08:05:06 Good morning everyone. Thank you for joining us today. Tuesday. 08:05:15 For our catch up funnels fundamentals of caches halos. 08:05:20 Let's see. Okay, as we began yesterday this week, focuses on how does gas flow out of galaxies. 08:05:30 It is our fifth week so we just made it past the halfway mark of the program. Hopefully people aren't all burned out, and are still excited for CGM science. 08:05:40 I know I am. 08:05:41 We had a wonderful event last night, social event, some of you I know we're able to attend. It was astronomy on tap, featuring some related public level talks by Jess work and Zach haven. 08:06:00 Last night at 7:30pm Pacific. 08:06:01 And then we, we had pub trivia afterwards, hopefully some of you were able to join and have fun. We will have more social events in the coming weeks. Next week, I believe we're going to do. 08:06:14 It's not using town town, gather but it's something similar. It's kind of a cross between that and zoom it's called room, RUME, and, and it allows you to join a zoom teleconference kind of thing and then have rooms, but you can see who's in the rooms 08:06:31 like see what's going on, but you don't necessarily hear the sound so it's it's experimental but we'll do it is kind of a beer our next week. I think we're talking about next Wednesday. 08:06:44 At 5pm Pacific, but there'll be more details on that and then we're going to definitely do another trivia, because that was so successful. Last week, So stay tuned. 08:06:56 Today, we will be joined by our keynote, our first keynote speaker of the week dr Dillon Nelson, and he'll start in a couple of minutes and go until the hour will have a short break. 08:07:11 Five minutes or so before we have a panel, consisting of Roman board lawyer john Chisholm revealed of a Django Kim and merely as well as Dylan, discussing and addressing questions that are brought up by you. 08:07:27 During the, the, the keynote, as well as kind of organic questions that arise on the topic of galactic outflows. I know there's a strong preference by people to respond in the zoom chat, but that doesn't really get preserved long term. 08:07:44 So if I could ask people to have their questions comments and discussion in the Slack channel Halo 21 week five outflows that will be better preserved and it allows threading so if somebody has a question then someone can respond as a thread on that and 08:07:59 it keeps it a little bit more organized and sorted. So please use that, that Slack channel and we can, we can have discussions, and I'll be drawing from the discussions there and the comments there to see the discussion for the panel after Dylan's talk. 08:08:16 One quick note you can use little emojis on individual questions that you think are really relevant or, or, kind of, upload them with like a plus or a thumbs up or something like that, which will also indicate to other members of the, of the program like 08:08:33 you think this is important and if you see a lot of of pluses on something then that might indicate that people should check it out or uploaded in that capacity. 08:08:42 You could even give thumbs down, I saw a few thumbs down so that's that's an option if you think it's a really dumb question or comment. 08:08:51 Okay, so our speaker for this, that our first speaker for this week's theme of how does gas flow out of galaxies is dr Dillon Nelson, Dylan, is we've finished his PhD at Harvard five years ago and has been at Max Planck in gardening for the last five 08:09:16 and he just began last fall, as a group leader at the Institute for theoretical astrophysics at Heidelberg University. He's really, and a young star in the field and has done an enormous amount of work in the, in the last 510 years. 08:09:26 He was co pi of the illustrious tng simulations and has done an enormous amount of work on the CGM both small scales, as well as highly organized, oxygen, a variety of different studies of the CGM and we're, We're happy to have him joining I'm happy to 08:09:45 have him joining So, do you want to, to share your screen, Dylan. 08:09:52 Yeah, thanks everyone I will take over. 08:09:58 I hope you can see that. 08:10:01 Yes. 08:10:02 and also that. Yes. That looks great. 08:10:07 Verbal so I hope everyone can hear me okay if I drop out of it, please just let me know and I'll kill my video, perhaps. So, yeah, thanks very much. Cameron for the kind introduction. 08:10:28 I'm very happy to take us into week five, and the topic the question I've been assigned here is how does gas flow out of galaxies, like to Calculus. The theory half, so we'll get the observational perspective from Tim Heckman on Thursday. 08:10:38 And 08:10:38 the question then. So, how does gas flow out of galaxies I want to kill the suspense. From the beginning, just give you the answer. And I think the answer is yes right we all agree. 08:10:46 Yes it does. 08:10:48 Yes they do make outflows, and in spectacular ways. So, although I've been told not to talk about observations whatsoever. I'm going to disobey that rule and give a very quick introduction to set the stage for this week and talking about outflows. 08:11:02 In particular, how do we actually see outflows and the real universe is a variety of techniques, right, direct imaging especially in the local universe in a mission, where we're typically thinking about broad narrow line decomposition with the broad component 08:11:17 representing the outflow in material, and also an absorption so down the barrel studies where gas is blue shifted representing gas flowing towards the observer, and more recently, perhaps also direct imaging but molecular gas see oh this is a red shoe 08:11:32 six example from Alma. I would also groups a lineman alpha Hey those into this category. So, a diversity of techniques for studying and characterizing the physical properties of outflows, and they these data give us a very nice perspective on how outflows 08:11:49 work, and what their properties are in terms of the galaxy population and I just want to flash through a few results which give a sense of that the kinds of influences that we can draw like how alpha properties relate to galaxies. 08:12:03 So the first year is from chisel 2014 It's a velocity of our flowing materials a function of galaxies during mass This is a call tracer silicon to which is zero of the last East increasing for more massive galaxies very broadly speaking, Similarly album 08:12:29 last weekend is a function of stellar stellar mass and even colder so the MD tracer also which is zero again slight trends of increasing velocity with galaxy mass. So Richard one magnesium to again alpha velocity versus stellar mass, increasing roughly but lots of diversity lots of 08:12:36 lots of diversity lots of scattered on the right I'm showing actually a mass outflow rate so solar masses per year of the outflow again as a function of galaxy stellar mass, just on color coded by black hole luminosity so presumably tracing something 08:12:50 here the connection between master control out of galaxies and what the black hole of the center is doing redshift to. So, this color is shine the fraction of galaxies with a strong detected outflow, in a sense, again as a function of galaxies stellar 08:13:04 mass and here the y axis is actually offset from the star from the main sequence. So, galaxies, at the fixed mass so going in a vertical line here which have higher star formation rates also tend to be more frequently detected with a strong outflow so 08:13:19 there's clearly star formation rate outflow connections. 08:13:23 And finally something different so strength of cold gas I'm using to come up with as a function of distance from galaxies. This is from Sloan magnesium too, and the green and the purple points are showing respectively sightlines oriented along the minor 08:13:38 versus the major axes of galaxies showing a preferential enhancement of cold, presumably outflow tracing gas along certain directions with respect to the galaxies themselves. 08:13:49 So this is just a kind of give a taste of the kinds of data and inferences we can draw and the observational space, and which is to say that we have a very diverse and complex set of empirical constraints on galactic outflows. 08:14:03 The question then is really how do galaxies, not get their gas which is another favorite topic of mine but how do they get rid of their gas. And if you'll indulge me I want to schematically orient set the stage here, just we're all in the same thinking 08:14:16 about the same regime. And so we're sitting inside a dark matter Halo there's a galaxy galaxy at the center. 08:14:23 Of course we have inflows we have cosmic gas accretion from larger scales, and then especially massive Halo, this will form into a quasi static hot atmosphere, right with an associated variable shock, this hot gas can cool in a uniform fashion and as 08:14:38 we've also heard a lot about lately. We can also have localized cooling phenomena Institute kind of cooling phenomena. 08:14:54 Gotta don't live in isolation, right there are also other galaxies and galaxy galaxy interactions and mergers, so the galaxy moving the other way will be going undergoing stripping processes and then of course galaxies are driving outflows outflows and 08:15:02 winds on us the two terms interchangeably. And let's not forget that not only central galaxies in the universe do such processes also all the neighboring galaxies also may be driving their own respective outflows. 08:15:14 Some of these may escape the halo others of course dumped on a farm fountain or recycling flows and and collectively that buzz word here the course is positive baryons cycle. 08:15:25 All of this complex physics, taking place inside the CGM, and gaseous halos which is why we love to study this topic so much. And today I'm really focused on just one, one element of this bear and cycle which is the outflows themselves driven from galaxies 08:15:41 across time. 08:15:43 And so in this context I see really three very broad questions right one is about the small scale launching physics heavy supernova and black holes actually produce outflows. 08:15:53 Second is what are the properties of those outflows as they propagate away from galaxies, as a function of the galaxy types galaxy properties themselves so that is so what is the connection between outflows and the galaxy population, and three is what 08:16:06 do altos actually do to shape the galaxies, which is the topic for the future weeks of this conference. And so today I'm really going to be focused on the second question which is, what are the properties, the physical properties of altos and how do they 08:16:20 they relate to galaxies. 08:16:22 And in the sense I'm going to propose here that a fundamental quantity, we like to think about is the mascot flow rate so m dot out in terms of solar masses per year as a unit. 08:16:35 Now this has of course many dependencies. I think we can all agree that the amount of mass flowing away from a galaxy depends on the spatial location. 08:16:45 so where you are with respect to the galaxy. 08:16:48 We also agree that gas flows in different directions and with different speeds, so the after the rate depends on philosophy. 08:16:55 And then, generally speaking on phase, which I right here is temperature, density and velocity of the gas. 08:17:02 And so if you allow me I'll simplify velocity just to a scalar, so radio velocity and I'll simplify the position into just distance from the galaxy and angle, some galactic centric angle with respect to the galaxy and so what we really want to measure 08:17:16 and understand them, is the rate, how much mass and how much energy is flowing away from galaxies. How does this evolved with radius, what is the geometry of these flows How fast are these flows. 08:17:31 What phase, what is the phase structure and terms of density in terms of temperature and are these flows smooth or are they clumping what is their subtraction. 08:17:38 And finally, what is their metal content, and here of course I've admitted many other things also carried by outflows which I want, unfortunately have time to talk about today. 08:17:48 For instance dusts magnetic fields and cosmic ray components, also, how flying in the gas. 08:17:54 Okay, but this is great this is in some sense a very high dimensional function that we would like to characterize from the simulations and also empirically, but unfortunately this is only half the story there's there's, it's only half of the dependencies 08:18:08 here because the outflow rates from galaxies also depends upon the galaxy itself. I think we also would all agree. 08:18:17 So if we measure all these properties and how gas is flowing. That answer depends also on the Galaxy mass itself, say the stellar mass of the galaxy. 08:18:27 And at fixed mass, it also depends plausibly on time on cosmic epoch on redshift, and it fixed mass and fixed redshift it also depends upon a star formation right, I think we also all would agree. 08:18:41 m dot star related to the number of supernova type to going off. 08:18:46 Okay, I think we're on the same page for now maybe I start to raise some eyebrows when I say the outflows also depend critically on the rate of energy injection from supermassive black holes. 08:18:59 And indeed, as I'll come back to later in this talk that interplay between these two terms is crucial. Indeed. 08:19:06 But what else so other properties of galaxies that outflows depend on here I just give one example Kappa gas, so by kappa i mean the circularity parameter, I mean a measure of the amount of rotational supports of a galaxy, which is to say, is it risky, 08:19:21 or how to ski or not to ski is it in the gashes components. Maybe it's not something that a lot of us think about how can they morphology of the gas of the disc actually influence and regulate the properties of outflows safety amount of our play mass 08:19:37 is a function of the temperature and the velocity, but it can come back to that later so the questions that we're asking here on the bottom half of these dependencies are really, how do altos evolve with mass galaxy or halal mass with redshift, and how 08:19:52 how are they set by the available energetics from supernovae, and by supermassive black holes and what is the interplay between those two processes. And finally, other galactic properties such as morphology, which can actually be quite relevant for galactic 08:20:09 outflows. 08:20:10 Okay so, from my perspective this is a computational problem right I'm given the theory perspective and we would like to understand outflows, which is to say, feedback physics from galaxies, from the point of view of simulations unfortunately galaxy formation 08:20:25 is a very difficult problem primarily because it's multi scale, and it has an enormous range spatial dynamic range, like illustrate that just by zooming in, right so starting from the large scale structures of the universe on Giga parsecs scales down 08:20:58 to a halo scales, say mega parts that were dark matter halos and Serkan galactic media or for me down to individual galaxy scale say on the killer per sec levels, but down further to par second indeed micro power second milli parsecs scales where say 08:21:01 the accretion disk around the supermassive black hole is launching object or a wind. 08:21:04 Now the problem here is that this is not just the problem of collapse, unfortunately what happens on these very small scales actually back propagates up to larger and larger scales, we know that feedback physics, say, on the scales of the equation just 08:21:19 of the black hole can actually impact the baryons, and not only the variance also the dark matter, distribution, saying on enormously larger scales. This is a 10 to the 12 orders of magnitude across these two scales here in the best simulations, have 08:21:34 logical simulations which exists to date capture more like 10 to the six orders of magnitude so we're still about a million times short, in terms of directly resolving these processes. 08:21:44 This is why understanding outflows in the context of galaxies is of course a challenging problem and that leads us to the issue of how do we actually create these houses via feedback processes, which is to say, how do we develop these kinds of models 08:22:01 in simulations. Now the question here is, how do we produce a model for galaxy formation for feedback and for outflows, which has some physically predictive power given these numerical limitations, let me just make 233 claims here. 08:22:18 So the first is that direct numerical simulations of DNS is impossible for the problem of galaxy formation and it will always be impossible that simply saying that there are unresolved physical scales and there will always be such. 08:22:32 So what do we do, we need to develop approximations, so approximate models or sub grid scale models which tell us, approximately what is going on below the resolution limit of numerical simulations and this is of course what we mean by some good physics. 08:22:49 Now I and many of my colleagues think that it's critically important to address this boundary, to identify and to develop models which specifically address the problem rather than ignoring it as is often the case in numerical studies for instance of diffusion 08:23:07 of magnetic field amplification on small scales or turbulent energy dissipation due to a cascade to small scales for instance, and of course it's. Ideally, nice if this scale separation between the resolve scales and the unresolved scales occurs at some 08:23:26 physically convenience or meaningful boundary and current cosmological simulations for instance resolving or not to giant molecular clouds is such a boundary. 08:23:36 So, outflows are a computationally difficult problem to tackle and this means obviously that a number of different approaches have been developed to do so across a number of scales we've, we've heard a lot in the first few weeks about cloud scale simulation 08:23:51 so wind tunnel cloud crushing works. 08:23:56 Next, of course, moving up in scale we have patches pieces of the ASM of discs, often in the form of these tall box simulations like silk antivirus. 08:24:05 Of course these citations are not by any means exhaustive here I'm just giving a few canonical examples or recent examples. 08:24:13 Next, of course we have global idealize disk simulations. So now we have a global galaxy disk which is of course critically important for resolving the evolution of outflows and say a fountain structure. 08:24:26 But idealised and here often talking about studying the impact of supernova so setting off bonds in this disk and seemed to kind of help flows that propagate away. 08:24:37 Next, of course we have idealize global halos. So, in such a simulation we might not even have a desk or care about the galaxy in the center, or rather a ball of top gas and hydrostatic equilibrium within mfW potential same here often the focus is on 08:24:52 the black holes, studying black hole physics feedback physics and jets, and again understanding how how flows are then generated and how they propagate away from the center of these handles. 08:25:03 Next, of course we have cosmological zooms in relation so this is the first instance where the initial conditions are cosmological, which, by which I mean motivated by kind of lambda CDN constraints. 08:25:15 And in practice. What this means is that the situation becomes much more complex, right, particularly in the inflows and the lack of cemeteries, which are present in some of the similar approaches. 08:25:27 finally of course we're moving from individual zoom simulations of individual galaxies and halos to full cosmological volumes where we are sacrificing more even the resolution, we have in order to gain this kind of global perspective on the Galaxy population 08:25:44 as a whole. 08:25:46 I think you agree from left to right here there's that there's a different scales, being resolved in different scales being probed and each of these approaches, all the way from the second sub scales of interfaces and clouds to Halo scales in the middle 08:26:04 and cosmological scales on the right. And I met. So in the in the next 2025 minutes of this talk I'm actually going to focus, almost exclusively on cosmological simulations, but this is what I've worked on and this is primarily where we can make this 08:26:17 population wide connection between the properties of galactic outflows and the properties of the galaxies themselves. 08:26:25 So, let's talk about supernova and black holes in galaxy simulations. First thing we know I think very clearly is that stars released, energy, and actually quite a bit of energy. 08:26:36 So this is a classic plot from aggregates 2012 showing essentially the energy injection rate for a stellar population as a function of time. 08:26:44 The white line shows the volumetric radiation the blue line stellar wins and the red line type two supernova after some appropriate delay time. One as of course are not shown here. 08:26:56 The issue is we have all this energy but what do we do with it in the simulation, and the unfortunate answer is that what we do with it makes a big difference. 08:27:06 So that's what I'm showing you here from this comparison study by Rosedale This is six different simulations of an idealized isolated galaxy disk with six different implementations of what you do with the energy available from stars, so no feedback you 08:27:20 dump it in a thermal energy you dump it in in a more stochastic sense and delayed cooling model and you give a kinetic energy input or a hybrid mechanical type model and the point is that all six of these have been used in r&d, they are used in galaxy 08:27:35 simulations different simulation projects. 08:27:38 And as you can see by I'm the structure of the desk here so face on an edge on depends a lot on the numerical implementation of the supernova feedback. 08:27:50 So very similar study from Matthew Smith and 2018 showed that the outflow properties are also similarly affected. So on the top row is that outflow rate, and on the bottom row is outflow velocity as function of time left panel shows one kilo per sec above 08:28:07 the disk, and the right panel tempo parsecs above the disk and again the seven colored lines here are seven different models for how you do supernova feedback and the point is simply that the outflow rates, and the outflow velocities from such this can 08:28:21 vary by factors of many. When you change how you input stellar energy in a type of galaxies simulation like this. 08:28:32 Okay, so that's stars. We also know that supermassive black holes released energy and indeed, a lot of it the typical argument goes that the the mass density of black holes is something like point 1% that of stars in the universe. 08:28:46 If the energy available from bike holes is roughly 10% radio efficiency of 10% times the MC squared, and compared to usual argument for supernova of roughly kind of 51 herbs per hundred solar masses formed. 08:29:02 If you take the ratio of these two you actually find that it's almost immunity, which means that the total feedback energy available from black holes is somehow comparable to that from supernova. 08:29:24 Across the Universe, and so you better not neglected. 08:29:17 Now of course, very detailed small scale simulations of black hole accretion discs, start are starting to probe the physics of how this energy is actually released from the Christian disk and from the emergence of relativistic jets, saying, and to be 08:29:34 honest we have not yet coupled very closely galaxy scale simulations to these kinds of small scale black hole simulations, I want to give you one example of how we're starting to do that though. 08:29:46 So this slide shows you the radiative efficiency of a black hole is a function of its in. And this is an answer obtained from small scale simulations of McKinney and as implemented in the new. 08:30:01 New Horizons simulations, which are cosmological gaps information simulation. So the point here is that if we can measure or think that we can track the black hole spin evolution in the simulation we can actually use this answer from the smaller scale 08:30:13 simulations to tell us what the radiative efficiency should be. 08:30:18 And again, the radio efficiency is this free not free grammar but is this parameter sitting in between. 08:30:25 m dot c squared the rest master creative energy and the amount of feedback energy we have available to us in our galaxy simulation to drive outflows okay but if you're not doing GRRMHD, which means general relativistic radiative Magneto hydrogen chemical 08:30:44 simulations of accretion discs around supermassive black holes and I think I am. Ron same page here not many of us here are doing that, then while you're doing in practice with this energy is very simple, right we're actually developing very simple models, 08:30:59 some good models for how this energy released from supermassive black holes can couple to the galaxy and stripe outflows. And what I like to do actually is give a very practical sense of how we implement supermassive black hole energy injection in galaxy 08:31:18 scale simulation. So, this is a kind of schematic diagram of a CGM of a halo out to the view radius and you should imagine that there's a black hole sitting in the very center, so you'll create some mass you have some available energy, what would you 08:31:30 like to do with it you could dump it in as thermal energy into local vicinity of the black hole. For instance, you could inflate a thermal bubble offset out in the ICM out in the halo perhaps this is mimicking and unresolved relativistic jets, which goes 08:31:47 out and then fleet such a bubble. 08:31:49 You can create large big energetic bubbles or maybe you create a population of smaller smoother gentler and time. 08:31:59 Thermal energy injections or maybe you do something completely different. Maybe you inject momentum so kinetic energy. 08:32:05 Maybe you do that locally and give it a particular direction, maybe you do this in the bipolar sense. Or maybe you make an ice a tropic energy injection. 08:32:14 Maybe you do none of those things maybe you simply capture what you think the black holes actually do right which is to offset the cooling flow. So maybe you simply shut off the cooling by hand, so to speak, in the halo gas. 08:32:29 Now it's interesting I think is that actually all of these different methods are used by different simulations in different contexts. And this, they're actually all used in practice subjects again to add additional detail so for instance if you're dumping 08:32:42 and thermal energy is it continuous and time or is it bursting, and some simulations do both. 08:32:48 Just one of these modes act all accretion rates all the time whereas one only in the high accretion state or the Christian state of the black hole and again you have different simulations implementing different assumptions, different model choices in 08:33:01 this regard. 08:33:04 So, now they've scared you enough, I think, as to how we actually implement feedback in galaxy scale simulations I want to jump right into the question of what kind of outflows are produced in the simulations. 08:33:19 So, there's an overarching theme to the next second half of this talk, and that is as follows despite this simplicity, which I've just shown you right the simplicity of the sub grid models for feedback, which is simplicity at the scale of the energy injection, 08:33:36 which is the resolution scale, the very small scale on the simulation. 08:33:40 The resulting flows which develop and propagate away from galaxies and through the CGM are quite interesting and they display a number of complex and almost emergent behaviors and properties which are somewhat unexpected. 08:33:58 So I'll start with point number one, something I think very not controversial whatsoever. At least I hope, bigger galaxies have faster outflows. 08:34:09 I'm showing you scaling of alpha velocity as a function of galaxy seller mass so this is from continue simulations and Richard to the blue points are showing you the 75th percentile of velocity and the orange the 95th percentile. 08:34:23 So very roughly speaking, outflows when you're thinking about 10 kilo parsecs away from the galaxy has speeds of order 100 200 kilometers a second for low mass galaxies, increasing up to say 400 800 kilometers per second at the Milky Way masters gene. 08:34:42 If you actually look at the distribution of outflow velocities around galaxies of different masters is quite interesting so this is showing you the amount of mass flowing as a function of velocity. 08:34:54 Each line shows the stat average for galaxies at different stellar mass is increasing from 10 to the eight or small galaxies up to 10 to the 11.5 in pink. 08:35:06 Now I see is that around, small galaxies of velocities of outflows are relatively small right few hundred kilometers a second. It's the kind of take the shape of this compact core which gradually grows to higher and higher velocities and this again this 08:35:23 is low mass galaxies. This is where supernova type two are driving the energetics and the production of these flows, and they result in these mild, I would call them outflow velocities, until we get to some critical stellar mass galaxy mass which here 08:35:39 is about 10 to the 10.5 stellar mass, this is where, and this model, black hole start to act, and it's where they really start to have an influence on driving outflows away from galaxies and what the black hole feedback does in superposition to the supernova 08:35:52 feedback is it produces these very high velocity tales. So, the black hole feedback is what results and these simulations, in our losses which exceed 1000 kilometers a second now too many thousands of kilometers per second. 08:36:07 Now these numbers start a little bit I put here, the average outflow speed for a milky way mass galaxy Richard zero. So the media and and it's something like 150 kilometers a second so relatively mild, that's point number one point number two, which I 08:36:24 is quite important to point out the properties of outflow and gas in these kinds of galaxies and cosmological simulations are distinct from what you put in, which is to say what you get out is not what you put in. 08:36:40 I think that's very clearly seen here on this plot on the right. So, this white line for low mass galaxies actually shows you a model assumption in these tangi simulations, which is a minimum wind velocity of 350 kilometers per second. 08:36:54 Now what this is. 08:36:56 mimicking is unresolved Physics for low mass galaxies. 08:37:00 But when we go and actually look at the flows which emerged from the systems that any resolvable scale we see that they actually move lower velocities then this supposed minimum, which means that the properties of the outflows when you see them propagating 08:37:14 through the halo have decoupled from the assumptions from the model inputs at the resolution scale, which I also see, which I also call that the scale of energy injection. 08:37:28 And we'll see that game come back. Okay so point number three. This may be either extremely obvious to you are extremely controversial and I'm quite curious which. 08:37:37 So, point number three is that outflow slow down with distance away from a galaxy. 08:37:44 They do not speed up again I'm showing you that here also velocity is a function of galaxy stellar mass This is the medium relation, when you're considering flows at different distances so red, blue, green are 1020 and 40 kilo parsecs away from that galaxy 08:38:06 and you can see that over the last few drops by roughly 50% or a factor of two. 08:38:06 Now if I compare that to the scape say the Delta escape velocity to go from 10 kilo parsecs to the very old ladies so we'd have to be on this line to move from 10 kilometer six out to the edge of the halo if gravity was the only thing holding us back. 08:38:23 And we can see the for instance of this red line at tempo bar six is above this point, which means that this is showing the 19th last velocity percentile so the high velocity tailed gas can be kind of sufficient velocity to escape the system, which is 08:38:38 of course not true. If we look at the 50th percentile so the bulk of the flow, which is a much lower velocities but still you see as you move out to bed, blue and green increasing distances the velocity slow down as you move away from the galaxy. 08:38:53 I want to show you a perspective of this from the eagle simulation so this is analysis from Mitchell 20, this is even more direct so it's alpha velocity as a function of distance away from the galaxy normalized spite of your radius. 08:39:06 The solid line shows all gas which is sitting there in space and the dash line which you should focus on actually shows the gas which has left the ice dam which has been ejected from the galaxy. 08:39:16 And the alpha velocity is a monotonic Lee decreasing function of distance everywhere in the resolved, Halo to resolve flow think the point here is that wins propagate into something they propagate into the CGM into the pre existing CGM and not a vacuum. 08:39:35 And I think to capture the proper evolution of wins we should think more about this difference between simulations which capture the mass and the properties of the medium into which altos propagate as opposed to models which model propagation into effectively 08:39:53 a vacuum. 08:39:54 And the context of the eagle simulations here the interpretation of this declining speed with distance is a general conversion rate from kinetic to thermal support for the wind. 08:40:08 Point number four is that winds are not nice a topic. 08:40:12 And so I show this image again it's one of my favorites. It's showing you the geometry of a flow, emerging from this galaxy, so the color in the background is the gas density map, and more importantly are the streamlines on top, which show the velocity 08:40:25 field of the outflow. 08:40:28 And inflow for that matter, and this is a discount CF oriented edge on, which means it's in the horizontal orientation, if you like. And what you see is that outflows, as indicated by the arrows are preferentially propagating along the minor axis of this 08:40:43 galaxy whereas inflows tend to be found along the other directions more aligned with the major x galaxies. 08:40:51 We can see this more quantitatively by looking at the amount of gas outflow as a function of angle. So from pi to minus pi, which is going around a circle from zero to 360 degrees, but the minor axis of the galaxies I've indicated at these two points. 08:41:07 And so where you see yellow that's where the mass, how flow rate is the highest and you see how outflows are not isotopic they're not equally spread an angle of a rather preferentially are propagating the line with the mind access galaxies. 08:41:21 Why is particularly cool in the context of simulations, is that when we injects energy from supernova so this is a 10th of the 10 stellar mass regime I'm thinking about this is where supernovae are doing the work. 08:41:36 When we inject energy, it is completely isotopic at the injection scale, which means there is no preferred direction whatsoever on the scale of 100 parsecs or so. 08:41:51 This is a scale of the energy injection in this simulation. However, when we go and look at the resolved flow which emerges from the galaxy, you see it picks up this natural combination effects are unavoidable consequence right of having a disc in one 08:42:02 direction and having the pressure gradients the path of least resistance can be other direction, but this is a very nice example of what you get out is not what you put in and how simplified models for sub grid physics can lead to more interesting behavior 08:42:17 on resolve scales, 08:42:21 another scene point is that this combination requires the disc to be here right if there was no gashes disc impeding the flow in the major access direction, it would be a nice of tropic flow and indeed a high redshift galaxies are not discs. 08:42:35 So we actually have a very clear expectation from the simulation which is that only as you go to lower lower redshift as galaxy discs form in the universe, which is to say the gas settles into rotation Lee supported thin structures. 08:42:49 That is when we expect such a combination to emerge, very obviously by Richard for one, but in this simulation is very hard to find such a signal average of two or high. 08:43:01 Okay, point number five, essentially, is that black holes make very cool outflows, it's just my excuse to show the interesting structure morphology and properties of what is the supermassive black hole driven flow, and he's tangy simulations. 08:43:18 So, these five rows show time evolution. Each has about 100 mega years apart, and in the very initial phase here what you're seeing is the first onset of strong black color activity in this galaxy. 08:43:30 This is the onset of quenching because this Baikal activity will eventually push this galaxy off the main sequence and quench it star formation rate, the left panel is showing you velocity and the right panel temperature where you can actually see already 08:43:43 in this first image, you look at very small scale just kind of 12345 or so loops of high temperature gas, these are five discrete energy injection events from the black hole which have already taken place and are now propagating outwards. 08:43:59 You see what this flow emerged or flow evolves into is a very high velocity. So exceeding 1000 kilometers a second flow as it passes the variable radius of the Halo, which is indicated by the white circles on very large scales. 08:44:13 This is a very high temperature flow. So it's almost a yes high pressure high temperature bubbles which are being inflated by the action of the supermassive black hole which then expand into the CGM and out into the intergalactic medium. 08:44:29 I want to point out here that if I zoom in, so the galaxy in this image is miniscule, the galaxy I've oriented it edge on again but in the vertical direction so it looks something like this if I zoom into the center in the region of influence of the black 08:44:42 hole is this little white circle here, and it's. That's the region where energy is injected, leading to these large scale outflows. 08:44:53 Again, I want to point out to kind of cool features here. So the first is this, this guy, you see it this is a satellite galaxy, which runs through the middle of this halo and gets destroyed. 08:45:05 Right. So what you see at the final time is this population of cool high velocity gas clouds moving outwards, which might be mistaken for cold gas and trained, or or injected in a black culture and women's right but in reality this is simply stripped 08:45:21 interstellar medium gas, have a satellite galaxy. 08:45:26 medium gas of a satellite galaxy. Medical thing if you can see a very small here so particular particularly intrigued by this these are cold clouds, it looked to be raining down onto the galaxy in the orthogonal direction to the outflow so along the major 08:45:39 axis here, you're going to see a bit more structure in the zoom in right so this seems to be a bit of a condensation or precipitation type process remains to be D. 08:45:49 What exactly is going on here but it seems to be a signature of fueling. 08:45:54 Again, in a volume of the halo distinct from the outflow. 08:45:59 We add on two additional panels here the same outflows to density and the listing and density you can clearly see the morphology of this flow is a pile up of gas and almost a shell. 08:46:30 So, this succession of outflows have driven shells of over dense gas which leads to this under dense, expanding bubble feature, which is also dredging up very enriched metal right so before this black hole really get started the intergalactic medium which 08:46:28 you can see right on the outskirts here is sitting down at split once told her or less actually. But after a few hundred mega years this black hole German outflow is actually taken supersoldier gas from the galaxy itself on very small scales polluted 08:46:42 the entire circle back to medium and indeed the near field, IBM, as well. 08:46:52 Point number six is that these black hole driven outflows also produced in these tg simulations, a multi phase structure and the outflows which I think is quite interesting. 08:47:01 So similar to before what I'm showing you here is the distribution of our flowing mass in terms of its temperature and each line shows the stack for galaxies of different masses and kind of the eight, all the way up to 10 to 11.5. 08:47:15 The main component you hear you see is this a cult hot phase. So it's this large, calcium, of kind of distribution with the peak of shifting to higher and higher temperatures basically tracking the variable temperature of this of these increasingly massive 08:47:31 galaxies right until we get to the purple line here. Again, this is 10 to the 10.5. This is where black holes start to do their work. And at this point we actually see the emergence of this secondary peak. 08:47:46 So, the emergence of a bind modality in the temperature distribution between a colder components and a warmer components. 08:47:53 This is a signature of the direct objective nature of the black hole feedback literally throwing out cold gas from the central galaxy has quite high intermediate velocities out to very large distances, away from galaxies. 08:48:11 It's also important to point out that these phases at different temperatures have different kinematics. 08:48:18 So that's what I'm showing you here 08:48:23 is pop shows the plane of velocity alpha velocity and gas temperature were again, yellow and green indicate where the mass flux is. 08:48:32 And so what you see is that the colder component tend to the four to five Kelvin has moderate velocities a few hundred kilometers per second, and the bulk of hot component does as well. 08:48:42 However, this tale towards extremely high velocity is contributed only by the hots components. 08:48:48 So the different phases have distinct kinematics, and I think another important points here to mention, you can see clearly in the left panel again is that the colder peak is much smaller than the hotter peak at this is saying that actually the mass flux 08:49:02 at these distances for these kinds of massive galaxies is dominated by the hot phase, which, unfortunately, is of course a very hard to observational they characterize beyond the local universe so if these kinds of expectations are true, from the simulation, 08:49:18 it means that it's going to be very hard to find the bulk of the outflow in mass. If it's in such 10 to the six to 10 to the seven Kelvin, or phases. 08:49:30 Now this link between kinetics and phase is very generic so here I'm showing you a plot from em Snyder's recent paper I flipped it i've i've inverted it flipped it horizontally, so that the axes are actually the same very similar to the plot on the right. 08:49:43 So, increasing temperature, and increasing outflow of last year and you see this exact same banana shape again showing you that the hot tail, sorry that the high velocity tail is coming from the hotter components, whereas the colder components of the 08:49:57 outflow is that more reasonable velocities. 08:50:01 Okay, point number seven mass loading so I managed to get 70% of the way through this talk without ever mentioning mass loading which is a bit strange for an outflow talk. 08:50:11 The definition of course to get us on the same page is the mass outflow rate normalized by the star formation rate of the galaxy itself. So let me show you the relation of mass loading of outflows is a function of galaxies daily mass This is again from 08:50:25 the tangi simulations are focused on the blue line. So Macedonian factors are high for low mass galaxies and then they dropped from say 10 minutes 1015 maybe even 100 for very small doors down to effect to a few Muslims have a few for the Milky Way mass 08:50:41 regime. 08:50:43 But at this point, you see that something rather strange happens this relation actually has an inversion, and the mass looting starts to go up again. 08:50:51 And this is because again at this mass scale black holes, start to dominate the energetics sorry the inner energetics of outflow production. It's no longer the supernova doing the work but instead the black holes which Dr very strong outflows, and at 08:51:07 at the same time, so they're pumping up the flow rate but the same time quenching the systems and depressing the star formation rate, which leads to this very pronounced feature. 08:51:18 This is also seen in the ego simulation, which I find very interesting so this is complimentary similar analysis from Peter Mitchell's recent paper. 08:51:26 The tng simulations are shown in blue very similar to the lines on the right on the left and the ego simulations are shown in yellow. 08:51:34 where in general they find slightly lower mass loading factors at the same galaxy masses on the main sequence here, but also this inversion, which an eagle simulation is also due to the startup of strong black hole feedback, again at this mass scale temper 08:51:52 temper temper 10.5. 08:51:55 I think is very interesting here is that this feature in Eagle is only seen in two out of four of these yellow lines. If you notice, it's only seen at the large distances, 50 k BC and the various ways. 08:52:11 And in contrast near the galaxy, the mass flow rates are very small 10 to 26. 08:52:19 This is actually a completely opposite to the behavior seen in the tg simulations, where you see the mass of started mastering factors or large attempt ABC and then they start to drop to larger distances so the relation is actually the complete opposite. 08:52:34 With respect to distance. This is a signpost that in these two simulations with these two models the propagation of wins through the CGM is very different. 08:52:49 In tangi, they are declining, they're getting weaker and weaker was an eagle they're actually getting more and more mass loaded and more mass flows out across the various radius and across the galaxy itself implication here is that this should be very 08:52:57 different fountain flows right very different recycling properties in these halos in these two different models. 08:53:05 Let me slide those over to the left to give a broader perspective again the mass loading factors. So on the top panel here I'm showing a fire simulation and on the bottom me how and the pictures are very similar so in fire you have very large mass audience 08:53:18 right for dwarf galaxies dropping to have order unity around enough to a mass scale and similar in the new house simulations massively in the border 100 dropping down to unity around the Milky Way mascot I particularly like this figure. 08:53:34 At the bottom, because the blue open symbols actually so the mass leading factors which would be inferred if you only considered very cold gas so less than 10 to three Kelvin, which would be observable as a molecular phase, which makes it very clear that 08:53:48 the inferred mass loading factors depend sensitively on the phase which are observing in order to make this estimate. 08:53:56 Okay so, general trends versus stellar mass, I think it's very common feature, large mass loadings declining to have order unity or a few at the Milky Way to tend to the 10.5 still in mass think it's quite important to point out, only on the left here. 08:54:12 Do we see this feature of the inversion of the mass loading, because this is due to black hole feedback, which is not present in the fire simulations or the neon simulations for instance, as a function of radius, it depends very strongly on the model. 08:54:29 So, in some cases, massively in fact just can decline with radius as is the case in tg or fire, but in other cases they can actually increase with radius, and this should again be reflected in the global gas content in the baryonic closure of these halos 08:54:44 how much gas is left, and should be I think observation and constraint. In that sense, versus redshift is a similar issue. Some models shows by the increasing mass loadings others slightly decreasing as Richard goes to zero at fixed mass. 08:55:01 So it really depends. And the comment on Richard let me, let me show or gives me an excuse to show one more plots from a different simulation This is from the horizon AGN salvation. 08:55:11 So no longer Macedonia factor but actually Rob mass outflow rate, solar masses per year as a function of galaxy stellar mass. So, these outflow rates are actually a bit smaller a bit weaker than in most of the other simulations I've showed you, but when 08:55:26 you don't do the normalization we can get a more clear sense right big galaxies drive very strong outflows with enormous outflow rates, and as you go from low redshift up to hybrid shift. 08:55:39 Hi, Richard galaxies drive enormously strong, how flows, just so we don't lose track of that fact hybrid of galaxies also drive very fast outflow so this is again off of last year versus galaxy mass in the tangy model, Greg from redshift six down to Richard 08:55:59 zero and that fixed mass. Hi, Richard outflows can be a factor of two factor three faster. This is the outcome of this model, in particular, okay point 10 is a very brief money back to the mass loading factor and this feature of the inversion. 08:56:14 So, the definition of mass loading has a normalization of the star formation rate, which is in essence and implicit assumption that the energetics involved in creating this outflow come from supernova type two. 08:56:31 And I think we can see regimes where this is not true. And where this assumption, or rather this normalization can lead to very biased conclusions. That leads me to make a very provocative x here to cross this out and just say, let's not do that. 08:56:47 Let's instead stick to the simpler, more direct measurement of the outflow rate, itself, 08:56:55 I think, okay, I'll say that briefly, I think, despite having crossed out the star formation rates in the masturbating factor. Let me point out that outflow properties do of course relate, in some cases directly to the star formation properties of the 08:57:10 galaxy. So let me walk you through this video very briefly. This shows you, and galaxies on the plane of star formation main sequence offset versus mass. 08:57:21 So by starvation main sequence offset any distance from the main sequence to the Main Sequences here sitting in zero and points up here have higher star formation rates and points down here have a lower star formation rates. 08:57:33 The color shows how fast velocities are emerging from those galaxies, relative to the average so red means that galaxies Drive faster outflows than the average galaxy at that mass blue means galaxies drives slower outflows. 08:57:50 What we see is a clear correlation right outflow velocity correlates with offset from the star for me main sequence correlates with star formation rates. 08:58:00 In particular, if you're above the star formation main sequence you're driving faster outflows in this model. 08:58:06 But as we saw before, at high mass, this actually inverts this relation. So for massive systems. These red pixels down here show you that it's actually low star formation rate galaxies which start to drive the fastest outflows. 08:58:22 So this inversion of the relationship between alpha velocity and star formation right properties is another example of the complication introduced when you also are considering the role of black hole feedback. 08:58:36 Okay, this is my rant on comparing simulations to observations, which is to simply say, we need to work more on these comparisons because I don't think we're doing good enough, so far. 08:58:48 And I want to end on a slight and pop more positive notes, which is does the fact that we saw how inflows and outflows occupy different volumes of the Halo is this imprinting any sort of signature which we can observe in the CGM properties. 08:59:04 so this was a study, led by Surrey saline Peru. And the idea was that if we look at the metal the city of CGM gas as a function of angle with respect to the galaxy. 08:59:16 We should see a preferential signal of enhanced metal along the minor axis of galaxies right due to the preference for outflows to occupy this space, as opposed to metal for employers to occupy that major access aligned directions and indeed if we go 08:59:30 the simulations, that's exactly what we find. So a clear prediction for correlation between gas middle of the city, and as a musical angle between a galaxy and an absorber say at 100 kilobits or sex, or you have lower medalist fees along the major access 08:59:44 and higher medalist these along the minor access, something which has not yet been observed yet but hopefully will be in the future. 08:59:54 Okay so, instead of concluding anything I simply labeled the conclusion slide points to argue over the first, the mass flow rates in the mastering factors, especially at the low mass galaxies, is a very high in the simulations, these are traditionally 09:00:07 very low in the data, are we measuring these things accurately observational, are we comparing these things correctly between the simulations and the data. 09:00:33 regimes and these important regimes can include, I would claim, the Milky Way Halo mass tend to the 12 itself. How flow driving feedback does not mean supernova. 09:00:36 There are other physics of play, and indeed the energetics of the supermassive black holes, change the entire outflow galaxy connection, which put in is not what you get out, we have the signatures for instance in the column nation, and in the divergence 09:00:52 of wind at how flow properties as they are Hydra dynamically evolved through the Halo, basically, which leads to many complex behaviors and these kinds of cosmological simulations. 09:01:06 And finally, that the angular variation, which I think is something we should think about more angular variation of certain galactic medium properties have the ability to nicely prob any aspects of outflows. 09:01:18 So, and just by flashing again these two very high dimensional functions, so to speak, and to propose that we work and think about how to measure quantitatively, how much gas is moving as a function of all these physical properties and these galaxy properties 09:01:36 both in simulations and empirically. 09:01:39 Thank you very much. 09:01:44 Thank you very much done. Excellent, excellent presentation really covered a lot of ground there and a lot of discussion in the Slack channel to follow up for our panel. 09:01:57 I will give everyone, let's let's read commence. It's 902. Now, how about we take an eight minute break, to allow everyone to get a drink of water use the bathroom and check over the slack and go through the comments, there's a lot of discussion there. 09:02:14 I will draw from that for for the discussion panel and. And so people can add to it or, you know, thumbs up things that they think are relevant comments or thumbs down ones we had a thumbs down during the talk so controversial points, but this is great 09:02:39 for discussion so we'll see as you see everybody in about seven eight minutes. And again thanks Dylan, excellent Talk. Thank you.