1 00:00:00,210 --> 00:00:10,790 So it's my very great pleasure to introduce Professor James Binney to give us the 10th definition Memorial Lecture Lecture. 2 00:00:11,520 --> 00:00:15,300 He is sponsored by the Department of Physics and by also his college. 3 00:00:17,730 --> 00:00:23,430 And so, James, I know James for a very long time since I was a graduate student. 4 00:00:23,430 --> 00:00:29,640 In fact, when Dennis Sharman proudly introduced me to the rare visits, then to Oxford. 5 00:00:30,990 --> 00:00:36,390 And meeting James, even at that young age, it was clear he had a brilliant future in front of him. 6 00:00:37,140 --> 00:00:45,840 And it's quite remarkable to look at the many fields that he has worked in. 7 00:00:50,250 --> 00:00:57,630 These include everything from stars to galaxies to galaxy clusters. 8 00:00:57,900 --> 00:01:06,330 Cosmology. So he's made a major imprint in astrophysics, areas of physics, galaxy dynamics. 9 00:01:07,770 --> 00:01:14,010 But more than that, his books textbooks have become legion in the field. 10 00:01:15,240 --> 00:01:23,879 I think books like Binney and Tremayne are known to just about every astrophysics graduate student throughout the world. 11 00:01:23,880 --> 00:01:32,100 I would say has, as you know, the the handbook for commencing their research and to have, you know, 12 00:01:32,700 --> 00:01:40,050 combined both writing these major books along with a brilliant research career, I think is a wonderful achievement. 13 00:01:40,770 --> 00:01:55,680 So James has, after his Ph.D. in Oxford, he spent a postdoctoral stint at Princeton and then was on the faculty for a couple of years. 14 00:01:56,160 --> 00:02:11,430 And then in 1981, he moved to Oxford as a university lecturer and has been here ever since and has been professor of physics since 1996. 15 00:02:12,060 --> 00:02:17,070 And ever since he's been Oxford has been a fellow of American college. 16 00:02:18,660 --> 00:02:23,670 So his his various honours include being a fellow of the Royal Society. 17 00:02:25,890 --> 00:02:35,370 So before and also the Maxwell Prize and being true to physics and other prizes from the American Stroke Society. 18 00:02:36,000 --> 00:02:44,940 So before I introduce James, let me just mention that he was a student of Dennis Sharma, this lecture series, 19 00:02:44,940 --> 00:02:53,910 this is the 10th lecture, annual lecture, and the series is in honour of Dennis's brilliant success in generating. 20 00:02:55,770 --> 00:03:06,330 Students have gone on to really wonderful things in the world of astrophysics and in related disciplines and in other areas too. 21 00:03:06,990 --> 00:03:13,320 And so it's a very great pleasure that I introduce James to tell us about galaxies and the intergalactic medium. 22 00:03:14,190 --> 00:03:32,159 Thank you, Joe. So it's just about 42 years since I came to work with Dennis as his part of this second generation of graduate students. 23 00:03:32,160 --> 00:03:35,910 And just in the three years that I was in Dennis's hut, as it was called, 24 00:03:35,910 --> 00:03:43,320 it was a prefabricated building provided by the Radcliffe Trust that has since been knocked down and replaced by some part of physiology. 25 00:03:45,270 --> 00:03:53,280 There were five people I mean, there were six people in that hut who went on to have already gone gone on to be fellows of the Royal Society. 26 00:03:54,000 --> 00:04:01,320 And that, I think, is stupid. He had an amazing record as a research supervisor in Cambridge Martin Rees, 27 00:04:02,190 --> 00:04:08,100 Stephen Hawking and so on were his students and this is beginning of his Oxford collection. 28 00:04:08,880 --> 00:04:15,480 What made Dennis such an amazing the many factors obviously made Dennis an amazing supervisor, director of research. 29 00:04:15,960 --> 00:04:19,740 He had an infectious enthusiasm, a very wide knowledge of physics, 30 00:04:19,740 --> 00:04:27,059 a love of everything to do with physics, an enormously generous and outgoing and stuffy character. 31 00:04:27,060 --> 00:04:31,290 So a talent for friendship and tremendous loyalty to the people who'd worked with him. 32 00:04:32,550 --> 00:04:35,879 But he also had, I think, good taste in his choice of problems. 33 00:04:35,880 --> 00:04:41,550 And he was interested in quantum gravity long before anybody much was interested, or very few people were interested in quantum gravity. 34 00:04:42,180 --> 00:04:49,530 And the topic of my talk today, I felt I had to talk about this topic because it was something that Dennis was working on. 35 00:04:49,530 --> 00:04:59,759 As this nature paper indicates, when I when I came to join his group in 1972, so this was with Roland Hunt, 36 00:04:59,760 --> 00:05:03,720 who had been his research student in Cambridge and was a post-doc in Oxford at that time. 37 00:05:06,030 --> 00:05:13,559 So here's a sort of outline of what I want to talk about. I want to know about the cosmic baryon budget and star formation, 38 00:05:13,560 --> 00:05:16,680 what we know about something that's a little bit of what we know about star formation, 39 00:05:17,850 --> 00:05:26,340 something about the phenomenon of extra plane, ah1 that's neutral hydrogen, that doesn't lie in the equatorial plains of just galaxies like our own. 40 00:05:26,700 --> 00:05:35,280 And about a fountain model that can explain that too, has been built to explain that and does so well, 41 00:05:35,520 --> 00:05:41,669 and then talk about the role that the exchange of gas between galaxies, 42 00:05:41,670 --> 00:05:49,350 the disk, the star forming disks of galaxies like our own in the intergalactic medium plays in the big picture of galaxy evolution and cosmology. 43 00:05:50,580 --> 00:05:52,800 Of course, in science. What? 44 00:05:52,950 --> 00:05:59,489 What we achieve, what's done, what's worthwhile is largely it's very largely done by one's collaborators, is always done with collaborators. 45 00:05:59,490 --> 00:06:09,629 And what I'm going to talk about today is a programme of work which started when Philip Alfreton early was came to Oxford as a marie Curie fellow, 46 00:06:09,630 --> 00:06:11,460 and we've worked together on this programme. 47 00:06:11,970 --> 00:06:20,880 Well, we, I say, but largely Filippo has with his research students in Bologna because he's returned to Bologna and is on the faculty there. 48 00:06:21,600 --> 00:06:29,910 But I'm also going to mention so influential in this has been work that Colin Potty who's also on the faculty in Bologna has done with me and to. 49 00:06:30,090 --> 00:06:36,420 And towards the end I'll mentioned stuff ideas which emerged out of work that I did with two students, 50 00:06:36,430 --> 00:06:40,320 Gavin de Boer nearly 20 years ago and Henri Goma about ten years ago. 51 00:06:42,000 --> 00:06:50,220 So the Baryon budget well from the cosmic microwave background analysis of the fluctuations on the sky in the microwave background, 52 00:06:50,520 --> 00:06:58,919 we know what the contribution is of baryons to the to the overall mass budget, 53 00:06:58,920 --> 00:07:03,120 the overall energy budget of the universe and what the contribution from dark matter is. 54 00:07:04,560 --> 00:07:12,480 So in, in cosmology people use this symbol. Omega was introduced, I think by the Russians to mean the ratio of the density contributed. 55 00:07:12,480 --> 00:07:17,190 The average density contributed by some component to the closure density of the universe, 56 00:07:17,190 --> 00:07:23,460 the critical density that would, in the absence of cosmological constant lead to the expansion drifting to a halt. 57 00:07:24,570 --> 00:07:30,990 So an edge here is the Hubble constant in units of 100 kilometres a second per megaparsec. 58 00:07:31,770 --> 00:07:41,579 So h squared is essentially a half, just call it a half. So the point is that Baryons contribute maybe four or 5% of the closure energy density of 59 00:07:41,580 --> 00:07:47,180 the universe and dark matter contributes maybe 22% of the closure density of the universe, 60 00:07:47,190 --> 00:07:55,979 something like that. In other words, there are about five times as much of dark matter is of light matter from Galaxy Redshift surveys, 61 00:07:55,980 --> 00:07:58,260 particularly the Sloan Digital Sky Survey. 62 00:07:58,260 --> 00:08:12,120 The study says we know what the how much luminosity in some band the band is emitted on average per unit of matter density. 63 00:08:12,360 --> 00:08:18,810 The answer is about 220 solar masses of matter density are needed to produce one solar luminosity of radiation. 64 00:08:19,920 --> 00:08:27,719 But given that most that there's five times as much dark matter as there is baryonic matter, 65 00:08:27,720 --> 00:08:34,049 that means that there are something like 40 solar masses of ordinary matter required to 66 00:08:34,050 --> 00:08:38,250 produce one solar luminosity of radiation on the average and the universe as a whole. 67 00:08:38,910 --> 00:08:50,639 If you count stars in the Milky Way or you do population synthesis, let's say you do construct in a computer you stellar models, 68 00:08:50,640 --> 00:08:58,050 you you you make models of stars of every different size and appropriate metallicity and ages and so on. 69 00:08:59,580 --> 00:09:05,340 You can you can come up with some kind of estimate of how much mass in stars you need to emit. 70 00:09:05,340 --> 00:09:11,040 From a typical population of stars. You need to produce the solar luminosity of radiation. 71 00:09:11,460 --> 00:09:16,680 And the answer is numbers that vary but are around around three. 72 00:09:16,710 --> 00:09:22,230 Here is the actual plot of that. It's the, it's the, it's the mass required. 73 00:09:23,550 --> 00:09:30,510 Uh, sort this colour down here and this is the mass light ratios is this massive rate of luminosity. 74 00:09:30,510 --> 00:09:35,340 This is slightly different band, the g band used by the SS, but it makes little difference. 75 00:09:35,550 --> 00:09:40,770 And the individual galaxies the dots here and you can see this is a function of colour. 76 00:09:40,770 --> 00:09:47,250 Red galaxies need rather more mass per unit luminosity blue galaxies need rather less mass per unit luminosity, 77 00:09:47,550 --> 00:09:51,480 but your typical value is the square root of ten i.e. three. 78 00:09:52,680 --> 00:09:57,960 So that's, that's about how much mass you need to produce luminosity once you get going with some collection of stars. 79 00:09:58,620 --> 00:10:04,980 But actually in the universe as a whole, as I've said, we know that the the average number is something like 40, 80 00:10:05,580 --> 00:10:11,489 even though in individual galaxies it's it's more like three. And in the Milky Way, we can do much more exact work. 81 00:10:11,490 --> 00:10:17,610 And it's currently something that my group is working on quite vigorously at the moment. 82 00:10:18,240 --> 00:10:24,180 We can say that the mass in in in baryons is less than something like five by 83 00:10:24,180 --> 00:10:28,980 ten to the ten solar masses and the total mass is something like ten to the 12. 84 00:10:29,250 --> 00:10:36,989 So the, it's a very small portion of the galaxy which is made up of baryons and is producing the relevant luminosity. 85 00:10:36,990 --> 00:10:43,110 So there's a really big problem of the missing baryons in the sense that there are 4% of luminosity, 86 00:10:43,110 --> 00:10:47,670 there are something like 40 baryons out there, the microwave background and galaxy redshift surveys tell us. 87 00:10:48,570 --> 00:10:52,020 But actually in the galaxies there are numbers more like three. 88 00:10:52,020 --> 00:10:54,240 There's a there's a very large factor of missing burials. 89 00:10:55,770 --> 00:11:02,729 When in galaxy clusters, which are the largest relaxed structures in the universe, the missing barrier islands are not missing. 90 00:11:02,730 --> 00:11:08,840 The barriers that are not in galaxies and not in stars are present and correct and have been and been known for a long time. 91 00:11:08,850 --> 00:11:13,409 So enrich here is the coma cluster, one of the richest, most magnificent clusters of galaxies. 92 00:11:13,410 --> 00:11:18,540 These are all these fuzzy things are all galaxies, two really huge galaxies, a load of satellites around them and so on. 93 00:11:18,810 --> 00:11:24,450 And this space around the galaxies is filled with plasma at the variable temperature, 94 00:11:24,450 --> 00:11:30,630 at the temperature at which there's enough pressure for the gas to resist the enormous gravitational field of the whole cluster of galaxies. 95 00:11:31,710 --> 00:11:41,730 And in if you look in X-rays so this is from ROSAT, quite an ancient German satellite, you see a surface brightness distribution like this. 96 00:11:41,730 --> 00:11:48,360 If you look in Planck, you look at the influence which the fast moving electrons, 97 00:11:48,360 --> 00:11:54,780 thermal electrons associated with all this hot gas have on the cosmic microwave background photons. 98 00:11:55,020 --> 00:12:03,540 You can detect the gas and get a plot like this. So we see this gas and it has an enormous mass. 99 00:12:03,540 --> 00:12:09,510 It has a mass which is many times several times the mass of what's in the galaxies, just like cosmology says it should be. 100 00:12:11,430 --> 00:12:15,270 A. Yeah, I think I've said those things. 101 00:12:15,990 --> 00:12:24,600 And another very important fact is that this plasma that lies between the galaxies emits spectral lines associated with iron, 102 00:12:24,600 --> 00:12:30,570 silicon and other elements that still have some bound electron at these enormously high temperatures, 103 00:12:30,900 --> 00:12:35,010 the temperatures being ten to the 7 to 1038 degrees Kelvin, 104 00:12:36,540 --> 00:12:38,609 which when you do the accounting on this, 105 00:12:38,610 --> 00:12:48,540 you conclude that on the order of half of the heavy elements that that meaning everything from carbon oxygen right up to ion uranium, 106 00:12:48,540 --> 00:12:56,700 whatever half of the elements are in hydrogen helium which have been synthesised by stars, which are presumably more or less in these galaxies don't. 107 00:12:56,910 --> 00:13:02,610 These heavy elements are not in the galaxies. They're outside the galaxy, the galaxies in this intergalactic medium. 108 00:13:04,740 --> 00:13:07,880 Right. So what do we. Sorry. 109 00:13:07,930 --> 00:13:08,830 Backwards. Backwards. 110 00:13:09,850 --> 00:13:18,610 If you look in groups of galaxies, so you go from an enormous cluster like the coma cluster down to smaller groups of galaxies and do the same thing. 111 00:13:19,870 --> 00:13:23,190 Surveying with X-ray. You find an interesting plot. 112 00:13:23,200 --> 00:13:29,450 This is from the group of Trevor Polman in Cambridge sorry in Birmingham as a function of temperature along here. 113 00:13:29,470 --> 00:13:34,720 So ten KV so ten to the seven K up there, ten to the six K or so down here. 114 00:13:34,930 --> 00:13:43,720 Logarithmic scale. You have different groups and estimates of the fraction of the mass of the group which is contained in the intergalactic gas. 115 00:13:44,440 --> 00:13:50,200 And there is clearly a tendency for this fraction to increase as you go to higher temperature. 116 00:13:50,200 --> 00:13:54,969 And that in practice means more massive groups which have stronger gravitational potential. 117 00:13:54,970 --> 00:14:01,420 Well, that's why the temperature is higher. The variable temperature is higher because you need more kinetic energy in the in the 118 00:14:01,420 --> 00:14:08,170 in the plasma in order to to be extended in the gravitational field of the system. 119 00:14:09,190 --> 00:14:13,509 So. So there are two conclusions from this kind of study. 120 00:14:13,510 --> 00:14:23,650 One is that a smaller fraction of the of the mass of the system is contained in intergalactic gas as you go to lower mass clusters groups. 121 00:14:25,240 --> 00:14:30,490 And also that the another important fact which is not evident from this diagram, 122 00:14:30,730 --> 00:14:38,830 is that the central entropy, the the the heat oddness of the of the gas sits at the centre of the group, 123 00:14:41,200 --> 00:14:46,779 lies more and more above the value that you would estimate if this plasma have been heated 124 00:14:46,780 --> 00:14:52,900 simply by falling downhill into the deep into the gravitational potential well of the group. 125 00:14:53,170 --> 00:15:01,360 So in the simplest picture, you imagine that there's gravitational clustering of all this dark matter, and so gravitational potential will arise. 126 00:15:01,570 --> 00:15:13,120 And the baryons, the the ordinary gas falls in there and gets compressed and shock heated when it gets as it's brought to a halt by other gas, 127 00:15:13,120 --> 00:15:16,060 which is already taking up the space in the middle of the group. 128 00:15:16,630 --> 00:15:25,510 So from that model you can compute how what the entropy of the gas would be, which is being heated in this way at the middle of the group. 129 00:15:25,960 --> 00:15:31,150 And the X-ray observations indicate a discrepancy between that value of the entropy 130 00:15:31,150 --> 00:15:35,590 and the entropy you actually see in this gas in the middle of these groups, 131 00:15:35,800 --> 00:15:41,709 which increases becomes larger and larger as you go to lower and lower temperatures so that there's 132 00:15:41,710 --> 00:15:48,880 evidence what that what that what that the interpretations are here that the intergalactic medium is not, 133 00:15:48,880 --> 00:15:54,670 in fact, only heated by stuff falling down into the gravitational potential wells of these, 134 00:15:54,670 --> 00:16:01,150 of these, uh, gatherings of these, these, these gatherings of galaxies and dark matter. 135 00:16:01,450 --> 00:16:13,420 It's also heated by the energy, by the energy of ultimately of thermonuclear origin, which is with which the gas is flung violently out of galaxies. 136 00:16:15,310 --> 00:16:21,520 And so we've had two pieces of evidence that that heating by gas being flung out of galaxies is important. 137 00:16:21,520 --> 00:16:30,020 One is that the heavy elements produced in the galaxies, in clusters like coma to a large measure, 138 00:16:30,040 --> 00:16:35,260 not any longer in the galaxies, but are in the intergalactic gas outside the galaxies, they must have been pushed out. 139 00:16:35,710 --> 00:16:43,240 And the other is this indication that that when you're looking at relatively low mass clustering of galaxies, 140 00:16:43,240 --> 00:16:47,500 which wouldn't, by gravitational landfall, heat the gas to a very high temperature, 141 00:16:47,800 --> 00:16:53,379 you can notice that the actual temperature of the gas is higher than it would be if it was only being gravitationally heated. 142 00:16:53,380 --> 00:16:56,590 And you can infer that it's also, as it were, from a nuclear heated. 143 00:16:58,250 --> 00:17:05,690 So the key points of my argument so far are that forming galaxies have ejected large quantities of gas and 144 00:17:05,690 --> 00:17:13,490 that most of the baryons that were are in the universe actually lie between galaxies and not in galaxies. 145 00:17:15,950 --> 00:17:18,290 So something a little bit about star formation. 146 00:17:19,820 --> 00:17:29,120 So the start, these diagrams and others show that the star formation rate is very strongly correlated with the surface density of cool gas. 147 00:17:29,150 --> 00:17:32,630 It's not exactly surprising since stars have to form out of cool gas. 148 00:17:33,080 --> 00:17:42,700 But what these diagrams show. So this is this is this is from this paper, from orientale as a function of galactic centric radius. 149 00:17:42,710 --> 00:17:46,850 So this is a fraction this is our 25, which is a very large radius, 150 00:17:46,850 --> 00:17:52,550 is the radius of which the surface brightness of the galaxy is 25 magnitudes per square second. 151 00:17:52,850 --> 00:17:58,910 In sum, in some bands, probably the blue band. The point is that it's a very large radius. 152 00:17:59,270 --> 00:18:06,080 So as you go, what you're doing is you're going here as you're going out and you're going out in a lot of different galaxies. 153 00:18:06,740 --> 00:18:12,950 And what you've got up here are logarithmic scale is the star formation efficiency. 154 00:18:13,220 --> 00:18:18,590 The star formation efficiency is defined to be the surface density of star formation. 155 00:18:18,590 --> 00:18:30,920 That's the number of stars formed per unit area per unit time divided by the surface density of gas, obviously the amount mass of gas per unit area. 156 00:18:31,430 --> 00:18:36,770 So this the star formation efficiency has units of one over time. 157 00:18:37,670 --> 00:18:45,829 And you can see that it's it's fairly constant in the inner regions and then tails off and the different colours 158 00:18:45,830 --> 00:18:53,150 here indicate that this is a region where we're talking about where most of the gas is in molecular form, 159 00:18:53,540 --> 00:18:58,070 whereas the further out most of the gas is in atomic form. 160 00:18:59,000 --> 00:19:03,830 So this is already an indication that it's the density of molecular gas that's important. 161 00:19:05,990 --> 00:19:09,620 And this is in. These are these are small, low mass galaxies. 162 00:19:09,620 --> 00:19:14,809 These are large mass galaxies. And in low mass galaxies, there's much less molecular gas. 163 00:19:14,810 --> 00:19:16,940 Much less of the gas is molecular. The important point, 164 00:19:17,360 --> 00:19:27,350 one important point is that is that the efficiency is fairly constant in the region that's dominated by molecular gas rather than by atomic gas. 165 00:19:28,760 --> 00:19:39,710 So that's a relatively central, more dense region. And and the other point is the numerical value that the inverse of this. 166 00:19:41,120 --> 00:19:46,520 This thing here, which is the time if the star formation continued, at this rate, 167 00:19:46,700 --> 00:19:52,430 it would use up that gas in the time, which is the inverse of this star formation efficiency. 168 00:19:52,670 --> 00:20:00,230 So it would use up the available gas according to that in a in three, four giga years in less than a Hubble time. 169 00:20:01,400 --> 00:20:04,970 Now, actually, as we've said, the gap the box isn't closed. 170 00:20:05,150 --> 00:20:08,690 Gas star formation definitely pushes gas out of galaxies. 171 00:20:08,690 --> 00:20:12,410 There's a great deal. I've spoken of a little evidence of that, but there's much other evidence that. 172 00:20:13,950 --> 00:20:17,850 So one conclusion of this is perhaps we should just go over. 173 00:20:18,510 --> 00:20:24,180 Sorry, I keep going backwards is look at this diagram here which shows the staff 174 00:20:24,180 --> 00:20:30,419 measured efficiency in terms of just using molecular gas and then it's really, 175 00:20:30,420 --> 00:20:33,659 really flat as a function of a narrow range. 176 00:20:33,660 --> 00:20:42,899 It's true. It lacks a centric radius, so the existing gas will be exhausted in 2 to 3 Giga Years because even if you didn't blow any gas out, 177 00:20:42,900 --> 00:20:49,320 you'd run out at present rates of star formation in in four or five Giga Years. 178 00:20:51,390 --> 00:21:00,330 And also, you would indicate that if if there was if what if the way galaxies worked was they started off with a fixed mass of gas and they 179 00:21:00,330 --> 00:21:07,560 started to turn them into into stars at a rate which depended upon the which was proportional to the surface density of gas. 180 00:21:08,160 --> 00:21:12,900 They would have had much higher surface densities of gas in the past and much higher star formation rates. 181 00:21:13,260 --> 00:21:18,000 And we really would be at very much at the end of the star formation process in galaxies like ours. 182 00:21:18,840 --> 00:21:25,919 So the star formation rate in the past would have been much higher if if the galaxy started with a fixed supply of gas, 183 00:21:25,920 --> 00:21:34,340 which they then steadily turned into stars. But if you look at the mix of stars near the sun, 184 00:21:36,950 --> 00:21:44,480 you can you can form an opinion about how many how how many relatively young stars there are vis a vis how many old stars there are. 185 00:21:44,780 --> 00:21:50,809 You conclude that the star formation rate has declined by maybe a factor of three, something like that, 186 00:21:50,810 --> 00:21:56,300 only in the entire ten Giga Years or so that it's taken to form the disk of our galaxy. 187 00:21:57,490 --> 00:22:05,450 So. So I think the bottom line from all of this is that galaxies don't start off with a fixed supply of gas. 188 00:22:06,260 --> 00:22:11,870 They start off with some gas. They form some stars, they eject some gas and they accrete some gas. 189 00:22:11,870 --> 00:22:18,650 That this exchange two ways is an important both the they have to both accrete gas to sustain star formation. 190 00:22:18,890 --> 00:22:24,110 Well if they do it Greek gas to sustain star formation and this keeps their 191 00:22:24,410 --> 00:22:31,420 their gas densities so that the surface density of gas is falling only slowly. 192 00:22:31,430 --> 00:22:41,240 Therefore, the therefore the level of star formation rate is falling only slowly because they almost balance their income 193 00:22:41,420 --> 00:22:50,090 coming from accretion and their expenditure locking gas up in stars and ejecting gas into the intergalactic medium. 194 00:22:51,800 --> 00:22:55,430 So now what to do about the phenomenon of extra plain a H1. 195 00:22:56,000 --> 00:23:04,700 So this is a galaxy that's really fairly similar to our own galaxy, NGC 891, which we happen to see almost perfectly edge on. 196 00:23:04,710 --> 00:23:10,720 So it's the classic galaxy to study. If you want to study an edge on galaxy, it's nearby. 197 00:23:10,730 --> 00:23:15,530 It's very similar to the Milky Way. It has a slightly more vigorous star formation than the Milky Way. 198 00:23:16,670 --> 00:23:20,560 And. This is these. 199 00:23:20,650 --> 00:23:30,460 These are contours of the surface density of neutral hydrogen seen in the 21 centimetre hyper fine line of hydrogen as measured in 1979. 200 00:23:31,090 --> 00:23:36,400 Now, of course, as the electronic engineers have gone on. They have got more and more sensitive. 201 00:23:37,390 --> 00:23:44,470 They built more and more sensitive radio telescopes. And so the the extent people. 202 00:23:45,310 --> 00:23:52,750 This is all done in Westerbork in the Netherlands, in fact. The the the limiting surface density has gone from five by 20, 203 00:23:52,750 --> 00:24:00,850 the ten to the 20 atoms per square centimetre through seven by ten to the 19 is that through 1.7 times ten to the 19. 204 00:24:01,030 --> 00:24:07,960 So there's been a drop by almost no magnitude and a half in the surface density that can be detected. 205 00:24:08,140 --> 00:24:16,000 And in that and in that technical progress, what has happened is that this galaxy has grown an h one halo. 206 00:24:16,390 --> 00:24:24,580 It's it's we're seeing here gas, which is not in the plane of the galaxy. 207 00:24:24,610 --> 00:24:29,769 Now, you might worry that what you're seeing is a disk of gas rotating around the galaxy, 208 00:24:29,770 --> 00:24:37,089 which isn't seen edge on which you can rule out that possibility based on the kinematics of this gas, because this gas is detected in line radiation. 209 00:24:37,090 --> 00:24:41,260 So you know very precisely how fast it's moving up and down the line of sight. 210 00:24:41,560 --> 00:24:45,310 And that kinematic information tells you that this gas is in fact, 211 00:24:46,570 --> 00:24:52,930 this galaxy is enveloped in gas which lies well off its galactic plane, which you see here. 212 00:24:52,930 --> 00:25:01,000 Edge on the Galaxy has grown in this direction as a result of the increased sensitivity, much more than it's grown in this direction. 213 00:25:03,160 --> 00:25:07,510 Of course, it's growing in every direction, but it really has been growing away from the plane. 214 00:25:07,840 --> 00:25:16,720 So this galaxy like this keeps something like a quarter of its supply of neutral hydrogen, more than a killer parsec off its disk. 215 00:25:16,960 --> 00:25:23,590 That's an amazing feat, because the temperature, the characteristic temperature of neutral hydrogen is less than ten to the four degrees K. 216 00:25:23,860 --> 00:25:28,569 The variable temperature of this gravitational potential well is over a million degrees K. 217 00:25:28,570 --> 00:25:38,380 So this gas has less than a 10th, less than 100th of the temperature required for pressure to to resist, 218 00:25:38,440 --> 00:25:43,120 to be in effect, to provide effective resistance to the gravitational field of this galaxy. 219 00:25:43,120 --> 00:25:48,010 And yet it keeps this gas up there. And that's what I want to talk about next. 220 00:25:49,770 --> 00:25:56,280 So this is another galaxy, a galaxy seen face on, also seen from Westerbork in the Netherlands. 221 00:25:56,520 --> 00:26:00,750 This is what this galaxy, NGC 6946 looks like in the optical. 222 00:26:02,190 --> 00:26:10,530 This is what it looks like in 21 centimetres. So it's one of these galaxies which has a very huge extended disk of neutral hydrogen. 223 00:26:12,990 --> 00:26:16,530 And if you look in this disk, 224 00:26:16,530 --> 00:26:20,639 there are various bright spots where there's lots of 21 centimetre emission and there 225 00:26:20,640 --> 00:26:25,890 are a few very small dark spots where there's very little 21 centimetre emission. 226 00:26:25,890 --> 00:26:34,290 And if you look at the distribution of neutral hydrogen along this line, you get you see a plot like this. 227 00:26:34,290 --> 00:26:40,710 This is from blacksmith's theses. So what's plotted here is line of sight velocity. 228 00:26:40,860 --> 00:26:49,230 So the point about 21 centimetre measurements or measurements in a in a in a spectral line of gas like 21 centimetre is that it generates a data cube. 229 00:26:49,500 --> 00:26:57,300 You have two dimensions on the sky and then you have a frequency dimension which you interpret as a velocity one because Doppler shifts and all that, 230 00:26:57,690 --> 00:26:59,490 because you're looking at a particular spectral line. 231 00:26:59,760 --> 00:27:06,210 So as you scan along this along this line, that means moving along here from the middle of this line, 232 00:27:06,450 --> 00:27:10,890 you see emission at a velocity, which shifts because the galaxy is rotating. 233 00:27:11,250 --> 00:27:16,230 And you see places like this where there's an absence of gas, that's one of these holes. 234 00:27:17,160 --> 00:27:24,810 And then you see material here, which is at velocities, which are different from the velocities of gas that lies in the plane. 235 00:27:24,840 --> 00:27:34,020 So the interpretation here is that some cluster of supernovae occurring at this point in the disk. 236 00:27:34,230 --> 00:27:38,790 So somewhere here have gone off in a very short time interval. 237 00:27:38,790 --> 00:27:48,209 One after another created a blast wave which has swept up the the coolest gas and flung it off the disk. 238 00:27:48,210 --> 00:27:51,090 And here you're seeing some of it flag flung off the disk. 239 00:27:52,320 --> 00:28:00,060 So personally and I set out to model this process in the following with the following very naive model. 240 00:28:00,390 --> 00:28:08,130 So here is our disk. Here is some point on the disk where star formation takes place here. 241 00:28:08,280 --> 00:28:12,959 In the case of our own Milky Way, here is the sun and we were able to stir up. 242 00:28:12,960 --> 00:28:18,840 So we'll do some of that. But for the moment we're applying this model to NGC 891 and some other external galaxies. 243 00:28:19,110 --> 00:28:24,090 And in that case, the sun isn't here. The sun's over there somewhere or out here, out in the lecture room. 244 00:28:25,140 --> 00:28:30,240 So then. So supernovae go off here, they blast a cloud of gas upwards. 245 00:28:30,630 --> 00:28:40,560 Here is the cloud of cool gas moving upwards through an ambient atmosphere of temperature of gas, which is at the very low temperature of the galaxy. 246 00:28:40,800 --> 00:28:51,720 This cloud is rather cool and as it flies through the ambient medium, it there's a flow of ambient medium across its top. 247 00:28:51,990 --> 00:28:55,350 The ambient medium turbulent mix is in. I'll show you a simulation of this. 248 00:28:55,350 --> 00:28:56,910 This is just a cartoon. 249 00:28:57,570 --> 00:29:07,500 The flow it flows over this thing you you get mixing of the of the intergalactic medium and this cool thing of gas and you get a wake here forming. 250 00:29:08,370 --> 00:29:12,719 And I want to show you some evidence that this model works very well. 251 00:29:12,720 --> 00:29:19,260 So this is this is called a galactic fountain, not surprisingly, because the gas goes up, 252 00:29:20,310 --> 00:29:23,790 reaches some maximum height and falls back down again at some other radius. 253 00:29:24,540 --> 00:29:30,690 Some other point. So we we built a very simple model of this of this process. 254 00:29:31,660 --> 00:29:35,880 And what were the what were the headline conclusions from this? 255 00:29:36,450 --> 00:29:40,260 So what we were doing was we were making a model of the complete data cube. 256 00:29:41,250 --> 00:29:46,140 This is this is this is why it's a little bit difficult to show you the, uh, 257 00:29:46,590 --> 00:29:51,629 the way in which the degree to which the model fits the data because the data of three dimensional models. 258 00:29:51,630 --> 00:29:59,990 Three dimensional. And so it's, it's. You rely on numerical tools to tell you whether you're doing well or not. 259 00:30:00,000 --> 00:30:08,250 But anyway, the Cube's required that the h one clouds as they flew through the intergalactic medium, 260 00:30:08,250 --> 00:30:11,460 as they went on their passage up from the disk and back down to the disk. 261 00:30:11,760 --> 00:30:14,070 These clouds had to grow in mass. 262 00:30:15,400 --> 00:30:22,000 And that was not what we were expecting, because it's intuitively obvious that what happens is that these clouds suffer ablation, 263 00:30:22,000 --> 00:30:31,210 that on their surfaces material is swept off and and mixed in with the hot gas and lost to 21 centimetre observation. 264 00:30:32,860 --> 00:30:40,540 And because they grew in mass as they flew through the intergalactic medium, when they landed, 265 00:30:40,540 --> 00:30:45,970 they had more mass than they set off, which meant that as a result of this process, the disk was gaining mass. 266 00:30:46,360 --> 00:30:55,000 And how fast was it gaining mass? Very much at the same rate as the rate at which the one sex peaked as ready, 267 00:30:55,070 --> 00:30:59,500 which the observations indicate that gas was being used up in star formation. 268 00:31:01,600 --> 00:31:02,080 So. 269 00:31:03,700 --> 00:31:10,900 So that was the inclusion of those two studies of two galaxies, the galaxy I've shown you NGC eight, nine, one, and another galaxy which is down here. 270 00:31:11,260 --> 00:31:14,889 This is NGC 2403, which is a very different sort of galaxy from this. 271 00:31:14,890 --> 00:31:18,190 It's lower in mass, it's seen much less edge on. 272 00:31:18,400 --> 00:31:21,370 So the the nature of the observational material is very different. 273 00:31:21,580 --> 00:31:27,879 But one in the same model could make it would could do a very good account of both of the datasets, 274 00:31:27,880 --> 00:31:30,310 the data cubes belonging to both of these galaxies. 275 00:31:30,850 --> 00:31:45,280 Then Fred Phillips from Tenaris graduate student Federico marinucci simulated the hydrogen recall flow over of gas over such a gas cloud. 276 00:31:45,280 --> 00:31:54,280 So, so in our phenomenological model we just said the cloud moved through the intergalactic medium and it gains mass or it loses mass in some way. 277 00:31:55,510 --> 00:32:02,589 Parameterised And then we in this way fit it to some, to some, to some data. 278 00:32:02,590 --> 00:32:04,480 Cube We've got parameters of our model out. 279 00:32:04,720 --> 00:32:11,410 But, but, but Marinus you actually did a hydrogen chemical simulation of the flow of hot gas over cold gas. 280 00:32:11,860 --> 00:32:22,360 And that study led to the conclusion there's a critical line in the in the the metallicity pressure and t and is the number density t is 281 00:32:22,360 --> 00:32:30,490 the temperature is 20 is the pressure in appropriate units and f over h is a is the is stands for the metal abundance in the medium. 282 00:32:30,790 --> 00:32:40,000 So there's a in the in the Corona's plane of possible metal metal abundances and pressures. 283 00:32:40,270 --> 00:32:47,710 There is a plane above which when the metallicity is high and the pressure is high and all the pressure is high, 284 00:32:48,760 --> 00:32:51,850 the mass of cold gas increases as the cloud flies. 285 00:32:51,850 --> 00:32:58,330 And if you're below this critical surface, the mass of your clouds decreases as it flies through. 286 00:32:58,870 --> 00:33:02,439 The reason for this is simply that if you increase the metallicity, 287 00:33:02,440 --> 00:33:09,190 you increase the ability of the gas to cool through spectral lines of the heavy elements. 288 00:33:09,190 --> 00:33:13,510 Heavy elements are very good at enabling plasma to cool. 289 00:33:14,260 --> 00:33:18,520 And also if you increase the pressure, you increase the density. 290 00:33:18,520 --> 00:33:22,809 That also increases the rate at which cooling occurs because cooling occurs at a 291 00:33:22,810 --> 00:33:26,020 rate that's a rate per ASM that is proportional to the number density of atoms, 292 00:33:26,020 --> 00:33:31,780 the number of things there are for electrons to bump into. So here is some evidence for this. 293 00:33:31,780 --> 00:33:36,370 Here is one of Marinol, which is simulations of the cloud which started off spherical, 294 00:33:36,610 --> 00:33:43,089 has been flattened by the by the pressure as it ploughs into the into the coronal gas. 295 00:33:43,090 --> 00:33:46,329 That's the variable temperature gas up here. It's been flattened. 296 00:33:46,330 --> 00:33:49,900 And then this wake is turbulent, wake downstream is developed. 297 00:33:50,170 --> 00:33:53,650 And in this turbulent weight downstream, one of two things can happen. 298 00:33:54,010 --> 00:33:59,830 Well, what necessarily happens is that the cold gas mixes with the hot gas in all those vertical rings. 299 00:34:00,640 --> 00:34:03,850 Then the hot gas can possibly evaporate. 300 00:34:03,850 --> 00:34:12,790 The cold gas quickly raised the cold gas to high temperatures where cooling is inefficient or the cold gas can succeed in cooling the 301 00:34:12,790 --> 00:34:22,930 hot gas by mixing with it and and lower its temperatures to two to values at which it's dense enough that it cools very rapidly. 302 00:34:24,010 --> 00:34:28,719 And if you so if you if you study the mass of cold gas, 303 00:34:28,720 --> 00:34:34,180 this is the difference between the mass at time t in the mass originally in the cloud of cold gas. 304 00:34:34,780 --> 00:34:39,700 That's gas with a temperature of less than five by ten to the five k as a function of time. 305 00:34:39,940 --> 00:34:48,430 Then it declines like this. This is just a convergence study showing the result of using different numerical resolutions. 306 00:34:49,360 --> 00:34:52,929 Whereas if you if you take the higher the medium resolution, 307 00:34:52,930 --> 00:35:02,860 so the medium resolution simulation and do the computation with a more metal rich and a more dense coronal plasma, 308 00:35:03,070 --> 00:35:06,979 you find that the mass increases over time. So that's that critical. 309 00:35:06,980 --> 00:35:14,230 There's that critical thing of so flushed with success for four months. 310 00:35:14,290 --> 00:35:22,839 Standing what was happening in the in the explaining modelling successfully the data cubes of external galaxies. 311 00:35:22,840 --> 00:35:25,060 We turn to the data cube of our own galaxy, 312 00:35:25,810 --> 00:35:31,750 which is very different in nature from the data cubes of external galaxies, because we sit right inside our galaxy. 313 00:35:32,530 --> 00:35:36,130 And then when you look out, you have absolutely no idea. 314 00:35:36,310 --> 00:35:42,880 You see a mission. And as a function of angle, the direction you point your telescope at various velocities, 315 00:35:43,120 --> 00:35:47,050 but you have no idea where along the line of sight this emission lies. 316 00:35:47,500 --> 00:35:57,370 So the early 21 centimetre surveys around 1960 found that there was 21 centimetre emission coming from very high galactic latitudes, 317 00:35:57,670 --> 00:35:59,860 which was which surprised people at the time. 318 00:36:00,200 --> 00:36:06,969 And what's more, you saw emission happening at velocities that you simply couldn't explain if you thought in terms of gas, 319 00:36:06,970 --> 00:36:11,500 the cold gas in the galaxy rotating in a nice orderly flow around the galactic centre. 320 00:36:11,500 --> 00:36:15,340 That's the so-called forbidden gas or gas of seen it forbidden velocities. 321 00:36:15,790 --> 00:36:19,300 And the nature of this of the gas that gave rise to this emission, 322 00:36:20,290 --> 00:36:25,930 this high velocity emission, as it was called, was was fiercely debated for decades. 323 00:36:26,230 --> 00:36:37,420 So as late as 1999, Blix and CO created something of a stir by arguing that this gas was at megaparsec distances. 324 00:36:37,690 --> 00:36:46,209 And if this gas was at very large distances because it produces a certain amount of flux at the telescope, 325 00:36:46,210 --> 00:36:49,180 it had to have you needed an enormous mass of gas. 326 00:36:49,180 --> 00:36:55,900 The further away you put this gas to produce the given flux at your telescope that you measure the mass of the gas has to increase. 327 00:36:56,200 --> 00:37:04,450 So if this stuff was at megaparsec distances, it would have it would be present in quantities that were suitable for forming whole galaxies. 328 00:37:04,630 --> 00:37:08,680 Whereas if the gas is very close to you, we are talking about rather trace amounts of gas. 329 00:37:09,880 --> 00:37:12,130 So this was hotly debated for a very long time. 330 00:37:12,400 --> 00:37:19,240 But in the in the in the not in the naughties, as it were in the last decade, this question was finally resolved. 331 00:37:21,210 --> 00:37:23,700 That by ultraviolet absorption measurements. 332 00:37:24,780 --> 00:37:31,050 It turns out that this gas lies at distances of less than the killer particle, so some of it's much closer than that. 333 00:37:32,430 --> 00:37:36,510 And that means that it's in red. It represents relatively small amounts of gas. 334 00:37:38,270 --> 00:37:46,309 So it it was an interesting exercise to ask whether this model of extra playing a gas 335 00:37:46,310 --> 00:37:51,560 that we had developed for these external galaxies would could account for the data cube, 336 00:37:51,560 --> 00:37:57,470 this data cube for our own galaxy, which is very different in nature. 337 00:37:58,550 --> 00:38:06,890 And so that so another student of Antonio Morasco, another student of Fratantoni, 338 00:38:08,150 --> 00:38:19,160 fitted the model that we built for these external galaxies to the neutral hydrogen observed in the light, an Argentine bond survey. 339 00:38:19,160 --> 00:38:23,090 So a recent reworking of the data cubed of our own galaxy. 340 00:38:24,520 --> 00:38:34,270 And here are some diagrams. This is what we're seeing here is gas at some particular line of sight velocities, 341 00:38:34,270 --> 00:38:39,230 this gas that's coming towards us at about 66 kilometres a second in sum. 342 00:38:40,120 --> 00:38:45,070 And here we have galactic longitude. So that's that's angled around the galactic plane. 343 00:38:45,310 --> 00:38:48,940 And here we have galactic latitude, 20 degrees, 40 degrees, etcetera. 344 00:38:48,950 --> 00:38:54,520 So this is quite a big piece of the sky in some sense. This is the data. 345 00:38:54,730 --> 00:38:58,570 This is this is what you actually see. And this is the model. 346 00:38:58,750 --> 00:39:02,740 This down here is, broadly speaking, the allowed gas. 347 00:39:03,140 --> 00:39:08,380 This is is gas belonging to that that we interpret as being long belonging to the galactic fountain. 348 00:39:09,340 --> 00:39:17,370 So, of course, it's not a perfect match here because this fountain involves individuals, groups of supernovae which send out particular plumes of gas. 349 00:39:17,380 --> 00:39:22,180 It's a noisy process you can't expect to get. You can only talk about the average. 350 00:39:22,180 --> 00:39:30,610 If we would look at this at some 200, 300 mega years later, this plot would be rather different. 351 00:39:31,390 --> 00:39:33,760 But I think this is a reasonable representation of it. 352 00:39:34,420 --> 00:39:38,830 An indication, though, that we have got the physics right is, I think, given by these pictures up here. 353 00:39:39,220 --> 00:39:43,750 What these are showing is the. The. 354 00:39:44,740 --> 00:39:48,490 Well, this is showing the court. 355 00:39:48,550 --> 00:39:53,250 Read it from here. Yet. 356 00:39:55,980 --> 00:39:59,580 Sorry. Yeah. The fraction of condense mass and the plot. 357 00:39:59,590 --> 00:40:07,350 Exactly. So I want the bottoms of the bottoms mass. So this is showing for this issue. 358 00:40:07,740 --> 00:40:11,340 So you fit this model to the to the data cube. 359 00:40:11,700 --> 00:40:15,329 The model is a phenomenological model and it involves things like how rapidly does a 360 00:40:15,330 --> 00:40:20,399 cloud gain mass as it sweeps through the intergalactic medium or through this corona, 361 00:40:20,400 --> 00:40:22,530 all this hot gas, very low temperature gas? 362 00:40:22,920 --> 00:40:29,550 How how much does it slow down as it moves through the medium because it's sweeping up material, etc., etc., etc. 363 00:40:29,970 --> 00:40:39,240 So you, you the the, the phenomenological model fitted that provides this fit to the data route produces these 364 00:40:39,240 --> 00:40:45,570 lines because numbers come out of it which enable you to produce those straight lines from, 365 00:40:46,020 --> 00:40:50,489 from Mariana, which is hydrodynamic simulation, which I showed you earlier. 366 00:40:50,490 --> 00:40:55,230 This one one's like this. Here you get the data points. 367 00:40:57,260 --> 00:41:02,060 And these are two enormously different computations. 368 00:41:02,990 --> 00:41:04,549 One is, 369 00:41:04,550 --> 00:41:15,200 is the mariner actually calculation involves numerical hydrodynamics and detail cooling functions with atomic physics and all that sort of thing. 370 00:41:15,560 --> 00:41:26,450 The other thing is a phenomenological model fitted to a data cube produced by ready observers and is completely ignorant of atomic physics or a. 371 00:41:29,020 --> 00:41:31,900 Yeah, that kind of thing. And hydrogen and hydrogen, that mix. 372 00:41:32,590 --> 00:41:37,120 And yet we get this degree of agreement, and this, I think, is an indication that we're doing the right thing. 373 00:41:37,300 --> 00:41:41,440 Another indication that we're doing the right thing is you can you can say. 374 00:41:42,190 --> 00:41:46,360 So let's go back to that cartoon. This cartoon. 375 00:41:46,810 --> 00:41:53,290 So when you if you are line of sight to some some ultraviolet emitting source up here, 376 00:41:53,290 --> 00:42:00,220 which might be a quasar or might be some some very hot star far out in the galaxy. 377 00:42:01,030 --> 00:42:11,649 If your line of sight passes through this turbulent wake, you should be able to see absorption lines because this turbulent wake contains 378 00:42:11,650 --> 00:42:16,990 gas that on the order of ten to the fifth K and it will contain ions of carbon, 379 00:42:17,680 --> 00:42:24,250 oxygen, silicon, etc. which which have characteristic absorption lines and when. 380 00:42:24,950 --> 00:42:32,350 And you will have a very sensitive probe, you'll be very sensitive to quite small amounts of gas containing these, these things. 381 00:42:33,040 --> 00:42:46,550 So. And indeed, when you look out to these ultraviolet sources with space telescope, you see that that's the data points here. 382 00:42:47,360 --> 00:42:51,660 Could you move away from the sorry, could you move away from the projector? 383 00:42:51,680 --> 00:42:58,069 Yeah. You see, so these these dots here are actual observations. 384 00:42:58,070 --> 00:43:06,180 These are lines of sight down to quasars or ultraviolet emitting stars, which you can get absorption spectra with HST. 385 00:43:06,200 --> 00:43:10,070 This is the work. Work of Lana et al. Yeah, there we are. 386 00:43:10,100 --> 00:43:18,259 There. And the colour indicates the, the, the velocity of a kind of an absorption system. 387 00:43:18,260 --> 00:43:18,470 Right. 388 00:43:18,480 --> 00:43:28,160 This is, the colours indicate velocity from from minus to the deep blue is -200 kilometres a second to two red is is plus 200 kilometres a second. 389 00:43:28,520 --> 00:43:34,249 And the that's sort of patchwork all over the place, all the predictions of our model. 390 00:43:34,250 --> 00:43:40,430 So in our model we've got a number of we've got a number of clouds moving around and we can, which is noisy, 391 00:43:40,430 --> 00:43:47,240 of course, and then we can, uh, we can in each of these wakes, we have a certain amount of turbulence, 392 00:43:47,510 --> 00:43:55,100 so we can say how much gas we expect to find in this pixel here at because because of the because of which clouds 393 00:43:56,510 --> 00:44:02,960 which have a turbulent wake which might scatter which might scatter a small volume of gas into that velocity. 394 00:44:03,230 --> 00:44:09,200 So the so the the, the, except when you're outside one of these circles, you're looking at one of our model. 395 00:44:09,380 --> 00:44:12,860 When you're inside one of these circles, you're looking at the actual data. 396 00:44:13,040 --> 00:44:19,660 And I think you have to agree that there's pretty decent agreement. I think that's all I have to say about that. 397 00:44:19,780 --> 00:44:24,969 You can also ask down which lines of sight do you expect to see? 398 00:44:24,970 --> 00:44:32,620 Absorptions, these black dots are these are plots of velocity versus galactic longitude in different a different galaxy. 399 00:44:32,820 --> 00:44:36,100 Each plot is belongs to a different galactic latitude. 400 00:44:36,100 --> 00:44:44,200 This is, uh. So this is -50 degrees, -20 degrees plus 50 degrees, I think, plus 20 degrees. 401 00:44:44,530 --> 00:44:50,290 Anyway, the point is that the model gives you probabilities which are, which is signalled by the, 402 00:44:50,560 --> 00:44:54,940 by the coloured contours and the black dots where you actually detect the absorption. 403 00:44:55,690 --> 00:44:59,020 And again, there's reasonable agreement, you can show that statistically pretty decent thing. 404 00:44:59,380 --> 00:45:06,820 So the key points of my discussion so far are that we have a model of this fountain 405 00:45:06,820 --> 00:45:11,440 process that yields a coherent picture of a very large body of disparate data, 406 00:45:11,950 --> 00:45:18,880 the 21 centimetre data cubes for external galaxies, the very different 21 centimetre data cube for our own galaxy, 407 00:45:19,210 --> 00:45:25,420 and these measurements of ultraviolet absorption line systems in our own galaxy. 408 00:45:26,590 --> 00:45:33,579 And it implies that the star formation driven fountain draws gas out of the corona, 409 00:45:33,580 --> 00:45:41,110 out of the the gas at the very old temperature that fills intergalactic space at a rate 410 00:45:41,110 --> 00:45:44,979 that's roughly equal to the rate at which gas is being used up by star formation. 411 00:45:44,980 --> 00:45:58,930 And therefore it replaces losses by star formation, and it requires that star forming galaxies have to be enveloped in a corona, in the corona, 412 00:45:58,930 --> 00:46:05,920 which above the disk is almost but not quite cool enough and dense enough to to 413 00:46:05,950 --> 00:46:10,600 to make the transition from the very old temperature down to low temperatures. 414 00:46:10,870 --> 00:46:14,800 So it needs a little help and it gets this help from the cold gas sent up. 415 00:46:15,010 --> 00:46:19,210 But if the corona were much hotter than it is or than we think it is, 416 00:46:19,480 --> 00:46:23,780 this process wouldn't work and the gas you sent out would become evaporated and less would come down. 417 00:46:23,810 --> 00:46:31,700 You'd run out of gas on a fast rate. So I want to connect this now to the global picture of galaxy evolution. 418 00:46:31,730 --> 00:46:41,240 Here is a very famous diagram from from ten years ago when from analysis of the enormous statistics gathered by the Sloan Digital Sky Survey. 419 00:46:42,500 --> 00:46:47,510 So what this is showing is colour with blue colours at the bottom and red colours at the top. 420 00:46:47,840 --> 00:46:51,110 And what we're seeing here is absolute magnitude. 421 00:46:51,130 --> 00:46:55,640 So these are blue. This is high, very high luminosity at this end of a galaxy. 422 00:46:55,970 --> 00:46:59,180 This is medium to low luminosity at this end. 423 00:46:59,480 --> 00:47:09,380 And what you've got. So this survey saw millions of galaxies and was able to assign these galaxies luminosity is it was able to assign them colours. 424 00:47:09,620 --> 00:47:16,910 And now what's plotted here is the density, the number of galaxies, the number density of galaxies in this luminosity colour plot. 425 00:47:17,240 --> 00:47:21,980 And what you see is a by modality. There's, there's what's called a blue cloud here. 426 00:47:22,160 --> 00:47:29,630 A relatively low luminosity is in bluish colours. And there's the red sequence here at higher luminosity is and red colours. 427 00:47:31,100 --> 00:47:38,300 So the general. So in a huge amount of work has been going on over the last decade to try and understand this in a constructive way. 428 00:47:40,080 --> 00:47:48,850 Uh. Right. Another interesting fact is that the red galaxies tend to occur in denser regions, 429 00:47:48,850 --> 00:47:55,330 regions like clusters, rather than in the in the voids between clusters. 430 00:47:57,790 --> 00:48:01,929 And it's generally thought that galaxies evolve from the blue to the red, 431 00:48:01,930 --> 00:48:07,719 because the reason these galaxies are relatively blue is because they're still forming stars. 432 00:48:07,720 --> 00:48:18,459 They and because the stars that are blue are typically of the importance the blue colours are given to these galaxies by relatively massive, 433 00:48:18,460 --> 00:48:24,220 therefore short lived stars, and therefore only galaxies that are forming stars are capable of being this blue. 434 00:48:24,550 --> 00:48:28,690 So the general idea is that's accepted by everybody that these galaxies, 435 00:48:28,690 --> 00:48:34,419 that galaxies at some level migrate up in this diagram and they possibly migrate sideways as a result of merging, 436 00:48:34,420 --> 00:48:40,090 but they certainly migrate upwards from blue to red as their star formation peters out. 437 00:48:40,360 --> 00:48:45,640 And you're just left with the old low mass stars that will go on glowing for a very long time, but then they're red. 438 00:48:47,470 --> 00:48:59,080 And I think the basic reason why stars move from blue to red is that star galaxies start to move with respect to their intergalactic medium. 439 00:48:59,770 --> 00:49:08,800 So they haven't spoken much at all yet about about the motion of a galaxy with respect to the intergalactic medium that it finds itself in. 440 00:49:09,070 --> 00:49:15,190 But if you're in a cluster of galaxies, there is only one galaxy which can be stationary with respect to the intergalactic medium, 441 00:49:15,190 --> 00:49:20,349 and that's the galaxy at the middle of the cluster, because it's the only because the intergalactic medium will be stationary. 442 00:49:20,350 --> 00:49:25,060 With respect to the cluster, well, it might conceivably be rotating, but we don't expect it to be rotating very much. 443 00:49:27,520 --> 00:49:31,209 And the other galaxies are zipping around at random. And so they're moving. 444 00:49:31,210 --> 00:49:32,620 With respect to the intergalactic medium, 445 00:49:32,620 --> 00:49:39,070 it is very difficult for the processes that I've been discussing to function if you're moving fast with respect to the intergalactic medium. 446 00:49:40,780 --> 00:49:44,559 But there is a paradox in this in this in all this, 447 00:49:44,560 --> 00:49:53,200 which is that where the intergalactic medium is densest and most ready and cooling and radiating fastest, 448 00:49:53,980 --> 00:49:57,010 namely in the middle of rich clusters of galaxies. 449 00:49:57,010 --> 00:49:59,950 Remember, we knew we've known for 30 years, 450 00:49:59,950 --> 00:50:08,560 we've had vivid evidence that the clusters of galaxies are full of enormous masses of very low temperature gas, 451 00:50:08,770 --> 00:50:13,960 because it emits so strongly in X-rays that even the X-ray telescopes of 1970 could detect. 452 00:50:16,170 --> 00:50:23,790 So so in these dense these regions is exactly where the galaxies are reddest and dearest and not forming stars. 453 00:50:24,870 --> 00:50:28,880 And the explanation for that, I think, is a. 454 00:50:32,920 --> 00:50:38,499 Is that the cooling rate of the intergalactic medium is highest in a cluster of galaxies. 455 00:50:38,500 --> 00:50:43,720 It is indeed highest at the very centre of the cluster of galaxies where the pressure is largest. 456 00:50:46,290 --> 00:50:51,980 And the, the, the, the accretion rate onto the central black hole. 457 00:50:51,990 --> 00:51:03,090 So. So this is this is in fact, a radio continuum picture of a Galaxy M87 that sits at the middle of a relatively small cluster, the Virgo cluster. 458 00:51:04,620 --> 00:51:10,530 And in the middle there, there is an enormous black hole with a mass of some by ten to the nine solar masses. 459 00:51:10,890 --> 00:51:15,480 And the rate at which this thing accretes is a very sensitive function of the temperature. 460 00:51:15,540 --> 00:51:21,660 So as the temperature goes down, the accretion rate on the black hole goes up and the black hole accretes. 461 00:51:21,900 --> 00:51:28,890 It emits jets which here impact on the we see impacting on the intergalactic medium. 462 00:51:29,700 --> 00:51:35,520 This is this is a blow up of the middle of this. So here is the this is a much smaller scale picture of what you see in the middle here. 463 00:51:36,360 --> 00:51:42,360 So a jet comes out, it it impacts on the ambient gas and re heats it. 464 00:51:42,600 --> 00:51:52,950 So if the temperature in the middle of the cluster drops significantly, you get a major energy outburst from the from the engine in the middle, 465 00:51:53,130 --> 00:52:00,930 which reads the system and keeps this this variable temperature gas from falling to the kinds of temperatures less than 100 degrees K, 466 00:52:00,930 --> 00:52:04,830 maybe 30 degrees K required to form significant numbers of stars. 467 00:52:05,700 --> 00:52:10,800 And this is something on which I worked with Tibor and Henri Goma in the past. 468 00:52:12,130 --> 00:52:15,160 In our own galaxy, we have a rather wimpy black hole. 469 00:52:15,160 --> 00:52:18,520 Sagittarius a star which you can't see but sits there. 470 00:52:18,880 --> 00:52:25,420 This is an X-ray picture of the middle of the galaxy. I showed you previously a radio continuum picture of the middle of the Virgo cluster. 471 00:52:26,950 --> 00:52:32,769 And I think the argument would be that the role of Sagittarius a star is to prevent the 472 00:52:32,770 --> 00:52:37,630 corona gas of our own galaxy from cooling in the place where it's most ready to cool, 473 00:52:37,840 --> 00:52:40,660 which is right in the middle of the galaxy where the pressure is highest. 474 00:52:41,230 --> 00:52:45,730 So the question because the question I'm addressing is why is it that the star formation in our 475 00:52:45,730 --> 00:52:53,380 galaxy that occurs outside of is predominantly occurs while away from the centre of the galaxy, 476 00:52:54,220 --> 00:52:56,740 although it must be in the centre of the galaxy. 477 00:52:56,920 --> 00:53:05,809 Even though I argue that what sustained star formation in the Milky Way is accretion of gas from this corona of a very low temperature gas, 478 00:53:05,810 --> 00:53:06,410 and this corona, 479 00:53:06,410 --> 00:53:13,420 a variable temperature gas, may not be ready to cool where we are, but surely where it's densest in the middle of the galaxy is ready to cool. 480 00:53:13,420 --> 00:53:16,479 And the answer is the answer I would give you is Yes it does. 481 00:53:16,480 --> 00:53:22,510 It is keen to cool the air, but there sits the monster which is capable of reheating it very, 482 00:53:22,510 --> 00:53:29,500 very promptly in the event that it allows its temperature there to drop. 483 00:53:30,950 --> 00:53:36,620 So so that's the picture one has that the middle of the corona is prevented from cooling by the black hole. 484 00:53:36,890 --> 00:53:48,080 But out here, the the fountain, which is driven by star formation, reaches up into the corona and grabs gas, brings it down to the turns. 485 00:53:48,080 --> 00:53:56,149 Hot gas into cold gas brings it into the the disk of cold gas that circulates around the galaxy within which star formation happens. 486 00:53:56,150 --> 00:54:01,100 And this is what sustains. Star formation is only very slowly falling right over the life of the galaxy. 487 00:54:01,670 --> 00:54:04,670 But consider what what's likely to happen otherwise. 488 00:54:05,780 --> 00:54:16,160 If you have a major merger and our galaxy is falling together with the Andromeda Nebula M31 and we will tumble together 489 00:54:17,630 --> 00:54:26,120 in a very interesting orgy of stellar dynamics and hydrodynamics and star formation in about three giga years. 490 00:54:26,600 --> 00:54:32,479 I think that's the right number and when that happens, there will be a burst of star formation. 491 00:54:32,480 --> 00:54:39,170 This is something Jo Silk has worked on a great deal because the gas will be vigorously shot and so on and supernova 492 00:54:39,440 --> 00:54:47,479 will will quickly heat much of the gas and quite likely will will will succeed in heating all of the gas, 493 00:54:47,480 --> 00:54:51,260 all of the cold gas. So there's no cold gas left in that case. 494 00:54:51,560 --> 00:54:54,410 With no cold gas left, there'll be no star formation. 495 00:54:55,440 --> 00:55:00,510 There will be no fountain to reach up into the corona at distances safely removed from the black hole. 496 00:55:00,810 --> 00:55:05,760 They'll just be this nice, smooth atmosphere, a very low temperature gas, 497 00:55:07,080 --> 00:55:14,580 densest and in danger of cooling and forming stars only right on top of the new improved black hole that will 498 00:55:14,580 --> 00:55:20,550 form by the merging of Sagittarius Star and the black hole which sits at the middle of the Andromeda Nebula. 499 00:55:22,290 --> 00:55:32,910 The black hole will periodically thrash the variable temperature gas and keep make sure it doesn't get any ideas about cooling too much. 500 00:55:33,420 --> 00:55:33,960 And that will be it. 501 00:55:34,590 --> 00:55:46,110 And our galaxy, well, the new improved galaxy formed by us and and the Andromeda Nebula will grow red and dead like those things that we have seen. 502 00:55:47,150 --> 00:55:54,950 So here are my conclusions. Most of the baryons lie between galaxies rather than in galaxies. 503 00:55:55,940 --> 00:55:59,630 We've only a partial understanding of why of how it comes. 504 00:56:00,170 --> 00:56:09,469 It's a it is. It's still a little bit puzzling that star formation is so efficient at pushing gas out of galaxies. 505 00:56:09,470 --> 00:56:20,960 But anyway, it's a fact we have to bear in mind. Galactic Centre Black holes prevent substantial accretion of taking place in the natural place, 506 00:56:21,200 --> 00:56:26,990 which is at the centre of a corona of hot gas where the density is highest. 507 00:56:28,340 --> 00:56:31,600 But in blue galaxies are the blue clouds such as stars. 508 00:56:31,660 --> 00:56:35,809 Star formation has been sustained for the lifetime of the universe, 509 00:56:35,810 --> 00:56:43,430 essentially by accretion from the IG m sustained through the through this fountain process that I've described to you. 510 00:56:43,760 --> 00:56:51,410 We have a model of the impact of this star formation on the insular medium that was structured to explain data from 511 00:56:52,100 --> 00:56:59,899 external galaxies but does a super job of of of of the of the data for the very different data for our own galaxy. 512 00:56:59,900 --> 00:57:03,780 So I'm completely convinced that it's right. And galaxies. 513 00:57:03,800 --> 00:57:10,580 The conclusion from all this is that what makes galaxies red and dead is when some event, particularly a merger. 514 00:57:10,790 --> 00:57:17,750 Well, some event could be two things. One can be a merger, which is probably our fate when we fall together with with with the Andromeda Nebula. 515 00:57:18,050 --> 00:57:23,360 The other thing that can kill your star formation is that you fall in to a rich cluster of galaxies, 516 00:57:23,600 --> 00:57:32,000 and then you you fall through the very temperature gas of that cluster of galaxies, which blows away very quickly. 517 00:57:32,210 --> 00:57:36,590 Your own corona. Of of of thermal temperature. 518 00:57:36,590 --> 00:57:43,880 Gas which was stationary with respect to you. And from now on you're moving with respect to the variable temperature grass that surrounds you. 519 00:57:44,030 --> 00:57:50,509 And in those circumstances, if you are foolish enough to put up a cloud of cold gas into the corona, 520 00:57:50,510 --> 00:57:57,740 hoping to harvest some of this life sustaining coronal gas for your own purposes, you'll. 521 00:57:58,860 --> 00:58:06,479 Cold packet will simply be blown away and incorporated in the in the bigger entity's pool of variable temperature gas. 522 00:58:06,480 --> 00:58:09,440 And that's how galaxies become red and dead. Thank you.