and that's a cooling flow, okay and it's where your losses dominate your heating terms.
08:38:19 On the other extreme, you can have a halo where there is complete thermal balance where the cooling losses are balanced by your heating right so this is where you know feedback is able to do a perfect job and balance your cooling losses.
08:38:34 And now I want to be. There's been a lot of discussion about this stuff so I want to be really clear to say, I am talking about I'm showing here the two extremes, I do not think the universe lives at one of the extremes, there is a continuum, you know,
08:38:49 it might be that the losses are close to the heating, but not exactly balanced and so you've got some bit of a cooling flow and some bit of a you know a thermal balance and so I think it's important to, to keep that distinction in mind when I talk about
08:39:04 this and that this is really a continuum.
08:39:07 So let me just explain a cooling flow for those who don't think about it all the time it's quite a simple idea, which is just that the compression of gas balances it's cooling so you have a parcel of gas it's out in the CGM it cools it loses some entropy,
08:39:26 and it falls in as it does it's compressed, and it's heated in it. So what ends up happening is, it stays at almost exactly the same temperature as it flows in and as the entropy is decreasing.
08:39:40 I recommend people look at Jonathan's recent series of papers on this if you're interested in some very intuitive nice you know like three lines to understand these.
08:39:51 And there's one crucial property of a cooling flow which is that the cooling time is equal to the flow time. Okay, and by flow time I just mean, you know, our Overby you know how it.
08:40:01 Yeah, so your cooling time is balanced by the flow time.
08:40:05 And if you up the cooling, you up the info.
08:40:10 And you can keep doing that there's different ways you can up the cooling you can increase the density, you can, you know, decrease the overall temperature because you're in a lower mass Halo right.
08:40:20 Whatever you do, if you increase the cooling you're going to up the inflows and that's fine nothing changes, until. Right. Well, eventually as you do this, right your sound speed stays the same.
08:40:33 But if your velocity is going up, then Something's got to give because eventually you'll become supersonic.
08:40:41 And there's a bunch of really classic results on this dating back to.
08:40:46 Ball busting soaker and 89 and probably people before then, which showed that cooling flows become thoroughly unstable, when they are supersonic in this supersonic transition because of this occurs when your cooling time is less than your free fall time.
08:41:04 So, you know, this is, I think, a very common situation and really what you can think of. It's just, you're now, once you're supersonic, you're out of Sonic contact with your neighbors so perturbations can sort of run away and cool faster than its communicated
08:41:20 to their neighbors.
08:41:22 So, this is the picture of a supersonic cooling flow or really the transition from sub Sonic to Tran Sonic supersonic. And this is something Jonathan and I had been looking at in a bunch of different guises here's a recent paper that he included Andre
08:41:38 led looking at the fire simulations. This is a visualization of the temperature field and inner CGM at fairly closely separated in time, and the main difference in these halos what happens over this time.
08:41:53 Is it transitions from a supersonic on the left to a sub sonic boom flow on the right. So there's going to be lots of discussion about this in the after party marks after party on thermal instability so I suggest me.
08:42:08 You know talk more about this there but you can clearly see that this is a thermal a stable medium that's almost uniform temperature and this is a thermal unstable medium that has, you know, huge temperature variations.
08:42:21 Okay, so now there is this whole other branch of thermal instability which is actually I think really the hot topic and this is what we heard so much about from critique Sharma yesterday.
08:42:34 And so I'm not going to get into too much detail on this now because we did just discuss it, but I want to show the basic principle and also just reiterate that these two processes these two ways of looking at it or not, incompatible with each other that
08:42:52 they might be happening at different times in different places or some continuum of the two. And so, here, here is the other end of the spectrum where you have thermal balance and these are visualizations done by arena both ski, and a paper that we put
08:43:07 out last year, looking at how thermal instability changes with in the presence of cosmic rays. This is actually, since I'm not touching on non thermal pressure at all this week I'm, I'm showing the hydrogen AMEC only simulations.
08:43:24 And what we've got here is a hydrostatic stratified medium where at every height you balance heating with cooling, and you put in little perturbations, and the one on the left has very rapid cooling, and the one on the right has very slow cooling.
08:43:42 And they're going to evolve, very differently.
08:43:46 So the one on the left cools so rapidly that it doesn't even fall down. Prior to, you know, cooling out, and the ones on the right. They don't even cool at all.
08:43:57 And there's one very important point I should make about these simulations is, these are run for the same number of cooling times.
08:44:07 So, these simulations have very different freefall time so this simulation was run for 100 times longer than this simulation that's why you know this stuff hasn't even fallen down so this, the one on the left really has way, way more cold gas and this
08:44:23 is exactly the process that critique was talking about yesterday that Mike McCourt showed in his seminal 2012 paper followed up by critique seminal spherical paper you know and spherical coordinates there's, there's so much work on this I'm going to miss
08:44:38 references on almost scared to do this but you only did great work on this with Greg Brian switching g looked at it and MHD arena and I did this in cosmic rays building on some really, Really beautiful work theoretical work by Philip Kamsky.
08:44:55 So there's there's lots going on here. And, you know, I think we should discuss this more, coming, coming up, and I, that sort of all I want to do now for my, my overview of course my overview is taken most of my time and I want to stress before I move
08:45:13 on that.
08:45:15 These are not happening in isolation, you know, cosmological accretion satellite galaxies, they trigger thermal instability is right large mode density perturbations like creating Paul Chowdhry showed trigger thermal instabilities more easily just like
08:45:30 what, you know, I think Clark is Marion and Dylan Nelson are seeing in their cosmological simulations and those thermo instabilities then feed cold gas into the galaxy which then drive galactic winds so this is a rich interconnected system right and so
08:45:47 there's, there's a lot going on here and there's probably different observational characteristics of all of them that we should we should think about.
08:45:56 And so now that I've done that. Overview I want to sort of get a little more meta right and think about what is this question you know why are gaseous halos often multi phase.
08:46:08 And I'll give you my answer which is a, you know, it's sort of like you spin the Wheel of Fortune, and whatever you land on you can get multi-phase gas okay so you spin the Wheel of Fortune website and I do.
08:46:23 And you can get multi-phase gas because the cooling time is less than the cloud crushing time like Max and pen showed, or you can get multi-phase gas because the cooling time is less than the ram pressure stripping time like Stephanie Thomason show or
08:46:38 the Kelvin Helmholtz time, like, you know a bunch of us have been showing like near showed and Brent showed and, and I've been looking at. You can also get it if the cooling time is faster than the compression time like, you know, you've all and obviously
08:46:53 I showed us in 2003 or even you know back.
08:46:57 50 years from now, or whatever 50 years back when they first started thinking about this.
08:47:04 Cool. Okay, so you can also get it if the grueling time is less than the freefall time, or you know maybe 10 times at freefall time and, you know, to just sort of synthesize it the way I like to think about it it's just the mixing time.
08:47:18 All of these things are in a sense, the timescale on which you're mixed back away from multi phase you know if you take something that's in homogeneous these are sort of the time scales on what you become homogeneous.
08:47:36 You know your cloud crushing it's mixed back in your free fall time its buoyancy waves and, you know, whatever it is, it's effectively a mixing time so this is maybe a little too imprecise to really be useful and so I'm not trying to say this like quantitatively
08:47:51 because I do think the details really matter. But on a zero authority level this is my answer right because your cooling time can be less, and it's oftentimes less than, whatever the relevant mixing time you're looking at.
08:48:07 So I've got six minutes left by my clock and there's a few questions I want to address. Okay. So, this is almost a command performance because, oh, Jess is giving me 10 minutes left.
08:48:22 Okay, great. Well then, I'll, I'll slow down. Um, so cool so are the phases in pressure equilibrium is a question that's been coming up a lot in the observations that mark told everyone that was going to talk about so I better talk about it.
08:48:36 So, so let's let's let's get let's get into it a little bit and, again, this might be a little too simplistic to be super useful but I think it's a good, it's the way I think about it and I'm the keynote speaker so you're going to hear about it.
08:48:50 So, The answer is yes. Unless, okay.
08:48:55 And that unless is there's really three things. Although, actually there's four, but, so unless they're moving too fast so you should be in pressure equilibrium unless your flow time is shorter than your sound crossing time.
08:49:10 I'll get into each of these afterwards so you don't have to dig in too much now. Also they can be out of pressure equilibrium if they're cooling too fast so if your cooling time is less than your sound crossing time.
08:49:23 Or, if you don't have enough resolution so if your cooling time is less than sort of the sound crossing time of one of your cells in a numerical simulation.
08:49:33 And then of course you know there. I'm just completely punting on this I have tons to say on this but non thermal pressure is awesome arena booties paper on this is a really nice look at this I think in a controlled way with some nice models, but everyone
08:49:48 here really I think is chomping at the bit to discuss this so here's a, you know, setting up the discussion for next week.
08:49:55 So let's talk about these three and how these take you out of pressure equilibrium.
08:50:02 So first, let's consider supersonic motion Imagine you have some background hot phase that you're considering, maybe this is a wedge of the CGM or some out flowing galactic wind, and it has some pressure scale height h over which the pressure changes
08:50:21 Ah, over which the pressure changes appreciably.
08:50:24 Well, in this case, imagine you have a filament like near Mandelker considered. You could also consider a cold cloud flying out in the other direction.
08:50:32 And this filament is coming in at some velocity v.
08:50:46 And it's got some, you know, sound speed and some, some size are, then you are likely to be out of pressure equilibrium. If you traverse the scale height faster than pressure can communicate across the, the thickness of the filament right so this is,
08:50:59 you know, a very similar. This is not a.
08:51:03 I think a contentious statement right it's if your flow time is shorter than your cooling or then the sound crossing time then you are likely to be out of pressure equilibrium.
08:51:14 But this is not going to happen for a really long time. No, this is not likely to be long live and I think this is a point I want to stress supersonic motion doesn't go on for a long time sort of by the very nature of being supersonic it wants to make
08:51:29 shocks and slow things down and so so you're likely to slow down and not be out of pressure going in for too long.
08:51:37 You can also consider supersonic cooling and this is something that Mike McCourt looked at in his seminal shattering paper a few years ago and also Max and pen looked at really nicely just last year, in a very intriguing I would say letter that I just,
08:51:54 I want to dig into and understand even more and what they did is they start with potato as max calls it and in homogeneous cloud, and they let it cool.
08:52:05 And depending on the size of the cloud. It cools, in an ISO Barrick way, or it cools, in an ISO Couric way, you know, out of pressure equilibrium or in pressure equilibrium, and just focusing on the one that cools, out of pressure equilibrium here that's,
08:52:23 that's this guy.
08:52:26 Here's a visualization of the pressure of the simulation originally it's in perfect pressure balance, and then they let it cool and it goes out of pressure equilibrium, although it then quickly comes back into pressure you go, whoops, and the characteristic
08:52:41 size scale in the characteristic criterion is if you know the cooling time is shorter than the sound crossing time, or if you know another way to say that if the cloud radius is bigger than CST cool you know if the cloud is larger than a Soundwave can
08:52:58 communicate across the cloud, which they call l shatter because it does cool things to your cloud.
08:53:06 So, I'll say the, to my mind, the poster child for being out of pressure equilibrium because I think it's actually one that can be long lived is a supersonic cooling flow.
08:53:17 And the reason why this is the poster child is because it's both cooling too fast, and moving too fast.
08:53:25 Okay and and this is actually, I think a very common situation in the universe at high redshift where the densities are high, you can. You're, you're likely to get these supersonic cooling flows and and you know turbulence and feedback is only going to
08:53:40 make this even more out of pressure equilibrium.
08:53:43 And so the reason why your, your boat you're out of both 30, you know you're both moving too fast and going too fast, is because of you have both of these criterion.
08:53:59 And Jonathan visualize this really nicely in his paper, these are those same slices I showed you earlier.
08:54:01 But now showing the pressure deviation.
08:54:05 And what you can see is in the sub Sonic cooling flow over here in the right, the pressure deviates, you know, order unity, really. But when you go down here, the pressure deviates by, you know, a factor of 100 over relatively small scales.
08:54:22 And so this may be the state of the universe for large swaths of cosmic time and in particular, you know, high redshift observations we need to consider that this is the sort of medium that we're looking at, as opposed to you know this with some cold,
08:54:38 cold droplets in it.
08:54:39 So, so this is a point I think it's worth stressing.
08:54:44 I forget what's next. Oh yeah. Okay, so being out of being out of pressure equilibrium because you're under resolved.
08:54:52 I talked about this at the last KITPV ending so I won't get into too much detail, but I think it's worth noting because so many of us run simulations of the CGM.
08:55:02 If you take one of my turmoil simulations like I'm showing over here on the right the turbulent radiative mixing layer.
08:55:10 And you ask what's the phase distribution.
08:55:14 Right then, here is, here's one way of visualizing it so this is the mass, how its distributed in pressure, high pressure to the right low pressure to the left, vs entropy high entropy of top low entropy and the bottom.
08:55:28 And you see a very nice trend it's pressure equilibrium pressure equilibrium and then a dip to low pressure. Right. and so this is out of pressure equilibrium.
08:55:40 However, if we increase the resolution by a factor of for that dip diminishes.
08:55:45 If we increase by even more that dip vanishes.
08:55:50 Okay, so if you want to talk about, you know, the phase structure of your simulations, you better be resolving things properly and of course these are pure hydro simulations, you know, non thermal pressure can can really add to this picture and maybe
08:56:07 change the resolution requirements, but it's it's worth keeping this picture as sort of the baseline and then adding to it, I think, and we can understand this very simply.
08:56:19 In the color that I've over plotted this is at every point you can ask what's the sound speed, and the cooling time. And so here I'm now plotting how the sound speed times the cooling time changes over this domain.
08:56:34 And I've drawn a little fuchsia line here to highlight the resolution of this simulation and what you can see is in the low resolution simulation it remains isometric at constant pressure until it approaches the resolution limit where it dips down at
08:56:52 intermediate resolutions it dips right alongside the resolution limit where a very very high resolution, you know, the, the resolution limit is way over here so we're, we're totally safe.
08:57:03 Great.
08:57:04 Um, so this is something we want to keep in mind because these, these turmoil simulations, even this low resolution one is impossibly high resolution for a global simulation.
08:57:17 Maybe not impossible, because there's been some really great work in. In recent years, or really in the recent year or two.
08:57:25 Trying to enhance the resolution out in the Halo.
08:57:31 And I, I'm pretty confident they're not getting down to this sort of resolution.
08:57:36 But there are different questions one might want to ask if we're asking pressure equilibrium, you do need to resolve CST cool. I'm confident in that if you're simply asking how much code gases there.
08:57:49 Well that's actually a slightly different question, and you might get away with with resolving just the turbulent mixing which is a little less restrictive of a requirement, although I'd be keen to talk more about this with everyone in the discussion
08:58:06 and over the coming months, are coming. Yeah, I guess so.
08:58:21 In my final say three minutes I just want to give you know my perspective on what the observational tracers of the multi-phase CGM are actually probing.
08:58:23 So first, let's make a multi phase CGM, and we'll do it in my favorite way, which is to make the cooling time shorter than the mixing time. So, this is very simple take a turbulent box.
08:58:38 Stir it up, make the cooling time shorter than the mixing time and you get this, this very nice looking swirling bath that's clearly got, you know, to stable equilibrium, the hot phase and the cold phase and lots of intermediate temperature gas.
08:58:51 So I could stare at this all day. I won't, and instead all over plot, what this looks like in different ions so here is the number density of oxygen six carbon for neutral hydrogen in silicon two.
08:59:08 Okay. And I'll fix time but I'll swipe these through so you can see how the different structures look and I want you to imagine you're putting you know a quasar sightline through this.
08:59:21 Well, let's focus up here. Okay, so what you see here is, first let's go to, here's silicon to Okay, these cold clumps are traced really nicely by silicon to there's a whole bunch of them, you'd have a whole bunch of cool lines probably blended together
08:59:40 if you didn't have high enough spectral resolution, they'd also be traced in h1.
08:59:46 It'd be surrounded by a cocoon of strong carbon four and a little bit of oh six although not sort of a booming oh six signal.
08:59:57 If instead we look over here you can see there's no silicon to just a little bit neutral hydrogen.
09:00:04 Okay, some carbon for in a ton of oh six. So, I don't know if this is really useful but I just kind of wanted to put this up as a as an example of what I imagine in my head when I see your spectra and I see the bumps and Wiggles.
09:00:21 And I think we're getting pretty close to being able to like really understand the physics of multi-phase gas. And so, pretty soon I think we're, or maybe even we're there now it's time to start forward modeling these observations.
09:00:39 And really, really trying to understand them. Okay, so I'm going to stop here. And, you know, here's my beautiful cartoon summarizing everything that the CGM is a rich interplay of these things we've talked about.
09:00:54 Yeah, and I'll stop there.
09:01:00 Thank you. Yeah, let's everybody give driving a round of applause. That was a ton of incredible beautiful visualizations. As expected, did not disappoint, including a lot of really really interesting material for discussion.
09:01:19 So, the halo 21 week to multi-phase Slack channel is already blowing up. If you haven't been over there, read it over, add your own comments and questions.
09:01:29 We've already got, you know, I've written down a few things that we want to talk about with our panel when we come back, take five minutes to do that, let's come back at 9:06am pacific time and five minutes, where will resume with our panel discussion
09:01:44 and questions and head over it, make sure you head over to the slack Thanks everybody, thanks again Drummond.