1 00:00:01,150 --> 00:00:11,390 Of. Right. 2 00:00:11,390 --> 00:00:19,130 I'm conscious that it's already lunchtime. I hope you have something to munch on at the coffee break because I'm going to 3 00:00:19,130 --> 00:00:25,400 discuss what's arguably the biggest problem in cosmology and in fundamental physics, 4 00:00:25,400 --> 00:00:30,020 and I'll take at least half an hour to do that. I hope you'll bear it. 5 00:00:30,020 --> 00:00:35,420 So I want to start by asking if any of you here were contemporary with this man? 6 00:00:35,420 --> 00:00:46,310 Does anyone know him? Yes, yes. So then you can tell me the date when when he did his best DPhil, which was the thing with Christopher Lloyd Smith. 7 00:00:46,310 --> 00:00:53,570 I have 75 years was that? Is that about right? Yeah. So Ian Hinchcliffe is a distinguished particle physicist. 8 00:00:53,570 --> 00:01:04,670 He's at Berkeley Lawrence Berkeley National Lab, and he is, of course, well known for his contributions to particle physics. 9 00:01:04,670 --> 00:01:11,660 But he's in fact perhaps best known for having been credited with something called 10 00:01:11,660 --> 00:01:17,780 Hinchcliffe Rule that says if the title of the article is a yes no question, 11 00:01:17,780 --> 00:01:24,050 then the answer is no, right? Of course, if you think about it for a minute, as any logician would, 12 00:01:24,050 --> 00:01:31,640 they would come up with another question, and this is a real paper submitted to the archive. 13 00:01:31,640 --> 00:01:37,010 And in fact, this is their paper because it's so hedged. 14 00:01:37,010 --> 00:01:41,210 His assertion is false, but only if it is true, right? 15 00:01:41,210 --> 00:01:52,100 So it is in that context that I want you to see the title that I have chosen for my talk, which indeed has a question mark at the end. 16 00:01:52,100 --> 00:01:59,930 And I will try and present to you the evidence, which makes me wonder if this is not subject to Hinchcliffe rule. 17 00:01:59,930 --> 00:02:06,770 But you all know what this is about because this has been with us now for about twenty one years. 18 00:02:06,770 --> 00:02:08,900 It made the cover of science. 19 00:02:08,900 --> 00:02:18,050 It was arguably the biggest breakthrough in cosmology since the discovery of the cosmic microwave background, and it has been richly rewarded. 20 00:02:18,050 --> 00:02:24,470 Various prises of Philip just talked about supernova, a binary mergers making gold. 21 00:02:24,470 --> 00:02:29,780 They got their gold from Stockholm. And this has continued. 22 00:02:29,780 --> 00:02:35,870 It is very much part of the establishment. This is the bedrock of the standard cosmological model. 23 00:02:35,870 --> 00:02:42,740 And yet I will try and argue to you that we need to be conscious of that question mark. 24 00:02:42,740 --> 00:02:54,390 In fact, I'll present to evidence that very likely the observations do not actually tell us anything about the cosmological constant or dark energy. 25 00:02:54,390 --> 00:03:02,520 Now to do that, I have to take you back to how this whole business started, and it has really all got to do with measuring distances in cosmology. 26 00:03:02,520 --> 00:03:07,860 Now to measure distances, you need standard candles, things which have a known wattage. 27 00:03:07,860 --> 00:03:11,760 So you know from how fainter, how far they are. 28 00:03:11,760 --> 00:03:20,460 And to do this, you must find some of the property of the object that tells you how bright it is intrinsically and historically. 29 00:03:20,460 --> 00:03:27,300 The first such objects were so-called secret variables because they are found in the constellation of Seaforth. 30 00:03:27,300 --> 00:03:35,820 And as you can see here, there periodicity you can see flashing on and off correlates with their absolute luminosity. 31 00:03:35,820 --> 00:03:41,010 And this is very advantages to for using them as standard candles. 32 00:03:41,010 --> 00:03:51,360 And the most at first, most dramatic use of these was by Edwin Hubble, who discovered in the outer periphery of our nearest neighbour, Andromeda. 33 00:03:51,360 --> 00:03:56,700 M31 a little object there, but she has written v r for variable. 34 00:03:56,700 --> 00:04:04,230 And this lady, Henrietta Leavitt, had already shown that relationship based on the previous light. 35 00:04:04,230 --> 00:04:10,140 She had established that these could be used as distance indicators, and riding on that, 36 00:04:10,140 --> 00:04:16,380 Hubble was able to deduce that this what was a nebula cloud in our galaxy, 37 00:04:16,380 --> 00:04:23,550 because that's all we knew about the universe then, was actually another galaxy just like ours, but quite far away. 38 00:04:23,550 --> 00:04:31,680 In fact, it is two and a half million light years away. Our galaxy, for comparison, is about 30000 light years in scale. 39 00:04:31,680 --> 00:04:37,350 And so the universe certainly became a lot bigger. And I would say this was the beginning of what we today. 40 00:04:37,350 --> 00:04:45,930 Our cosmology, physical cosmology today. Now, of course, fast forward in time, we have better instruments now in orbit. 41 00:04:45,930 --> 00:04:56,160 The Hubble telescope and that's powerful enough to see a secret variable in the outer periphery of this galaxy and 100, which is in the Virgo cluster. 42 00:04:56,160 --> 00:05:03,000 This is a cluster of galaxies to which we belong. And you can see it flashing on and off. 43 00:05:03,000 --> 00:05:08,370 And that tells us that that galaxy is about 54 million light years away. 44 00:05:08,370 --> 00:05:13,380 This is the directly calibrated distance because we can see the speed. 45 00:05:13,380 --> 00:05:18,780 But if you waited a little time, then you will find that a supernova goes off in this galaxy, 46 00:05:18,780 --> 00:05:24,000 as Philip said, they go off roughly once a century and a galaxy of that size. 47 00:05:24,000 --> 00:05:27,780 And you can tell that it's a type one supernova. 48 00:05:27,780 --> 00:05:37,740 Here is the nickel 56 decay. Of course, this is the die curve, but the real data shows also the features that, he explained. 49 00:05:37,740 --> 00:05:46,260 And that allows us to now use calibrate this supernova because we know the distance to this galaxy already. 50 00:05:46,260 --> 00:05:51,690 And then when we see similar supernovae further away in the universe, we know the distances. 51 00:05:51,690 --> 00:05:55,960 So now we can measure distances to billions of light years. 52 00:05:55,960 --> 00:06:03,220 And of course, when you are doing that, we are looking back in time, that's why this is important for cosmology and you know that we see the Sun, 53 00:06:03,220 --> 00:06:10,930 as it was a little back in the past, said Drove our galaxy is about thirty thousand years ago, and yet his star is four years. 54 00:06:10,930 --> 00:06:16,750 I just showed you the Andromeda nebula. That's two and half million years. 55 00:06:16,750 --> 00:06:22,750 We see Virgo as it was when the dinosaurs were dying out. And in fact, when you look in the Hubble deep field, 56 00:06:22,750 --> 00:06:28,030 which is the deepest view of the universe taken by our telescope, we actually look back about 12 billion years. 57 00:06:28,030 --> 00:06:34,880 It's really travelling back in time. Time travel is possible only backwards. 58 00:06:34,880 --> 00:06:40,390 Now you all know that when you look at those galaxies, that is another feature that is interesting. 59 00:06:40,390 --> 00:06:47,620 Their spectra are all shifted towards the red as we go from a nearby star to the more distant galaxies. 60 00:06:47,620 --> 00:06:58,840 And this is the redshift the fraction of change in the wavelength. And you normally would attribute this to the same. 61 00:06:58,840 --> 00:07:04,600 Edwin Hubble, incidentally, has an Oxford degree. You might be surprised to know he wasn't that this department. 62 00:07:04,600 --> 00:07:10,600 He actually did a degree in jurisprudence of all things, and he was a Rhodes scholar. 63 00:07:10,600 --> 00:07:17,200 So he is credited with the discovery of the expansion of the universe, and here is his original, 64 00:07:17,200 --> 00:07:24,130 rather sketchy data showing the correlation between the velocity and the distance. 65 00:07:24,130 --> 00:07:28,090 In fact, he actually never interpreted this as the expansion of the universe. 66 00:07:28,090 --> 00:07:38,560 He said that task of interpretation should be left to theories, and I'll make it clear later why his data, of course, was rather close by. 67 00:07:38,560 --> 00:07:43,300 Today, we have data extending a lot further, as I mentioned. 68 00:07:43,300 --> 00:07:50,590 So therefore, we are in a position to say things and draw inferences that Hubble could not have done. 69 00:07:50,590 --> 00:08:00,640 And the first thing I want to emphasise is that those red chips are not due to, as one might imagine, the Doppler effect, which is familiar to us. 70 00:08:00,640 --> 00:08:08,560 It's the whole new effect. It is caused by the stretching of spacetime so that when a photon is emitted from 71 00:08:08,560 --> 00:08:13,540 a particular source at a given time later when you observe that same photon, 72 00:08:13,540 --> 00:08:19,000 its wavelength, of course, is the same regard to the space it is moving in locally. 73 00:08:19,000 --> 00:08:26,230 But the metric of spacetime has changed in the intervening time, and therefore we see it as a field on a longer wavelength. 74 00:08:26,230 --> 00:08:33,710 And that is the redshift. And this picture, of course, also makes it clear that this expansion, as it is, is everywhere. 75 00:08:33,710 --> 00:08:42,590 The same looks the same from any point on the sphere. So it's not really an expansion that begs the question what are you expanding into in all that? 76 00:08:42,590 --> 00:08:49,670 So let's now fast forward to 1998, when this discovery of the accelerating expansion was made. 77 00:08:49,670 --> 00:08:56,810 So two groups of astronomers were looking at the magnitude of titans of innervate versus the redshift. 78 00:08:56,810 --> 00:08:59,870 And as you can tell from this last here, 79 00:08:59,870 --> 00:09:11,060 the data were indicating that the expansion rate instead of slowing down as one might expect it does under gravity, was in fact speeding up. 80 00:09:11,060 --> 00:09:19,730 But this data is perhaps not convincing enough, but it was a strong enough case that they were given the Nobel prise for it. 81 00:09:19,730 --> 00:09:28,100 And when I applaud the same data in terms of the scale of the universe versus the time which is inferred from the redshift, 82 00:09:28,100 --> 00:09:32,600 we don't directly, of course, see 10 billion years back in time. 83 00:09:32,600 --> 00:09:38,470 We inferred from the rich ships that are measured of these objects. And that seems to indeed indicate that. 84 00:09:38,470 --> 00:09:43,790 But rather than be on a girl that was only always slowing down, 85 00:09:43,790 --> 00:09:51,590 the expansion has started accelerating in the last few billion years, which is really a redshift of order one. 86 00:09:51,590 --> 00:09:56,550 But. Now, I'm sorry to go through all this rather quickly, 87 00:09:56,550 --> 00:10:02,400 but I take it you have been exposed to it over the last 21 years and this has 88 00:10:02,400 --> 00:10:08,280 led ultimately to our formulation of the so-called standard model of cosmology, 89 00:10:08,280 --> 00:10:10,950 which I can right now on one page. 90 00:10:10,950 --> 00:10:18,450 In two equations, much as the standard model of particle physics can be now written, the Lagrangian can be on a t shirt. 91 00:10:18,450 --> 00:10:25,230 There are big differences between the two standard models, which I'll highlight. So what you have here is the metric of spacetime, 92 00:10:25,230 --> 00:10:34,620 which allows you to be the simplest form possible because we are assuming that the universe is isotropic and homogeneous, as Einstein and others did. 93 00:10:34,620 --> 00:10:40,380 When the theory of cosmology was first formulated nearly a hundred years ago now. 94 00:10:40,380 --> 00:10:46,230 And that allows a considerable simplification of Einstein's equations, 95 00:10:46,230 --> 00:10:51,720 which take that metric of spacetime and the curvature tensor that you construct from it, 96 00:10:51,720 --> 00:10:59,430 and of various other objects like the DC Tensor and the scalar and so on and related to the energy momentum. 97 00:10:59,430 --> 00:11:07,680 Because as you well know, that's what tells space how to curve and the curvature in correspondingly dense matter, 98 00:11:07,680 --> 00:11:10,410 how to move in with those famous words. 99 00:11:10,410 --> 00:11:18,870 So we can now see that sphere that I showed you earlier is depicted here with the radial coordinate being conformal time or redshift. 100 00:11:18,870 --> 00:11:21,450 We are sitting on the North Pole if you like, 101 00:11:21,450 --> 00:11:28,980 and we are looking at the sphere and light can only come to us on the surface of the sphere because that's all there is. 102 00:11:28,980 --> 00:11:37,200 And the antiporda point is the Big Bang. That's why the Big Bang fills our entire sky, even though it is a singular point. 103 00:11:37,200 --> 00:11:41,730 This geometry, non-nuclear and geometry makes that possible. 104 00:11:41,730 --> 00:11:45,930 And in fact, as we go back in redshift or conformal time, 105 00:11:45,930 --> 00:11:52,320 we come to a point that the universe was a thousand times smaller than it is today and thousand times therefore hotter. 106 00:11:52,320 --> 00:11:54,510 And that's enough to ionised hydrogen. 107 00:11:54,510 --> 00:12:01,290 So the rest of the time the universe is ionised and the photons starting out from the Big Bang at Brexit of infinity, 108 00:12:01,290 --> 00:12:07,800 which is the dashed line when random walk just like the four dogs trying to get out of his star. 109 00:12:07,800 --> 00:12:11,900 And then when they reach this point, they are emitted and come to us. 110 00:12:11,900 --> 00:12:16,020 So this is the photosphere of the universe. We are really inside. 111 00:12:16,020 --> 00:12:20,410 It's an inside out picture of a star. And we are in sight. 112 00:12:20,410 --> 00:12:25,070 OK, that's that is the analogy to what you've been hearing about so far. 113 00:12:25,070 --> 00:12:32,280 Now, mathematically, we can now reduce the very complicated equations that Einstein gave us to a very simple one, 114 00:12:32,280 --> 00:12:39,090 which is named after Alexander Friedman by using this metric to construct this object. 115 00:12:39,090 --> 00:12:48,360 And that tells us how the scale factor changes in response to the matter and the curvature and the object called the cosmological constant, 116 00:12:48,360 --> 00:12:50,700 which you are allowed to add, 117 00:12:50,700 --> 00:13:00,420 which we must add there, because that is reflecting the underlying symmetry general coordinate evidence that underlies general relativity. 118 00:13:00,420 --> 00:13:07,440 We have no option in that matter. So if I take this equation, this is called the Hubble Barometer. 119 00:13:07,440 --> 00:13:12,480 And if you divide this through, then I can rewrite it in this form. 120 00:13:12,480 --> 00:13:20,970 With these matter densities written in terms of the fractional contribution it makes to the critical density as defined here. 121 00:13:20,970 --> 00:13:29,070 And that scales with redshift as you see as one does that you just reflecting the dilution of matter with the expansion in the early universe, 122 00:13:29,070 --> 00:13:38,490 the rapid radiation. So then that would become one if to the fold, because the results of the the stretching of the photons, as I mentioned, 123 00:13:38,490 --> 00:13:46,940 then we have curvature, which goes as that square and there is omega lambda that goes as nothing because it's a constant. 124 00:13:46,940 --> 00:13:55,640 And this then gives us a so-called some rule, which says that these three guys are up to one and that there are only three 125 00:13:55,640 --> 00:14:00,890 terms because that is forced on us by these as you homogeneity and as a trophy. 126 00:14:00,890 --> 00:14:12,290 OK. That is an assumption. But we use that some rule to infer that because matter only makes about about a third of the universe and the curvature 127 00:14:12,290 --> 00:14:18,740 appears to be close to zero from the cosmic microwave background that there must be something which makes up the rest. 128 00:14:18,740 --> 00:14:25,400 And that has got to be this omega lambda. And that then turns out to be 70 percent or so of the universe. 129 00:14:25,400 --> 00:14:29,960 And because omega lambda is defined as lambda, over three is not square. 130 00:14:29,960 --> 00:14:37,340 That means the lambda is of ordered the present day Hubble parameter square, which is a rather strange. 131 00:14:37,340 --> 00:14:47,840 So you can see how omega lambda has come to be. How we got to this crazy direction that lambda has a scale of 10 to the minus 40 to G.V., 132 00:14:47,840 --> 00:14:52,790 because that is what the value of H.A. is in particle physics terms. 133 00:14:52,790 --> 00:14:58,250 And so if I try to compute what this corresponds to in terms of an energy density, 134 00:14:58,250 --> 00:15:04,070 I would get that it is of orders about a million electron volts into the minus 12 GV, 135 00:15:04,070 --> 00:15:12,530 which of course, I don't need to tell you is a lot smaller than any scale in fundamental physics that we are aware of. 136 00:15:12,530 --> 00:15:18,980 The weak scale. The Fermi scale, which is of order one hundred, gives where all the action is now. 137 00:15:18,980 --> 00:15:25,850 That's where we are. We have actually gone a factor of 10. Beyond that, we are doing experiments at the TV scale. 138 00:15:25,850 --> 00:15:32,270 And although this stuff is not going to affect anything we do at the Large Hadron Collider, 139 00:15:32,270 --> 00:15:42,220 we have to come to this conclusion from cosmology that actually most of the universe is made of a cosmological constant at this crazy scale. 140 00:15:42,220 --> 00:15:47,830 And this actually makes no physical sense, OK? 141 00:15:47,830 --> 00:15:52,990 There is a cosmological constant problem, we have known about it for 80 years, 142 00:15:52,990 --> 00:16:01,960 ever since one of those formulated Powley and the Lubitsch and others pointed out that the zero point energy the vacuum coupled to gravity, 143 00:16:01,960 --> 00:16:06,550 according to Einstein, and therefore it had to be very, very small. 144 00:16:06,550 --> 00:16:12,190 Otherwise, the universe would not exist as we find it. That's the possible logical constant problem. 145 00:16:12,190 --> 00:16:17,290 Not only has the problem not been solved by the best minds since then, 146 00:16:17,290 --> 00:16:25,860 we now have the astronomers telling us that it does exist at a level which is set by the present rate of expansion. 147 00:16:25,860 --> 00:16:32,460 It makes no sense, it infuriates me, OK, that this is the basis of our standard cosmological model, 148 00:16:32,460 --> 00:16:41,310 which is taken to be a standard model in the same way as the standard model of particle physics, which I believe has been somewhat better tested. 149 00:16:41,310 --> 00:16:46,620 And that is what theme of my talk is that we have had all these satellites w my 150 00:16:46,620 --> 00:16:52,830 plant and the Sloan Digital Sky Survey and the Dark Energy Survey and so on, 151 00:16:52,830 --> 00:16:57,930 which are telling us that the universe is made of mostly of dark energy. 152 00:16:57,930 --> 00:17:05,040 But the basic foundations of this model, which is this assumption of isotropic and homogeneity, which, 153 00:17:05,040 --> 00:17:13,920 as it says here, underlie this deduction, this rather important deduction rests entirely on these assumptions. 154 00:17:13,920 --> 00:17:21,390 We need to test them in the same way that the basic assumptions of the standard model of particle physics have been tested. 155 00:17:21,390 --> 00:17:26,440 They have been tested at the quantum level and it is passed every test super. 156 00:17:26,440 --> 00:17:31,270 What about testing the foundational assumptions of the standard cosmological model? 157 00:17:31,270 --> 00:17:36,100 I think that somebody needs to do this, which is why, although I mean the part of the theory group, 158 00:17:36,100 --> 00:17:43,480 I'm going to spend the rest of time telling you about astronomical data and the right way to analyse them. 159 00:17:43,480 --> 00:17:48,850 So you heard all about type one supernovae. There are various possibilities for how they come to be. 160 00:17:48,850 --> 00:17:50,740 You heard about the taxonomy. 161 00:17:50,740 --> 00:17:58,660 We are interested in these because these are closest to being standard candles and they are measured at various wavelengths with various names, 162 00:17:58,660 --> 00:18:02,600 bands and we learn all this jargon. 163 00:18:02,600 --> 00:18:08,050 It's kind of hard to come to terms with the fact that although it makes sense to show the 164 00:18:08,050 --> 00:18:13,180 logarithmic measure of the brightness because they do vary over orders of magnitude, 165 00:18:13,180 --> 00:18:16,750 for some reason, the astronomers then multiplied by minus two point five. 166 00:18:16,750 --> 00:18:22,690 I've never understood why this is very confusing, because it means the more negative the number, the brighter it is. 167 00:18:22,690 --> 00:18:31,810 OK, just keep that in mind. So here we are. Various light curves, as you also saw earlier, and it is quite clear that they've got a lot of scatter. 168 00:18:31,810 --> 00:18:41,530 You call that this is a log scale, right? However, Phillips, a astronomer at the Anglo-Australian Telescope, 169 00:18:41,530 --> 00:18:47,860 discovered that there is a nice and b correlation between the peak magnitude and the width of this light. 170 00:18:47,860 --> 00:18:54,010 Good, right? And it's slightly different, according to what band you observe them and what the colour is. 171 00:18:54,010 --> 00:19:01,960 Now, this is pretty interesting, and I should emphasise that this has been made possible only because we see the right time. 172 00:19:01,960 --> 00:19:08,020 You see this data points which are there before the supernova went off, and you might well ask, 173 00:19:08,020 --> 00:19:16,450 how did I know that a supernova was going to go off at some random point in the sky so that I start making measurements two weeks earlier? 174 00:19:16,450 --> 00:19:23,560 That is the great advance in the field that was made possible by just couple device cameras, 175 00:19:23,560 --> 00:19:28,240 which can keep scanning the sky over and over again so that when a supernova does go off, 176 00:19:28,240 --> 00:19:31,960 you can go back and look at the picture that you took two weeks earlier, 177 00:19:31,960 --> 00:19:39,700 and that was the real advance in supernova astronomy in the 90s, which led to this discoveries. 178 00:19:39,700 --> 00:19:45,640 This correlation, by the way, is not understood. Theoretically, that's the kind of thing doesn't always work on, but it is empirical. 179 00:19:45,640 --> 00:19:53,650 We can use it because it works. It reduces that stature to the point where you can actually use this things as standard candles. 180 00:19:53,650 --> 00:20:01,990 And James has told us about the basic underlying physics, which you can trust because it's got the word suddenly trigger mass in it. 181 00:20:01,990 --> 00:20:08,590 So, you know, there is some really sound physics underlying that, and therefore we can use these as standard candles. 182 00:20:08,590 --> 00:20:14,950 But in practise, we have to make corrections to the magnitudes a so-called stretch correction to stretch 183 00:20:14,950 --> 00:20:19,150 literally the light burst in the fall on top of each other and the colour correction, 184 00:20:19,150 --> 00:20:27,340 depending on at what redshift we measured them. Recall that when you see them at high redshift, the emission will be moved towards the infra-red. 185 00:20:27,340 --> 00:20:31,720 That's why you have to go out on the Hubble Space Telescope or something to look at. 186 00:20:31,720 --> 00:20:39,730 And this is the current state of art is called the salt, the spectral adaptive logic of template version two. 187 00:20:39,730 --> 00:20:45,040 So there you are. It's the latest now. How does this relate to cosmology? 188 00:20:45,040 --> 00:20:50,860 Well, I already showed you all the relevant formula. It's very simple. I mean, this model is 100 years old. 189 00:20:50,860 --> 00:20:58,300 So the magnitude you heard about the that translates into a luminosity distance to this relationship here, 190 00:20:58,300 --> 00:21:07,030 and the luminosity distance is related to the parameters of the model like all these omega three, which are omega matter or omega lambda, et cetera. 191 00:21:07,030 --> 00:21:13,600 Of course, we can extract these parameters, therefore, by looking at the magnitude Rateliff relationship. 192 00:21:13,600 --> 00:21:18,910 But in fact, we don't really need to look at the model because acceleration is a kinematic done. 193 00:21:18,910 --> 00:21:23,710 It doesn't need you to have a theoretical model of the cosmology. 194 00:21:23,710 --> 00:21:29,020 You just do a Taylor expansion of this luminosity distance. The first term is the velocity. 195 00:21:29,020 --> 00:21:33,310 The second term is acceleration, which is defined like. 196 00:21:33,310 --> 00:21:38,080 So it was defined with the minus sign because people expected the universe to 197 00:21:38,080 --> 00:21:43,410 be decelerating and therefore they wanted you not to be a positive number. 198 00:21:43,410 --> 00:21:51,190 The surprises that people claim it's a negative number. And of course, the third term is for some reason it's called the jerk and so on. 199 00:21:51,190 --> 00:21:54,820 So we can do this Taylor expansion. This is a small graph. 200 00:21:54,820 --> 00:22:02,500 It has got no model assumptions whatsoever other than, of course, this as a in homogeneity that attention. 201 00:22:02,500 --> 00:22:09,970 So here is the data. So these people in 2014 for the first time made the data public. 202 00:22:09,970 --> 00:22:14,950 740 supernovae from various different observations, 203 00:22:14,950 --> 00:22:23,350 all assembled together analysed uniformly using the same template that you designate service by making the data public. 204 00:22:23,350 --> 00:22:29,000 That is what science should be. It should be open. We should all be able to play. 205 00:22:29,000 --> 00:22:36,590 And we can now take this data and ask if the divisions that were drawn earlier do bear up. 206 00:22:36,590 --> 00:22:41,360 By the way, this column is of the mass of the host galaxy that might have an influence as well. 207 00:22:41,360 --> 00:22:47,120 So this may not be the end of the story. There may be other predictions you need to make, as I've indicated here. 208 00:22:47,120 --> 00:22:51,840 But for the moment, this is the state of play. They they don't correlate with each other. 209 00:22:51,840 --> 00:23:02,230 They're pretty good of the. Now we got into this of three years ago yet, and Ensign was my master's student and he, you know, 210 00:23:02,230 --> 00:23:07,570 students these days have to go to proper courses on statistics which most of us have never had. 211 00:23:07,570 --> 00:23:14,590 And the first thing they learn is that, you know, you construct something called the likelihood, which is the probability that you are seeing. 212 00:23:14,590 --> 00:23:22,010 Is this given data for as you model? And in this case, the data that you're observing at the magnitudes, 213 00:23:22,010 --> 00:23:28,220 the stretch and the colour corrections for some set of parameters of the model, which I called Peter, 214 00:23:28,220 --> 00:23:33,860 and it is easy to say that this tape this can be decomposed into a product of 215 00:23:33,860 --> 00:23:38,480 stuff that depends on the cosmology and stuff that depends on the supernova, 216 00:23:38,480 --> 00:23:48,250 because surely the two don't talk to each other, so you can assume that the variables that control the unknowns are decoupled from each other. 217 00:23:48,250 --> 00:23:54,790 Now, yet they noticed immediately that when you plot the directions that the observers had been, you know, 218 00:23:54,790 --> 00:24:01,900 painstakingly amassed that day, in fact, a rather reasonably described by Gaussian distributions. 219 00:24:01,900 --> 00:24:05,710 This is not perfect, of course. And in principle, you can do it numerically. 220 00:24:05,710 --> 00:24:12,670 What matters is the area under the curve, but it Gaussian is a good description because what that means is that this likelihood, 221 00:24:12,670 --> 00:24:22,630 which affects factories as says that the observed distributions are drawn from some underlying distributions of these special colour parameters, 222 00:24:22,630 --> 00:24:25,720 which are determined by the physics of the supernova. 223 00:24:25,720 --> 00:24:31,870 And if what I see is a Gaussian, then it's very likely that the period distribution is also Gaussian. 224 00:24:31,870 --> 00:24:38,650 And therefore, I can assume all this stuff to be Gaussian, which makes life very easy because I know how to integrate the Gaussian. 225 00:24:38,650 --> 00:24:43,630 That's my training in theoretical physics going in the hand. 226 00:24:43,630 --> 00:24:53,170 This is not entirely a joke because being able to derive it likelihood without running a massive Markov chain Monte Carlo, you know, 227 00:24:53,170 --> 00:24:59,530 analysing and dimensional parameter space, which is the standard workhorse in cosmology, 228 00:24:59,530 --> 00:25:03,550 is a big advance because you can check that what you have done is correct. 229 00:25:03,550 --> 00:25:07,060 In fact, the programme is on the webpage. It's a small Python code. 230 00:25:07,060 --> 00:25:11,680 You can run it. And when we did that, we found something rather amazing. 231 00:25:11,680 --> 00:25:15,280 We found that the evidence is nowhere near as strong as had been claimed. 232 00:25:15,280 --> 00:25:16,210 In fact, 233 00:25:16,210 --> 00:25:26,090 the cantos and the omega cosmological constant verses matter plane came down and touched the line that separates acceleration from deceleration. 234 00:25:26,090 --> 00:25:30,890 And this was a bit of a surprise because we had been told that these photos are way up 235 00:25:30,890 --> 00:25:37,130 there and establish exhilaration to the level where it was deemed worthy of a Nobel prise. 236 00:25:37,130 --> 00:25:44,930 Three Sigma is nothing in particle physics. We have animals coming and going every day, which are three sigma. 237 00:25:44,930 --> 00:25:53,600 Now, people objected to this. First, they tried to claim that our analysis method was to use the words of Brad Schmid, one of the Nobel laureates. 238 00:25:53,600 --> 00:25:57,680 He tweeted that it was rather unorthodox. 239 00:25:57,680 --> 00:26:04,980 Yep, he was quite shocked and I, until I assured him that Brian, like me, had really never done a course in statistics in his life. 240 00:26:04,980 --> 00:26:12,830 OK. It is not unorthodox at all. It was in fact called the greatest advance in 20th century statistics. 241 00:26:12,830 --> 00:26:19,770 Then people would complain that, Oh look, these people only can get to this in an empty universe and the universe is not empty. 242 00:26:19,770 --> 00:26:28,640 It has mattered to the right of this. Plus, the sum of these two things should add up to one, so we should be somewhere down the line way diagonally. 243 00:26:28,640 --> 00:26:33,350 This is only true to the standard cosmological model. It not be true in general. 244 00:26:33,350 --> 00:26:38,930 You might have other terms in the Friedman equation. You might have terms corresponding to, you know, 245 00:26:38,930 --> 00:26:44,930 viscosity of the fluid filling the universe or to the so-called back reaction neutrino, whether it is. 246 00:26:44,930 --> 00:26:53,870 So that is not an argument. But let me move on because I'm conscious of the time and get to this point about questioning. 247 00:26:53,870 --> 00:27:00,620 The fundamental assumption is the universe actually isotropic now this is what we lecture to our students all the time, 248 00:27:00,620 --> 00:27:03,890 and it is manifestly not true if you look at the sky. 249 00:27:03,890 --> 00:27:12,320 So the furthest back you can look is that redshift 1000 surface, the so-called last scattering surface of the microwave background, 250 00:27:12,320 --> 00:27:20,900 and that shows a huge dipole how the sky is hotter than the other half and the temperature varies in a well-defined way, 251 00:27:20,900 --> 00:27:25,850 which was in fact first worked out by Stewart and Sharma. 252 00:27:25,850 --> 00:27:28,040 I, in fact, came to Oxford to work with the NIA. 253 00:27:28,040 --> 00:27:35,510 Sharma was also James, the supervisor, and the other people that wrote about it were Peebles and Wilkinson. 254 00:27:35,510 --> 00:27:41,120 Peebles got the Nobel prise for cosmology this year, so you can see this as the distinguished history, 255 00:27:41,120 --> 00:27:45,530 and this dipole is interpreted as due to our motion. 256 00:27:45,530 --> 00:27:53,270 It's seen as a special relativistic effect. In fact, this is now a standard problem for undergraduates to work out that formula. 257 00:27:53,270 --> 00:27:57,590 When you move through a part of black body radiation, what would you see now? 258 00:27:57,590 --> 00:28:03,020 As it happens, we are actually moving in the opposite direction as we go down the centre of the galaxy. 259 00:28:03,020 --> 00:28:08,240 So our velocity is actually a bit twice as much as 620. 260 00:28:08,240 --> 00:28:13,820 And amazingly, this is not just us, the entire local group in the Magellanic clouds, 261 00:28:13,820 --> 00:28:19,030 et cetera, we are all moving coherently towards the particular direction. 262 00:28:19,030 --> 00:28:24,430 Now, why are we moving if the universe was homogeneous, then of course, there is nowhere to move to. 263 00:28:24,430 --> 00:28:28,690 Every place is the same. There must be some lump of matter attracting us. 264 00:28:28,690 --> 00:28:34,450 That's the received wisdom and it was given a name which was called the Great Attractor. 265 00:28:34,450 --> 00:28:39,520 Right now we have looked for the Great Attractor and we should have found it within about 266 00:28:39,520 --> 00:28:44,050 100 megaparsec because the claim is that although there are these enormous entities, 267 00:28:44,050 --> 00:28:52,450 when you average on scales bigger than 100 megaparsec, then you can sensibly describe the universe as homogeneous. 268 00:28:52,450 --> 00:28:58,900 But in fact, we don't find that this is the actual distribution of galaxies in our neighbourhood. 269 00:28:58,900 --> 00:29:04,360 I've drawn for your convenience, a sphere of one hundred May opposites. We are at the centre here. 270 00:29:04,360 --> 00:29:12,310 We are moving in this direction towards the shape of a cluster. But that is not the Great Attractor, as we as I show you. 271 00:29:12,310 --> 00:29:16,180 And we are being fooled by something beyond shapely. 272 00:29:16,180 --> 00:29:23,410 And the further away this object is, the more massive it has to be in order to make us move with the velocity that we do. 273 00:29:23,410 --> 00:29:31,720 And that creates a problem for the standard cosmology. How do you create a lump of matter, which is that massive light? 274 00:29:31,720 --> 00:29:33,520 So there are many ways to frame it. 275 00:29:33,520 --> 00:29:43,780 My point is simply here to emphasise that the standard assumption and cosmological analysis is that we are a so-called Copernican observers. 276 00:29:43,780 --> 00:29:49,930 The universe looks isotropic and homogeneous to us. And this is the same, too. 277 00:29:49,930 --> 00:30:01,430 This is not the case. So let us try to get something out of the data, here is a catalogue of a previous catalogue of 557 supernovae shown on the sky. 278 00:30:01,430 --> 00:30:06,920 The sizes of the squares tell you whether they are brighter than or less bright than 279 00:30:06,920 --> 00:30:12,860 the brightness that they ought to be a standard candles for their known distances. 280 00:30:12,860 --> 00:30:17,240 So what is evident is that these things are not all zero. 281 00:30:17,240 --> 00:30:19,130 This is not a flat distribution. 282 00:30:19,130 --> 00:30:28,040 And when we do this, we can actually map out the Hubble fluid ratchet because we can see the shelves of galaxies around us, 283 00:30:28,040 --> 00:30:30,680 each one hosting some supernovae. 284 00:30:30,680 --> 00:30:38,420 And we can therefore determine whether there is a bulk motion which is reflected in the dipole anacetrapib of the supernovae. 285 00:30:38,420 --> 00:30:43,910 And these residuals are actually mapping out the peculiar velocity flow and peculiar velocity is a good 286 00:30:43,910 --> 00:30:50,300 thing because the previous plot that I showed you of the galaxies may not be all the matter that there is. 287 00:30:50,300 --> 00:30:54,710 In fact, we know that is five times more dark matter, which you don't see in that picture. 288 00:30:54,710 --> 00:31:02,000 The only way to know the influence of all the matter is wait for it to do its stuff gravitationally, 289 00:31:02,000 --> 00:31:08,480 cause things to move, and we measured those motions, right? So then we are sure that we are sensitive to all the matter. 290 00:31:08,480 --> 00:31:16,280 And in fact, we find when you do this analysis in different ranges of Ratchet, as you see here, that we are actually moving. 291 00:31:16,280 --> 00:31:25,730 There is a dipole anisotropy in roughly the same direction as the CMB, which is that black dot there, the dipole from this is much cruder. 292 00:31:25,730 --> 00:31:29,960 Is that blue patch and that is the 90 percent confidence level and so on. 293 00:31:29,960 --> 00:31:39,350 It's much less, well, determined, but what is clear is that that velocity with which we are moving is significantly higher than 294 00:31:39,350 --> 00:31:45,920 the value that you expect due to just random fluctuations in the so-called lambda CDR model, 295 00:31:45,920 --> 00:31:52,610 which is the standard model of the field. Now these divisions are large, but they are also uncertain. 296 00:31:52,610 --> 00:31:57,320 So within one standard deviation or so, as you can see here, they are consistent. 297 00:31:57,320 --> 00:32:03,470 So this did not make such an impact. Although you might worry that all this data is on one side of the plot. 298 00:32:03,470 --> 00:32:09,560 That is clearly some systematic issue, and others have also found such flows on a smaller scale. 299 00:32:09,560 --> 00:32:14,630 This was the deepest we went out to shapely, which is a two hundred and sixty opposites. 300 00:32:14,630 --> 00:32:22,730 And we did not find any info on two shapely on the other side, which means that the thing that is pulling us is even beyond shapely. 301 00:32:22,730 --> 00:32:25,340 And this was confirmed later by various people. 302 00:32:25,340 --> 00:32:33,770 I just show you one set of data from the biggest such peculiar flow catalogue, which was done also at the Anglo-Australian Telescope. 303 00:32:33,770 --> 00:32:38,840 And that is all still show you this the bulk velocity versus the radius. 304 00:32:38,840 --> 00:32:48,050 And you see that data, which is this point here, is well clear of the dashed line, which is the expectation in the standard lambda model. 305 00:32:48,050 --> 00:32:57,930 Our result was this Green Line pinpoint without Airbus with other large, they have done a much better job with the biggest sample. 306 00:32:57,930 --> 00:33:01,850 Recall that for all these objects, you need independent distance measurements. 307 00:33:01,850 --> 00:33:06,050 They use something called the fundamental plane, which is good to about 15 percent. 308 00:33:06,050 --> 00:33:10,880 So this has been a very field fraught with lots of uncertainties and errors. 309 00:33:10,880 --> 00:33:18,560 But it is now becoming pretty clear that there is a problem. We are moving much faster than we should expect to be. 310 00:33:18,560 --> 00:33:26,570 In fact, we can ask how likely it is that we are moving that fast out that far and we can interrogate this large 311 00:33:26,570 --> 00:33:32,390 simulations that our colleagues in astrophysics do of the formation of structure in the universe. 312 00:33:32,390 --> 00:33:36,710 And we can ask how often might we find ourselves to be in this position? 313 00:33:36,710 --> 00:33:41,660 And then they can quantify that statement earlier that we are not Copernican by saying 314 00:33:41,660 --> 00:33:46,940 that less than one percent of observers in such a universe should see what we see. 315 00:33:46,940 --> 00:33:49,460 We are definitely special. 316 00:33:49,460 --> 00:33:57,590 That, you know, people have a problem with being sort of know because we started cosmology by saying the Earth is at the centre of the universe. 317 00:33:57,590 --> 00:34:03,140 I think people are psychologically reluctant to accept that there is something, anything special about us. 318 00:34:03,140 --> 00:34:11,690 Of course, there's nothing special about us other than the fact that we are not the same as everybody else. 319 00:34:11,690 --> 00:34:15,620 The question is what does it mean for the inferences that we draw from the data? 320 00:34:15,620 --> 00:34:24,790 That is important because only by that route do we come to this amazing conclusion that 70 percent of the universe is dug in. 321 00:34:24,790 --> 00:34:33,680 Now, if you are in the bulk flow as this yellow patch here, then the Red Devils will tell you that you are not a Copernican observer. 322 00:34:33,680 --> 00:34:35,720 You are a so-called tilted observer. 323 00:34:35,720 --> 00:34:43,460 A terminology that was first introduced by a South African relative is George Ellis, who is in fact just retired to Oxford. 324 00:34:43,460 --> 00:34:48,980 So you have the advantage of discussions with him and for lack of time, 325 00:34:48,980 --> 00:34:57,830 I shall skip over the details of how we can then deduce that if you happen to be in such a batch and observing objects 326 00:34:57,830 --> 00:35:05,530 which are in the batch and our data shows that three quarters of the observed supernovae are within this bug flow. 327 00:35:05,530 --> 00:35:10,660 Then you might find that the deceleration perimeter that you measured is not just the usual one, 328 00:35:10,660 --> 00:35:14,830 but as a correction done because of the bulk flow, which is the opposite sign. 329 00:35:14,830 --> 00:35:21,090 And that means that you might inferred a negative number when in fact the number is actually positive globally. 330 00:35:21,090 --> 00:35:28,270 OK, now the guy who has deduced this lady was Gonzaga's in this this. 331 00:35:28,270 --> 00:35:35,560 He has tried to compute these terms with without success because the necessary astronomical measurements have not been made. 332 00:35:35,560 --> 00:35:38,440 But there's a clear prediction that if this is true, 333 00:35:38,440 --> 00:35:45,400 then we should see a dipole in the deceleration or acceleration parameter as well in the same direction that we are moving. 334 00:35:45,400 --> 00:35:52,360 Is that true? Well, amazingly enough, nobody has thought of looking for this until now. 335 00:35:52,360 --> 00:35:54,980 This is the sky distribution of the supernovae. 336 00:35:54,980 --> 00:36:04,600 Then the giant like, like a light analysis catalogue, jails 11 40 guys and they are distributed in redshift, so half of them are pretty close to us. 337 00:36:04,600 --> 00:36:13,520 These are the Sloan Digital Sky Survey. This is the Supernova Legacy Survey, disposable things that the Hubble Space Telescope at High shipped. 338 00:36:13,520 --> 00:36:19,310 This is the direction you are moving towards the cmb, these are the directions of the bird flu, which are uncertain, 339 00:36:19,310 --> 00:36:29,030 but close to what we discovered is that of course, this peculiar velocities are well known in cosmology and people have made corrections for them. 340 00:36:29,030 --> 00:36:35,510 But what we discovered is that the people who have made these corrections astronomers have assumed that the bulk 341 00:36:35,510 --> 00:36:42,980 flow simply disappears at a distance of about one hundred and fifty may about six about 500 million light years. 342 00:36:42,980 --> 00:36:48,470 But our data shows that it continues to at least twice the distance and shows no signs of falling off. 343 00:36:48,470 --> 00:36:52,430 Besides which you can't have a velocity, certainly stop and go to zero. 344 00:36:52,430 --> 00:37:01,580 There has to be continuity right now, thanks to collaborator Ramis, who is very good at digging out what people have done with that data. 345 00:37:01,580 --> 00:37:06,770 We were able to uncover this and show this picture of shame as it were. 346 00:37:06,770 --> 00:37:10,460 It shows that there has been some dodgy analysis going on. 347 00:37:10,460 --> 00:37:14,720 So we undo all that. And now we're in a position for the punchline, 348 00:37:14,720 --> 00:37:21,590 which is that the supernova do have a clear dipole in the inferred acceleration in roughly 349 00:37:21,590 --> 00:37:26,540 the same direction as the one that you're moving into the CMB that shows the dipole, 350 00:37:26,540 --> 00:37:31,250 and this shows the log likelihood, which is almost flat in that direction. 351 00:37:31,250 --> 00:37:38,780 And here we see the magnitude of the dipole, which is about eight OK in this, according to this definition, 352 00:37:38,780 --> 00:37:44,990 which is about 50 times bigger than the monopole, which is only about minus 0.1 eight or so. 353 00:37:44,990 --> 00:37:51,380 In fact, the monopole, which is what people thought all this time it was, is negligible. 354 00:37:51,380 --> 00:37:58,760 It's been set it to zero and you, in fact, you get a slightly better fate, according to the so-called Bayesian information criterion. 355 00:37:58,760 --> 00:38:05,820 So it's quite clear that what you're really seeing is the dipole along the direction of a local bulk flow and. 356 00:38:05,820 --> 00:38:12,720 In fact, as I mentioned, we can possibly have this zero of divided this scale by 10, 357 00:38:12,720 --> 00:38:17,220 otherwise this thing would just be a vertical line because this is totally negligible. 358 00:38:17,220 --> 00:38:20,610 It's consistent with zero at one point four sigma. 359 00:38:20,610 --> 00:38:29,370 So now whatever you might want to attribute to a cosmological constant or dark energy has a significance of 1.4 sigma. 360 00:38:29,370 --> 00:38:38,430 OK? Ask yourself if you would believe that whereas this has a significance of three point nine sigma, which is rather convincing. 361 00:38:38,430 --> 00:38:47,400 So we conclude that cosmic acceleration is simply an artefact of our being located in a bulk flow rather than due to anything fundamental. 362 00:38:47,400 --> 00:38:57,570 Goodbye to dark energy. And I therefore can come to my conclusions, which is that, as I've just stated, this is all data driven. 363 00:38:57,570 --> 00:38:59,070 I've not used any theory. 364 00:38:59,070 --> 00:39:07,350 There was a theoretical motivation following from the work of cosmologists or tilted observers, but it's completely empirical. 365 00:39:07,350 --> 00:39:12,750 So you have to ask yourself if you believe this by simply running the same programme 366 00:39:12,750 --> 00:39:18,440 that you have done on the data and checking that you get the same results. Now, 367 00:39:18,440 --> 00:39:28,040 this is the new ingredient that we are in a bug flu that is stressing out much further than you expect and that you have to deal with that adequately, 368 00:39:28,040 --> 00:39:35,690 that makes us non-competitive and observers. We therefore cannot afford to analyse the data as we have been doing all this time 369 00:39:35,690 --> 00:39:40,580 using a model that was set up one hundred years ago and is no longer fit for purpose. 370 00:39:40,580 --> 00:39:46,310 We have to take that into account, and when we do so, we find rather astonishing results. 371 00:39:46,310 --> 00:39:54,320 This is due to appear. So this is literally hot off. The press is due to appear on Monday in astronomy and astrophysics letters. 372 00:39:54,320 --> 00:40:03,710 And the standard assumptions are questionable, but the good news is that some of those objects that are showed in that slide the nuclear satellite, 373 00:40:03,710 --> 00:40:10,940 the Large Synoptic Survey Telescope, the square kilometre array chips, they mentioned that our department is involved in these. 374 00:40:10,940 --> 00:40:15,920 This will enable definitive tests because they will provide us the kind of data that we need. 375 00:40:15,920 --> 00:40:24,580 So to date, we have all these 740 or 2000 supernovae, and this history will give us 10000 supernovae per year. 376 00:40:24,580 --> 00:40:30,860 Right. So I should live that long, but this is something worth waiting for. 377 00:40:30,860 --> 00:40:39,650 So I leave you with the words of two famous physicist, Lev Landau, who was often wrong himself, by the way. 378 00:40:39,650 --> 00:40:45,530 But that's what he thought of cosmology. But I think the last word really belongs to Wolfgang Pauli. 379 00:40:45,530 --> 00:40:58,203 It's not even wrong. Thank you.