WEBVTT

1
00:00:00.670 --> 00:00:01.460
STFC-RAL-CR03-RAL-R61-2.01: Bye, too.

2
00:00:04.500 --> 00:00:05.780
STFC-RAL-CR03-RAL-R61-2.01: Fantastic.

3
00:00:06.000 --> 00:00:10.500
STFC-RAL-CR03-RAL-R61-2.01: You have to… Hash, hash.

4
00:00:13.700 --> 00:00:15.430
STFC-RAL-CR03-RAL-R61-2.01: Seaboard.

5
00:00:15.650 --> 00:00:16.460
STFC-RAL-CR03-RAL-R61-2.01: Hash.

6
00:00:18.300 --> 00:00:20.469
STFC-RAL-CR03-RAL-R61-2.01: Oh, from that, from the tablet.

7
00:00:21.900 --> 00:00:23.730
STFC-RAL-CR03-RAL-R61-2.01: And… Yeah.

8
00:00:25.590 --> 00:00:42.099
STFC-RAL-CR03-RAL-R61-2.01: Okay, welcome to the seminar. We are pleased to have a seminar by one of our PhD students, and with that, we would like to advertise that this seminar is open to any future PhD students from our department.

9
00:00:43.380 --> 00:00:58.210
STFC-RAL-CR03-RAL-R61-2.01: Okay, let me introduce our speaker, Abi. She has recently submitted her PhD at Royal Holloway University of London, where she worked on improving the treatment of neutrino cross-section.

10
00:00:58.940 --> 00:01:06.699
STFC-RAL-CR03-RAL-R61-2.01: Systematic uncertainties for due to support precision long baseline observation measurements.

11
00:01:06.750 --> 00:01:20.909
STFC-RAL-CR03-RAL-R61-2.01: Before this, I've completed a master's degree by a researcher at Durham University, studying the origin of gamma-ray emission from globular clusters.

12
00:01:20.910 --> 00:01:32.340
STFC-RAL-CR03-RAL-R61-2.01: and an undergraduate degree at the University of Oxford, where her final year project focused on neon reference platinum and the snow-plus experiment.

13
00:01:32.690 --> 00:01:36.079
STFC-RAL-CR03-RAL-R61-2.01: And today, she will present a thesis work.

14
00:01:43.770 --> 00:02:01.420
STFC-RAL-CR03-RAL-R61-2.01: Okay, hi everybody, yeah, I'm Adam, and yeah, so my title of my talk today is Off-Actus and Model Independent, Near-to-Far Neutrino Interaction Extrapolation with Deep Rhythm, which is a bit of a mouthful, but, hopefully by the end of this talk, I'll be able to explain a bit more about what this means.

15
00:02:02.080 --> 00:02:03.630
STFC-RAL-CR03-RAL-R61-2.01: X.

16
00:02:07.030 --> 00:02:07.950
STFC-RAL-CR03-RAL-R61-2.01: Okay.

17
00:02:08.240 --> 00:02:13.649
STFC-RAL-CR03-RAL-R61-2.01: So before we go any further, I want to give a little brief introduction to neutrinos.

18
00:02:14.050 --> 00:02:29.619
STFC-RAL-CR03-RAL-R61-2.01: So neutrinos are the most abundant massive particle in the universe, so the only other particle that is more abundant is photons, which don't have any mass. So this diagram here shows the 18 orders of magnitude of energy over which neutrinos are produced.

19
00:02:29.920 --> 00:02:38.220
STFC-RAL-CR03-RAL-R61-2.01: So you can see this is a very large scale, and you can see on the y-axis here, you've got the distance, the different energy neutrinos have had to travel in order to reach us.

20
00:02:38.220 --> 00:02:48.430
STFC-RAL-CR03-RAL-R61-2.01: So you can see on the top right-hand corner here, you've got ultra-high energy neutrinos, which travel gigaparsecs, so from extragalactic sources, the origins of which we don't yet know completely.

21
00:02:48.430 --> 00:02:58.860
STFC-RAL-CR03-RAL-R61-2.01: And on the bottom left-hand corner, you've got neutrinos that have traveled much smaller distances, so much closer to home, so you have geoneutrinos, from… come from radioactive decaying rocks, and also from

22
00:02:58.900 --> 00:03:00.690
STFC-RAL-CR03-RAL-R61-2.01: Nuclear reactors.

23
00:03:01.100 --> 00:03:11.209
STFC-RAL-CR03-RAL-R61-2.01: So they've been observed to come in three flavors, so electron, muon, and tau neutrinos, and each of those have an associated charge, lepton with them as well.

24
00:03:11.290 --> 00:03:24.460
STFC-RAL-CR03-RAL-R61-2.01: So we've observed that they oscillate, between different flavors, so this was observed, using solar neutrinos and atmospheric neutrino observations, and this is only possible because neutrinos, have a non-zero mass.

25
00:03:25.060 --> 00:03:43.879
STFC-RAL-CR03-RAL-R61-2.01: So the neutrino mass is not actually something that's predicted in the standard model of particle physics, not part of it. So actually, neutrinos having mass and neutrino oscillations are direct evidence of, that there's physics beyond the standard model and particle physics we currently have, and there's, yeah, there's more to the story.

26
00:03:44.220 --> 00:03:54.369
STFC-RAL-CR03-RAL-R61-2.01: So there's lots of things we don't know about neutrinos. We don't know the mass ordering, we don't know the CP violation, and at the bottom of this plot here, you can see

27
00:03:54.420 --> 00:04:11.220
STFC-RAL-CR03-RAL-R61-2.01: how useful neutrinos are for studying lots of different physics and answering some of these questions at different ranges of energies. So I… this talk is going to be focused on this… the range of energy in the middle here for long-base light neutrinos, where we can study things like CP violation and mass ordering.

28
00:04:14.880 --> 00:04:26.009
STFC-RAL-CR03-RAL-R61-2.01: Okay, so a little bit of, maths and some theory. So neutrinos are produced and detected in weak interactions, in their flavor eigenstates, so electron, muon, and tau neutrinos.

29
00:04:26.130 --> 00:04:36.059
STFC-RAL-CR03-RAL-R61-2.01: But each flavor state is actually a superposition of the mass eigenstates, so the flavor eigenstates are not the definite, mass eigenstates, for neutrinos.

30
00:04:36.280 --> 00:04:50.999
STFC-RAL-CR03-RAL-R61-2.01: So this means that as they travel, the mass… the phases of these mass states can change, and they evolve as they travel. And this is how we get neutrino oscillations, and how the electron neutrino is able to change later to muon or tau.

31
00:04:52.000 --> 00:05:05.880
STFC-RAL-CR03-RAL-R61-2.01: So the way you relate mass, flavor eigenstates to the, mass eigenstates is through the PMNS matrix. So the PMNS matrix is a unitary matrix, and what that means is we can write it as a product of three, rotation matrices.

32
00:05:05.960 --> 00:05:12.050
STFC-RAL-CR03-RAL-R61-2.01: So the parameters in these rotation matrices are the mixing angles between the different, flavors.

33
00:05:12.820 --> 00:05:18.010
STFC-RAL-CR03-RAL-R61-2.01: sorry, between the different mass states and also the CPU-violating phase metrics.

34
00:05:18.180 --> 00:05:37.180
STFC-RAL-CR03-RAL-R61-2.01: So we can actually learn what these parameters are, within the rotation matrices, by looking at different energy neutrinos. So we can learn using atmospheric and non-baseline neutrinos to learn the parameters in the first matrix, with reactor and non-baseline neutrinos to learn,

35
00:05:37.270 --> 00:05:48.449
STFC-RAL-CR03-RAL-R61-2.01: sine of theta 1, 3, and charge people what the value of delta CP is, and then using solar and reactive neutrinos to measure the parameters of its biometrix here.

36
00:05:51.050 --> 00:06:00.200
STFC-RAL-CR03-RAL-R61-2.01: So we also know that neutrinos, so we know that neutrinos oscillate, and we want to know how frequently they oscillate, or the oscillation probability.

37
00:06:00.200 --> 00:06:12.179
STFC-RAL-CR03-RAL-R61-2.01: So, because this is a talk on June, I'm focusing the equations using muon neutrinos, because this is what the June neutrino being predominantly been made of.

38
00:06:12.420 --> 00:06:16.740
STFC-RAL-CR03-RAL-R61-2.01: So the first equation is, ME1, neutrino,

39
00:06:16.750 --> 00:06:32.260
STFC-RAL-CR03-RAL-R61-2.01: disappearance. So you can see that this is controlled by two parameters. You've got the theta 2, 3, so the mixing between the second and third, mass eigenstates, and this delta, 3-1 parameter. So this is controlled, by the mass splitting between the third and first, states.

40
00:06:32.260 --> 00:06:39.320
STFC-RAL-CR03-RAL-R61-2.01: And also L over E, so the length that the neutrinos have traveled, divided by the energy of the neutrino.

41
00:06:39.530 --> 00:06:49.119
STFC-RAL-CR03-RAL-R61-2.01: And this L over E quantity is known as the baseline. So the, the theta parameter controls, how… how close the oscillations are together.

42
00:06:49.120 --> 00:07:03.160
STFC-RAL-CR03-RAL-R61-2.01: And then the delta, controls where the oscillation maximum is. So for conveniently folks, so for 1300 kilometer at baseline, you see, you see that the oscillation maximum at 2.6 GeV.

43
00:07:04.060 --> 00:07:22.480
STFC-RAL-CR03-RAL-R61-2.01: The second equation, so this is, electron, neutrino appearance. So this is slightly more important for long baseline studies. So you have three terms in here. So you have the delta… the first term is the delta 1, 3 term, so this is kind of the leading order term, and this controls,

44
00:07:22.480 --> 00:07:26.839
STFC-RAL-CR03-RAL-R61-2.01: How big… how big the oscillations are. Also,

45
00:07:27.200 --> 00:07:33.629
STFC-RAL-CR03-RAL-R61-2.01: Yeah, and also exactly the position, the position of where and depth of where the minimum are.

46
00:07:33.960 --> 00:07:42.239
STFC-RAL-CR03-RAL-R61-2.01: The second term is a CP violating term. This is how we… we only term what the delta CP parameter appears.

47
00:07:43.220 --> 00:07:47.989
STFC-RAL-CR03-RAL-R61-2.01: And, the third term is just a small correction, which is the solar term.

48
00:07:48.240 --> 00:08:08.420
STFC-RAL-CR03-RAL-R61-2.01: So you'll notice in all of these, there's also this parameter A, so this is to do with the matter density. So as electron neutrinos travel through the Earth, they're gonna interact, with electrons within, the matter they're traveling through, and so this becomes quite important when you have a long baseline, so this is not a term you can not really…

49
00:08:09.890 --> 00:08:21.719
STFC-RAL-CR03-RAL-R61-2.01: So the math effect, grows with the baseline, and it means that you're sensitive to the, delta M squared 3-1, so the mass splitting. And I should also say that the sine of delta 31 and the

50
00:08:21.800 --> 00:08:30.699
STFC-RAL-CR03-RAL-R61-2.01: value of delta CP is going to switch, depending on whether you're using neutrinos or antineutrinos. So this is how you can,

51
00:08:31.000 --> 00:08:34.560
STFC-RAL-CR03-RAL-R61-2.01: That about adults, trainers and I can eat trainers.

52
00:08:34.789 --> 00:08:40.279
STFC-RAL-CR03-RAL-R61-2.01: You can also see on the plot on the bottom here, for different values of delta CP.

53
00:08:40.590 --> 00:08:50.989
STFC-RAL-CR03-RAL-R61-2.01: you can see that the position of the, the minimum, so how much suppression you get from the CP interference term, varies depending on what the value of, delta CP is.

54
00:08:54.080 --> 00:09:13.880
STFC-RAL-CR03-RAL-R61-2.01: So, oscillation parameters. So what do you know currently? So this plot here is showing the constraints that we have for the neutrino oscillation parameters, as a function of time, so starting in 1998 and going all the way to 2025. So you can see the enormous progress we've already made with a number of these parameters.

55
00:09:14.260 --> 00:09:19.619
STFC-RAL-CR03-RAL-R61-2.01: So the first two, mixing angles we know to buy good precision, so a couple of percent.

56
00:09:20.060 --> 00:09:28.229
STFC-RAL-CR03-RAL-R61-2.01: The third mixing angle, feature 2, 3, we're still not sure what octant is it, so what side of 45 degrees?

57
00:09:29.030 --> 00:09:40.559
STFC-RAL-CR03-RAL-R61-2.01: We know the first mass splitting to a few percent, but the second one we do not know the sign of. And these are… this is what's going to determine what the mass ordering of the neutrinos, is.

58
00:09:40.820 --> 00:09:49.749
STFC-RAL-CR03-RAL-R61-2.01: The sign of delta CP is probably still the biggest mystery, and this is what the next generation of long baseline experiments is hoping to determine.

59
00:09:49.890 --> 00:09:57.639
STFC-RAL-CR03-RAL-R61-2.01: And then I suppose, even bigger question, beyond Delta CP is, is the PM&S, description even correct? So…

60
00:09:58.040 --> 00:10:07.160
STFC-RAL-CR03-RAL-R61-2.01: the… these parameters are based on the assumption that PM and F is the correct description for neutrinos, but to be honest, we don't even completely know that.

61
00:10:07.450 --> 00:10:16.969
STFC-RAL-CR03-RAL-R61-2.01: So, we need to, basically, we need to precisely understand all these things and, validate that PM and F is the correct description in order to understand neutrinos.

62
00:10:19.320 --> 00:10:25.230
STFC-RAL-CR03-RAL-R61-2.01: Okay, so next I'm going to move on to talking about, the experiment I work on, which is June.

63
00:10:26.070 --> 00:10:35.270
STFC-RAL-CR03-RAL-R61-2.01: So as you probably all know, so June stands for Deep Underground Neutrino Experiment, and it's an experiment that's currently under construction in the US.

64
00:10:35.530 --> 00:10:43.970
STFC-RAL-CR03-RAL-R61-2.01: So you, currently, you have a neutrino beam, a high-power energy neutrino beam, that's produced at Thermilab, which is in Illinois.

65
00:10:44.340 --> 00:10:54.069
STFC-RAL-CR03-RAL-R61-2.01: Close to the source, you have a near detector, which detects the being before the neutrinos had a chance to oscillate. Neutrinos travel through the Earth.

66
00:10:54.900 --> 00:10:58.660
STFC-RAL-CR03-RAL-R61-2.01: And then the oscillators that neutrino,

67
00:10:58.840 --> 00:11:07.309
STFC-RAL-CR03-RAL-R61-2.01: will be detected at the fire detector, which is in South Dakota at the Stamford Underground Research Facility. It's about one and a half kilometres underground.

68
00:11:08.060 --> 00:11:24.680
STFC-RAL-CR03-RAL-R61-2.01: So June's been designed to do a wide range of, physics, so in terms of the long baseline programs, we recently have a near detector and a fire detector. We're hoping we can determine the mass ordering, and delta CP, to good precision,

69
00:11:24.750 --> 00:11:32.759
STFC-RAL-CR03-RAL-R61-2.01: within 10 years, and, part of that depends on exactly where in parameter space Delta CP is, whether nature is kind to us or not.

70
00:11:33.050 --> 00:11:50.420
STFC-RAL-CR03-RAL-R61-2.01: Beyond, just beam physics, there's a other wide variety of physics we can do. So the fire detector in particular is going to be, sensitive to supernova neutrinos, and we should get a unique sensitivity to newly, electron neutrinos that come from supernova.

71
00:11:50.420 --> 00:12:02.999
STFC-RAL-CR03-RAL-R61-2.01: With good directional information, as well as doing more studies for solar and atmospheric neutrinos. And beyond that, testing some beyond the standard model physics, so light testing, is the PM and S matrix even correct?

72
00:12:03.240 --> 00:12:07.050
STFC-RAL-CR03-RAL-R61-2.01: Yeah, so there's a wide range of physics that June is going to be capable of.

73
00:12:08.490 --> 00:12:27.559
STFC-RAL-CR03-RAL-R61-2.01: So on top of that, June… the June collaboration is the largest international neutrino collaboration, that's ever existed. There's over 1,500 scientists in it, from over 39 countries… 39 countries. This is the, meeting… a picture from the collaboration meeting, from the collaboration, at CERN from a few years ago.

74
00:12:27.630 --> 00:12:31.839
STFC-RAL-CR03-RAL-R61-2.01: And yeah, as you already know, there's a major UK contribution,

75
00:12:32.130 --> 00:12:35.010
STFC-RAL-CR03-RAL-R61-2.01: There's also been UK scientists working on gene.

76
00:12:36.320 --> 00:12:52.590
STFC-RAL-CR03-RAL-R61-2.01: Okay, so the next few slides, I'm going to talk about the different, components of the June experiment. So the neutrino beam, so it's produced at the LB&F, beamline CERNA lab, and it's producing 120 GCB protocols, which is high energy, from the main injector.

77
00:12:53.000 --> 00:13:02.830
STFC-RAL-CR03-RAL-R61-2.01: So these protons are… proton beam is produced. It's collided with a graphite target, which causes the protons, to decay… to decay into pions and kayons.

78
00:13:04.570 --> 00:13:16.650
STFC-RAL-CR03-RAL-R61-2.01: The pions and kons are charged, so they can be focused using magnetic focusing horns, and eventually these will decay into neutrinos and their associated leptons, which is how we get our neutrino being.

79
00:13:18.040 --> 00:13:31.430
STFC-RAL-CR03-RAL-R61-2.01: So depending on… so we're going to start with… it's going to be a very high-powered neutrino beam, the most powerful neutrino beam that's ever been built, reaching up to possibly beyond 2 megawatts, in the second phase of the experiment.

80
00:13:32.580 --> 00:13:45.630
STFC-RAL-CR03-RAL-R61-2.01: So the magnetic horns can be used, to change, the composition of the beam. So by… you can change it from neutrino mode to antineutrino mode, depending on which pions you're focusing, towards the beam line and which ones you're not.

81
00:13:46.690 --> 00:14:00.669
STFC-RAL-CR03-RAL-R61-2.01: On top of this, so you can see in this, in this plot on, on the right here, so this is the flux, expected from the near-detect beam. So you can see it's a broadband beam, so it covers a wide range of, energies.

82
00:14:01.200 --> 00:14:20.249
STFC-RAL-CR03-RAL-R61-2.01: And it peaks at about 2.6, sorry, 2.5 GeV, which is conveniently, the right energy to get the maximum oscillation, at 113… 1300 kilometers, which is where the fire detector is. So it's all been designed specifically,

83
00:14:20.390 --> 00:14:21.219
STFC-RAL-CR03-RAL-R61-2.01: To do not.

84
00:14:24.580 --> 00:14:43.409
STFC-RAL-CR03-RAL-R61-2.01: Okay, so moving on to the near detector. So the near detector is about 500, 600 meters away from, where the beam is produced, and it's going to measure the unoscillated beam and also, median neutrino, interactions, and it's going to be important for constraining, properties like the flux and cross-section.

85
00:14:43.410 --> 00:14:49.429
STFC-RAL-CR03-RAL-R61-2.01: Which I'll talk about a bit more later. So this is a schematic of what the near-detector will look like.

86
00:14:49.430 --> 00:14:56.119
STFC-RAL-CR03-RAL-R61-2.01: The most important component, probably, is NDLAR, which is the liquid arbor time projecting chamber,

87
00:14:56.640 --> 00:15:01.399
STFC-RAL-CR03-RAL-R61-2.01: Which is where you will measure this… measure the neutrino interactions.

88
00:15:02.020 --> 00:15:17.060
STFC-RAL-CR03-RAL-R61-2.01: You also have, TMS, which will be, which is a muon spectrometer, which will be replaced by Gaseous argon detector NDGAR in Phase 2, which will then be sort of the more capable, more precise near-detector for neutrino interactions.

89
00:15:17.270 --> 00:15:28.999
STFC-RAL-CR03-RAL-R61-2.01: On top of this, you have, SAND, which is, continuous beam monitored, because we need to, if we're not doing precision measurements, then we need to really understand what the neutrino beam is doing at all times.

90
00:15:29.980 --> 00:15:46.140
STFC-RAL-CR03-RAL-R61-2.01: Additional, additionally, and I will talk more about this later because it's important, June Prison, so, TMS and NDLAR are able to move relative to beam axis. So yeah, so you can move laterally, laterally to the beam axis, and this is going to be important,

91
00:15:46.240 --> 00:15:50.560
STFC-RAL-CR03-RAL-R61-2.01: for sampling different neutrino energies. I'll talk loads more about that later.

92
00:15:52.290 --> 00:16:12.050
STFC-RAL-CR03-RAL-R61-2.01: And finally, the fire detector. So the fire detector is currently under construction, underground at the Stamford Underground Research Facility. This is the picture of the fully excavated cabin, cabin, sorry. Apparently, it's huge, I haven't seen it in real life, but apparently this is, yeah, this picture doesn't really do a justice about how big it is.

93
00:16:13.310 --> 00:16:23.130
STFC-RAL-CR03-RAL-R61-2.01: The file detector is going to consist of, have the same technology as the near detector, so they're going to be LAR TPCs again. But you're going to have, four of them,

94
00:16:23.900 --> 00:16:33.369
STFC-RAL-CR03-RAL-R61-2.01: In order to measure, sufficient interactions, because… because, you're going to have a much smaller flux, given that the beam has traveled, 2500 kilometers.

95
00:16:33.470 --> 00:16:38.310
STFC-RAL-CR03-RAL-R61-2.01: And in that square. So you have a… there'll be a horizontal drift,

96
00:16:38.620 --> 00:16:53.269
STFC-RAL-CR03-RAL-R61-2.01: fire detector, as well as a vertical adrift fire detector in Phase 1. And then in phase two, there will be two additional modules. Exact technology of those, I think, has yet to be completely confirmed, but eventually there will be for

97
00:16:55.930 --> 00:17:11.349
STFC-RAL-CR03-RAL-R61-2.01: Okay, so how do we do a long baseline, oscillation analysis? So what we have to do is we have to measure the neutrino event rate as a function of the reconstructed neutrino energy, and we need to do that at the NIV detector and the FAR detector, so we're able to compare them.

98
00:17:12.040 --> 00:17:21.879
STFC-RAL-CR03-RAL-R61-2.01: So, the bed rate that you're going to get is a bit complicated, so it's an… it's an integral as a function of the true neutrino energy, which is something you can't actually measure.

99
00:17:22.470 --> 00:17:33.180
STFC-RAL-CR03-RAL-R61-2.01: talk about that later. And it's… it's a function of, various sets of parameters we have to model. So these include, the flux, the green flux.

100
00:17:33.350 --> 00:17:43.860
STFC-RAL-CR03-RAL-R61-2.01: The cross-section model, which is how the neutrinos interact with the argon in the detector, and also the detector efficiency. So how, yeah, how our detector

101
00:17:43.860 --> 00:17:53.489
STFC-RAL-CR03-RAL-R61-2.01: While our detector is able to measure particles, and also the smearing, so how the reconstructed neutrino energy translates onto the true neutrino energy that we can't measure.

102
00:17:53.950 --> 00:18:08.289
STFC-RAL-CR03-RAL-R61-2.01: And then it's the same as the file detector, except we have a slightly different box, because we're 31 kilometers away, different detector, and different smearing, and also the oscillation probability, which is what we're trying to determine to learn the oscillation parameters.

103
00:18:08.670 --> 00:18:16.990
STFC-RAL-CR03-RAL-R61-2.01: So what we have to do is we measure the event rate of the near detector and car detector, we use the ND… we use the NID detector,

104
00:18:17.050 --> 00:18:29.200
STFC-RAL-CR03-RAL-R61-2.01: event rate to constrained model parameters. We can then propagate, the model to predict, the flux, sorry, the fire detector event rate for different sets of oscillation parameters.

105
00:18:29.770 --> 00:18:43.039
STFC-RAL-CR03-RAL-R61-2.01: And then, well, once we've got the prediction, we can then compare what we actually observe at the file detector, find what is the best fit, and then measure, extract what our best fit for the different observation parameters are.

106
00:18:45.160 --> 00:18:46.230
STFC-RAL-CR03-RAL-R61-2.01: So…

107
00:18:47.260 --> 00:18:57.649
STFC-RAL-CR03-RAL-R61-2.01: why can't we… so why can't we not just… so why is it complicated to take the near detector and the file detector event rates and compare them? Because it's quite challenging.

108
00:18:57.920 --> 00:19:15.479
STFC-RAL-CR03-RAL-R61-2.01: So you can't simply take a ratio, unfortunately, between the near detector and the file detector. So there's… that's for two… two main reasons. So the first is that the fire detector event rate is different to the emitted spectrum. You can see that in the integrals. So this is… this is what the file detector spectrum would look like compared to the near detector.

109
00:19:15.490 --> 00:19:24.760
STFC-RAL-CR03-RAL-R61-2.01: And the cross-section effect, which is the only set of parameters that's the same at near and far, because they're both liquid argon experiments. So,

110
00:19:24.930 --> 00:19:30.530
STFC-RAL-CR03-RAL-R61-2.01: The near-detector effects don't get answered automatically, from the fire detector cross-section.

111
00:19:30.660 --> 00:19:49.340
STFC-RAL-CR03-RAL-R61-2.01: The second reason is because the true neutrino energy is not the same as the reconstructed neutrino energy. This is kind of very clear of the file detector. So these are drawn from the gene Monte Carlo, but they show the reconstructed neutrino energy event spectra compared to the true neutrino energy.

112
00:19:49.580 --> 00:19:54.549
STFC-RAL-CR03-RAL-R61-2.01: And what you can see, is that they look quite different. So, in particular.

113
00:19:54.660 --> 00:20:03.809
STFC-RAL-CR03-RAL-R61-2.01: You can see at the oscillation at Maxima at 2.3 GeV, the Ricoh and the tree are in different… are in different spaces.

114
00:20:04.240 --> 00:20:19.139
STFC-RAL-CR03-RAL-R61-2.01: So the Rico, the dip kind of fills it. But this, this dip, the oscillation maximum, is very important for working out, the mass split between 3 and 2, and also its depth gives the, one of the mixing angles. So we really need to get this smearing right.

115
00:20:19.640 --> 00:20:29.729
STFC-RAL-CR03-RAL-R61-2.01: But it's why we can't… it is a bit complicated, and we need to, model things carefully in order to translate from the near detector, to the far detector.

116
00:20:32.310 --> 00:20:35.450
STFC-RAL-CR03-RAL-R61-2.01: Okay, and I said I'd talk about June Prism.

117
00:20:35.580 --> 00:20:43.399
STFC-RAL-CR03-RAL-R61-2.01: So the near detector moves off-axis. It's planned that it will move up to 28.5 meters, relative to the beam axis.

118
00:20:43.490 --> 00:21:02.299
STFC-RAL-CR03-RAL-R61-2.01: So it's the same detector, but it's just moving, so we've got the same interaction physics, but at different neutrino flux at every position. So the reason for this is to do with the pion decay kinematics. So as you, so the energy of the pion depends on where it is relative to the beam axis.

119
00:21:03.880 --> 00:21:17.119
STFC-RAL-CR03-RAL-R61-2.01: This just comes from special relativity, and the energy of the neutrino that is produced from the pion is related to the energy of that pion, as well as the angle it is relative to the neutrino beam.

120
00:21:18.010 --> 00:21:31.110
STFC-RAL-CR03-RAL-R61-2.01: So you get this… you get this equation here, and this is what I plotted on this diagram here. So you can see when you're on axis, so the dark… the dark blue line, the neutrino energy is directly proportional to the pione energy.

121
00:21:31.670 --> 00:21:48.119
STFC-RAL-CR03-RAL-R61-2.01: So they're very correlated, but as you get further off-axis, so out to about 28 meters, the relationship is not quite so strong. So you can see that for a given, given energy, the energy of the neutrinos produced is much more similar for every pile of energy.

122
00:21:50.860 --> 00:21:59.719
STFC-RAL-CR03-RAL-R61-2.01: So what that means for the… yeah, so what that means for the neutrino flux as a function of off-axis position, which is what this plot is showing here, so…

123
00:21:59.720 --> 00:22:12.609
STFC-RAL-CR03-RAL-R61-2.01: Is that as you go off axis, so on axis, the neutrino flux that you get is a broadband beam, so this is a dark purple here. So you can see it's wide, and it's, yeah, peaks at 2.3 GBP.

124
00:22:13.650 --> 00:22:25.210
STFC-RAL-CR03-RAL-R61-2.01: As you move further off axis, the peak of the… peak of where the neutrino flux is, gets, smaller, and also the flux gets narrower. So this is going from blue to yellow.

125
00:22:25.860 --> 00:22:34.289
STFC-RAL-CR03-RAL-R61-2.01: So this matters because it means that you can control, basically, what your… what event spectrum your near-detector is seeing.

126
00:22:34.670 --> 00:22:44.390
STFC-RAL-CR03-RAL-R61-2.01: Because on axis, if you're trying to model neutrino interactions, and you're trying to see all energies at once, that can become very complicated. So, yeah.

127
00:22:44.680 --> 00:22:50.010
STFC-RAL-CR03-RAL-R61-2.01: Prism allows is a way of, constraining, some of these characters.

128
00:22:52.550 --> 00:22:56.419
STFC-RAL-CR03-RAL-R61-2.01: Okay, so next section, I'm going to talk about neutrino interactions.

129
00:22:57.400 --> 00:23:13.820
STFC-RAL-CR03-RAL-R61-2.01: So when the neutrino enters our detector and interacts with the argon nuclei, there's two types of interactions that can happen. So the first is a charged current interaction, where the neutrino… so the neutrino hits the nucleus, and you get a charged electron.

130
00:23:14.890 --> 00:23:23.879
STFC-RAL-CR03-RAL-R61-2.01: Which is the same flare as the absolute neutrino, and these are the events we use to measure oscillations, and this is the dominant source of interactions that we see.

131
00:23:23.900 --> 00:23:39.640
STFC-RAL-CR03-RAL-R61-2.01: There's also neutral current interactions, where you get an incoming neutrino that scatters, or you get an outgoing neutrino from the same flavor. So the neutrino leaves the detector and lists… so you can't actually detect the neutrino,

132
00:23:39.940 --> 00:23:47.350
STFC-RAL-CR03-RAL-R61-2.01: So, what you see is an energy loss in your detector. So this is actually an important background for oscillation, samples.

133
00:23:48.570 --> 00:23:52.289
STFC-RAL-CR03-RAL-R61-2.01: So, inside the nucleus,

134
00:23:53.100 --> 00:24:07.340
STFC-RAL-CR03-RAL-R61-2.01: So the target nucleus is not a free nucleon. So what that means is the nucleons inside the nucleus are bound, so they can't escape, but they're also moving around. So this means that the interactions going on inside the nucleus are a little bit of a black box.

135
00:24:08.100 --> 00:24:14.459
STFC-RAL-CR03-RAL-R61-2.01: So the haddrons produced inside the nucleus could re-interact or be reabsorbed, and this is called final state interactions.

136
00:24:15.600 --> 00:24:27.509
STFC-RAL-CR03-RAL-R61-2.01: So, this is why the true neutrino energy, so the energy of this incoming neutrino, is hard to measure from just measuring the energy from the outgoing particles.

137
00:24:27.590 --> 00:24:38.029
STFC-RAL-CR03-RAL-R61-2.01: So you can measure the energy of the outgoing particles to reconstruct the neutrino energy, but what you have to be aware of is that there will be some smearing because you don't know exactly what's happened inside the nucleus.

138
00:24:40.180 --> 00:24:48.039
STFC-RAL-CR03-RAL-R61-2.01: So, for charge-current interaction types, there's four main processes that could have happened inside the nucleus that we need to think about.

139
00:24:48.160 --> 00:24:54.340
STFC-RAL-CR03-RAL-R61-2.01: So the first one is part of the elastic scattering, so where a neutrino scatters off a child's neptron.

140
00:24:55.210 --> 00:25:03.500
STFC-RAL-CR03-RAL-R61-2.01: This is dominant at low neutrino energies. You can also have multi-nuclear scattering, sort of 2P2H interactions.

141
00:25:03.950 --> 00:25:18.900
STFC-RAL-CR03-RAL-R61-2.01: At the energy range that's important for June, particularly important for June, around the oscillation maxima, you get resonant pion production as well. So you, excite a delta baryon inside the nucleus, which then decays into a pion, and that has a particular signature.

142
00:25:19.200 --> 00:25:31.089
STFC-RAL-CR03-RAL-R61-2.01: And then, as you get up to higher energies, there's a growing fraction of interactions that break apart the nucleus, and you get hydraulic showers, and this is called deep inelastic scattering.

143
00:25:34.000 --> 00:25:47.170
STFC-RAL-CR03-RAL-R61-2.01: So, multiple processes can happen for any neutrino energy. So this, again, is the dune flux at the near detector and the file detector as a function of energy, and it's broken up in terms of interaction type.

144
00:25:47.860 --> 00:26:01.649
STFC-RAL-CR03-RAL-R61-2.01: So you can see for the… see, you've got all the different interaction types I've just mentioned, and neutral current interactions as well, but what you can see is that any given energy, you can't necessarily know what type of interaction is going to have occurred.

145
00:26:02.680 --> 00:26:10.890
STFC-RAL-CR03-RAL-R61-2.01: So, right across the entire energy range of June, multiple interaction processes could have happened, and not one of them is really dominant.

146
00:26:11.880 --> 00:26:13.900
STFC-RAL-CR03-RAL-R61-2.01: So, each process looks different.

147
00:26:14.090 --> 00:26:29.930
STFC-RAL-CR03-RAL-R61-2.01: And also, the way that the missing energy, so the way the true neutrino energy from the incoming neutrino translates to the energy of the outgoing particles is going to be different. So, in order to understand the true neutrino energy, we need to model all these processes correctly.

148
00:26:32.210 --> 00:26:51.639
STFC-RAL-CR03-RAL-R61-2.01: So, I've talked about reconstructing neutrino energy, but what do we actually measure in our detector? So I'm going to call what we measure in the detector, the visible neutrino energy, which is basically the sum of the energy from the electrons and hydronic particles that you're able to measure in your detector, comes from the ionized tracks and the argon.

149
00:26:51.640 --> 00:26:54.379
STFC-RAL-CR03-RAL-R61-2.01: And the electron drift readout and the RTPC.

150
00:26:55.390 --> 00:27:05.459
STFC-RAL-CR03-RAL-R61-2.01: So this is different from the reconstructed neutrino energy, because this is built from the particles that are actually leaving the nucleus, with some uncertainty about what's happening inside the nucleus that we can't see.

151
00:27:05.820 --> 00:27:13.850
STFC-RAL-CR03-RAL-R61-2.01: Whereas invisible energy is what the detector is actually able to see and reconstruct, from the positive… from the deposited charge.

152
00:27:14.810 --> 00:27:19.490
STFC-RAL-CR03-RAL-R61-2.01: So how closely these are related kind of depends on what type of interaction has happened.

153
00:27:20.320 --> 00:27:35.330
STFC-RAL-CR03-RAL-R61-2.01: So I guess we can think about this in terms of two scenarios. So, best case scenario is we end up with two outgoing particles, such as a muon and a proton, that are easy to measure, because the tracks are very distinct, and it's easy to tell what particles they are.

154
00:27:35.330 --> 00:27:42.790
STFC-RAL-CR03-RAL-R61-2.01: The tracks are nice and straight, nice and clean, and this, in this case, the visible energy is going to be closely related to the reconstructed neutrino energy.

155
00:27:45.160 --> 00:28:03.530
STFC-RAL-CR03-RAL-R61-2.01: worst case scenario, could be you get a much more, sort of, messy interaction, so you could have multi-pion FSI interaction, so you could have a pion, some photons and protons, some hadrons interacting inside the nucleus, and the result of this, and this one is taken from a microboom, perhaps, so…

156
00:28:03.570 --> 00:28:07.999
STFC-RAL-CR03-RAL-R61-2.01: Another neutrino interaction experiment, but what you basically see is a bit of a mess.

157
00:28:08.160 --> 00:28:18.339
STFC-RAL-CR03-RAL-R61-2.01: This is hard to reconstruct within the detector, and this probably means that the visible energy, so the energy you see, is probably a lot less than the reconstructed neutrino energy.

158
00:28:22.340 --> 00:28:27.529
STFC-RAL-CR03-RAL-R61-2.01: Okay, so now I'm going to talk about source of uncertainty in a long baseline oscillation analysis.

159
00:28:28.450 --> 00:28:33.360
STFC-RAL-CR03-RAL-R61-2.01: Okay, so there's two types of uncertainties, there's statistical and systematic.

160
00:28:33.510 --> 00:28:52.490
STFC-RAL-CR03-RAL-R61-2.01: So previously, so current and baseline, experiments like TGK and Nova are limited by statistics. So how many neutrinos have actually been able to measure? This is not going to be a problem for June, because we've got the most powerful neutrino beam ever built, and tens of kilotons of liquid argon to measure the neutrino interactions.

161
00:28:52.980 --> 00:29:00.659
STFC-RAL-CR03-RAL-R61-2.01: However, what that means is within a few years, the June statistical uncertainty is going to become comparable to the systematic uncertainty.

162
00:29:01.080 --> 00:29:10.580
STFC-RAL-CR03-RAL-R61-2.01: So what this means is, for June, we need to have a more careful treatment of systematic uncertainties, than we've had to do in the past, at TGK and NOVA.

163
00:29:12.850 --> 00:29:29.870
STFC-RAL-CR03-RAL-R61-2.01: So the sources of systematic uncertainty that we're going to be able to constrain at the near detector come from the flux, cross-section, and the efficiency. So the neutrino beam, source of uncertainty could come from hadron production, which we don't know too well, and also things to do with the beam alignment and configuration.

164
00:29:29.920 --> 00:29:40.429
STFC-RAL-CR03-RAL-R61-2.01: In this talk, I'm going to be focusing on neutrino interactions, so these are the uncertainties to do with how the neutrinos interacted with the nuclei, and also what energy is visible.

165
00:29:40.930 --> 00:29:43.770
STFC-RAL-CR03-RAL-R61-2.01: And these are arguably some of the hardest to model.

166
00:29:46.280 --> 00:29:51.180
STFC-RAL-CR03-RAL-R61-2.01: Okay, so we have a cross-section uncertainty model currently, but there are some problems with it.

167
00:29:51.290 --> 00:30:05.940
STFC-RAL-CR03-RAL-R61-2.01: So the first problem, set of problems with it is that the models actually disagree. So the theoretical models we have to describe neutrino interactions over the energy range that gene sees, there's not actually one theory that is capable of describing the whole energy range.

168
00:30:06.790 --> 00:30:12.769
STFC-RAL-CR03-RAL-R61-2.01: The second problem is the available models we have at a given energy all disagree with each other in different ways.

169
00:30:13.530 --> 00:30:24.329
STFC-RAL-CR03-RAL-R61-2.01: And there are already observed tensions from other experiments, such as MicroBoon and Minerva, between the data they see and models that they got through, the very same ones that Unum used, so we know this is a big problem.

170
00:30:24.710 --> 00:30:31.639
STFC-RAL-CR03-RAL-R61-2.01: And the fourth problem is if the models don't converge before June starts taking near-detector data, this is going to be a problem for the analysis.

171
00:30:33.700 --> 00:30:40.869
STFC-RAL-CR03-RAL-R61-2.01: So the second problem, apart from the models disagreeing, is that the uncertainty models, so the uncertainties surrounding the models that we have.

172
00:30:40.990 --> 00:30:42.320
STFC-RAL-CR03-RAL-R61-2.01: are not sufficient.

173
00:30:42.650 --> 00:30:48.470
STFC-RAL-CR03-RAL-R61-2.01: So we build the duty in Monte Carlo, so the interactions from the current, from current interaction models.

174
00:30:48.960 --> 00:30:59.549
STFC-RAL-CR03-RAL-R61-2.01: But the theoretical uncertainties that come from the models are… yes, that come from the models are not big enough to cover the tension, between different models.

175
00:30:59.770 --> 00:31:10.149
STFC-RAL-CR03-RAL-R61-2.01: So if you can imagine Model A and Model B having associated uncertainty with them, but it's not… the uncertainties are not big enough, to… to have them both.

176
00:31:10.410 --> 00:31:20.380
STFC-RAL-CR03-RAL-R61-2.01: So what this means is if the data that June takes falls outside the uncertainty bound the theory allows, we could have a big disagreement between the Monte Carlo prediction and the data we're collecting.

177
00:31:20.630 --> 00:31:40.250
STFC-RAL-CR03-RAL-R61-2.01: And what that would mean is then we'd struggle to get a good fit to the near detector, and if we can't get a good fit for the near detector data, it's not clear about how we'll be able to proceed to do an oscillation analysis, which would mean… could mean, worst case scenarios, we're not able to make precision measurements of the oscillation practice, which is obviously something we all want.

178
00:31:41.730 --> 00:31:57.600
STFC-RAL-CR03-RAL-R61-2.01: So the current cross-section uncertainty model has got about 50 parameters that come from the theoretical models and interesting processes that we mentioned. It's based on what's in the event generator genie and has been tuned for different bits of experimental data.

179
00:31:57.770 --> 00:32:10.509
STFC-RAL-CR03-RAL-R61-2.01: They're implemented as response functions, which basically means they're pre-calculated, before you do the fitting, and these diagrams are just a sort of schematic of what the response functions, look like.

180
00:32:10.920 --> 00:32:12.130
STFC-RAL-CR03-RAL-R61-2.01: I thought that.

181
00:32:15.490 --> 00:32:21.150
STFC-RAL-CR03-RAL-R61-2.01: Okay, so next thing I'm going to be talking about is, how do we learn e-truth if we can't measure it?

182
00:32:21.790 --> 00:32:30.129
STFC-RAL-CR03-RAL-R61-2.01: So we need to learn Etrue, because it's the true neutrino energy, because it's what the oscillation parameters are determined as a function of.

183
00:32:30.250 --> 00:32:34.410
STFC-RAL-CR03-RAL-R61-2.01: Which you saw in the, probability equations at the beginning.

184
00:32:35.320 --> 00:32:53.389
STFC-RAL-CR03-RAL-R61-2.01: So therefore, we need to understand how true neutrino energy maps to the reconstructed neutrino energy. And this is kind of shown in this histogram for the near detector for the Monte Carlo gene users currently. You can see it's quite diagonal, so this means that the true neutrino energy is highly correlated to what we reconstruct.

185
00:32:53.400 --> 00:32:57.359
STFC-RAL-CR03-RAL-R61-2.01: But there is some smearing around it, so there's some events around diagonals.

186
00:32:58.370 --> 00:33:07.779
STFC-RAL-CR03-RAL-R61-2.01: So we really need to understand this mapping, and we also need to understand what uncertainty we have around this mapping, because if we understand that the near detector, we can propagate it to the smart detector.

187
00:33:08.070 --> 00:33:11.529
STFC-RAL-CR03-RAL-R61-2.01: We also know that the mapping is going to be different at near and far.

188
00:33:14.350 --> 00:33:26.729
STFC-RAL-CR03-RAL-R61-2.01: So, how does PRISM help us learn this mapping? So, because PRISM is going to be measuring different reconstructed neutrino energies, what that means is they're going to be measuring different portions of this map.

189
00:33:26.730 --> 00:33:39.199
STFC-RAL-CR03-RAL-R61-2.01: So you can see on 2 meters on axis, you're kind of observing something that's very similar to what you get on axis, but as you get up to 28 meters, you're only observing events that are much, closer to that diagonal at a modular

190
00:33:40.270 --> 00:33:57.469
STFC-RAL-CR03-RAL-R61-2.01: So what this means is that instead of trying to learn the entire truth to Rico, from a broadband spectrum, you can learn it at much narrower slices and then piece them all together, which means that you're not… you're going to be able to learn the map without being sensitive to the cross-section model biases.

191
00:34:00.620 --> 00:34:14.560
STFC-RAL-CR03-RAL-R61-2.01: Okay, so what does the current model do to this mapping? So, taking a couple of examples here, so what these plots are showing are what happens if you take the uncertain… if you take the parameter at its nominal value.

192
00:34:14.690 --> 00:34:22.580
STFC-RAL-CR03-RAL-R61-2.01: And then take the parameter at one sigma uncertainty that we think it is, and then take the ratio of the event rates and see what happens.

193
00:34:23.530 --> 00:34:30.440
STFC-RAL-CR03-RAL-R61-2.01: So you can see that these, plots are, you know, broadly the same color, with some small amount of variation.

194
00:34:31.130 --> 00:34:40.819
STFC-RAL-CR03-RAL-R61-2.01: So what this is showing is that these parameters make some large-scale changes to the map, to the truth to reach out energy. So they affect events at all energies.

195
00:34:42.120 --> 00:34:57.649
STFC-RAL-CR03-RAL-R61-2.01: They also all look vaguely similar. So what this means is that the model that we currently have can only move the truth of the Ricoh energy mapping in directions it allows, and those directions are kind of making broad changes, everywhere.

196
00:35:00.680 --> 00:35:05.849
STFC-RAL-CR03-RAL-R61-2.01: So, the current… so with the current model, we want to know what constraint we can get on each route.

197
00:35:06.340 --> 00:35:11.250
STFC-RAL-CR03-RAL-R61-2.01: So these… so in the… in reconstructing neutrino energy.

198
00:35:12.170 --> 00:35:21.199
STFC-RAL-CR03-RAL-R61-2.01: So the plot on the middle here shows the size of the uncertainty. So you can see that in the reconstructed neutrino energy, this uncertainty is tiny.

199
00:35:21.310 --> 00:35:31.210
STFC-RAL-CR03-RAL-R61-2.01: And that is expected because we have such high statistics of the near detector, so that means we get negligible residual error when we fit in reconstructed neutrino energy.

200
00:35:31.820 --> 00:35:33.410
STFC-RAL-CR03-RAL-R61-2.01: So that's all fine.

201
00:35:33.520 --> 00:35:52.499
STFC-RAL-CR03-RAL-R61-2.01: But when we look at the true neutrino energy spectrum, which is what we've predicted from the RICO, you see that you get a very similar size uncertainty. So this is telling us, and from our current cross-section uncertainty model is telling us, that we should know the uncertainty on each true to less than 1%.

202
00:35:53.490 --> 00:35:56.919
STFC-RAL-CR03-RAL-R61-2.01: A lot less than 1% at lower end, higher energies.

203
00:35:57.380 --> 00:35:59.919
STFC-RAL-CR03-RAL-R61-2.01: The trouble is, we know from our model that

204
00:36:00.130 --> 00:36:13.720
STFC-RAL-CR03-RAL-R61-2.01: We know that we don't… we're not certain of our cross-section uncertainty model to this, level of, to this level, so we know we don't know at 1%, but the result that it's giving us is suggesting that we do.

205
00:36:14.920 --> 00:36:20.040
STFC-RAL-CR03-RAL-R61-2.01: So if we have real data, what this would mean is we would get a very small uncertainty.

206
00:36:20.440 --> 00:36:30.000
STFC-RAL-CR03-RAL-R61-2.01: And if our model is wrong, then we're confidently, predicting the wrong, true neutrino energy with a big uncertain… with a very small uncertainty.

207
00:36:30.880 --> 00:36:38.509
STFC-RAL-CR03-RAL-R61-2.01: And we can't do anything else with the current, parameter and set of cross-section parameters, because we don't… because it doesn't allow us to.

208
00:36:38.680 --> 00:36:54.080
STFC-RAL-CR03-RAL-R61-2.01: So the risk with this is that we get an unaccounted-for bias that we pass to the file detector with a false confidence. So we get the true neutrino energy spectrum wrong, or the truth to Ricoh mapping wrong, and we get it wrong with very high vision because the uncertainty found is so small.

209
00:36:55.030 --> 00:36:56.300
STFC-RAL-CR03-RAL-R61-2.01: So we wouldn't know.

210
00:36:58.320 --> 00:37:11.240
STFC-RAL-CR03-RAL-R61-2.01: So, what this means is that we need a new approach, because if the near-detected data doesn't fall… falls outside the uncertainty envelope, and the cross-section model can't absorb the difference, then we might get biased oscillation parameters.

211
00:37:11.470 --> 00:37:23.900
STFC-RAL-CR03-RAL-R61-2.01: So we need to come up with a new approach. We want the new approach to do something different from what we currently have, so we wanted to make local changes to the trees to Rico mapping to not affect all events or all energies, all events.

212
00:37:25.300 --> 00:37:42.410
STFC-RAL-CR03-RAL-R61-2.01: Given that the cross-section models are all wrong, we want it to be model agnostic, so going to build it in terms of interaction kinematics, so how the neutrino energy, momentum, etc, which are just products of the interaction, rather than being dependent on what particular theoretical models are chosen.

213
00:37:43.690 --> 00:37:53.859
STFC-RAL-CR03-RAL-R61-2.01: And the third thing we wanted to be is flexible, so we can capture the discrepancies that we're going to see between our Monte Carlo and the data that we can't, look, yeah, alone we can't describe.

214
00:37:56.620 --> 00:38:03.189
STFC-RAL-CR03-RAL-R61-2.01: So, I've tried to come up with a little, sort of diagram, little animation to show what this is doing.

215
00:38:03.800 --> 00:38:23.180
STFC-RAL-CR03-RAL-R61-2.01: So if you imagine, this is… this is the data that we collect. So this truth to Rico mapping looks a little bit different to the one I showed from my Monte Carlo, but it's… it's kind of very similar. So this is the actual truth-to-Rico mapping that we might get from our data. So this is what our uncertainty model has to reproduce and be able to move the Monte Carlo

216
00:38:23.290 --> 00:38:24.920
STFC-RAL-CR03-RAL-R61-2.01: In order to reproduce.

217
00:38:27.680 --> 00:38:29.490
STFC-RAL-CR03-RAL-R61-2.01: So, with the current model.

218
00:38:30.000 --> 00:38:35.279
STFC-RAL-CR03-RAL-R61-2.01: We have a few large parameters that are able to make big changes across the mapping.

219
00:38:35.550 --> 00:38:47.419
STFC-RAL-CR03-RAL-R61-2.01: So, you get a few large fixed shapes… a few large fixed shapes that can move, but you can't reconstruct the, the… your actual data, no matter how you arrange them.

220
00:38:50.200 --> 00:38:54.789
STFC-RAL-CR03-RAL-R61-2.01: So what we want is we want a new approach that's going to be able to,

221
00:38:55.270 --> 00:39:05.639
STFC-RAL-CR03-RAL-R61-2.01: to make small, small, smaller changes, and do them on more local, in more local energy regions. So the idea is that as each piece moves, and each piece is varied.

222
00:39:05.760 --> 00:39:15.440
STFC-RAL-CR03-RAL-R61-2.01: they can reproduce the, what your data is showing. So you can see that this is reproducing something that looks rather like what our, data is showing.

223
00:39:15.880 --> 00:39:18.440
STFC-RAL-CR03-RAL-R61-2.01: You can see that this is nowhere… nowhere near.

224
00:39:19.170 --> 00:39:21.449
STFC-RAL-CR03-RAL-R61-2.01: So this is what we have currently.

225
00:39:21.600 --> 00:39:23.259
STFC-RAL-CR03-RAL-R61-2.01: This is what we're aiming for.

226
00:39:27.850 --> 00:39:34.430
STFC-RAL-CR03-RAL-R61-2.01: Okay, so this next section of the talk, I'm going to talk about how… my attempt to do this, and how I've… how I've done it.

227
00:39:35.620 --> 00:39:40.610
STFC-RAL-CR03-RAL-R61-2.01: So, there's lots of different ways we could bin in terms of… you could,

228
00:39:40.850 --> 00:39:59.200
STFC-RAL-CR03-RAL-R61-2.01: categorize your cross-section model in terms of interaction kinematics. So one way of doing this, and probably the simplest way of doing this, is choosing a 2D… two parameters, which are the true neutrino energy, which we want because we're doing an oscillation analysis, and this is what we want to learn, and also the bias, neutrino energy bias.

229
00:40:00.080 --> 00:40:18.399
STFC-RAL-CR03-RAL-R61-2.01: Which is basically the difference between the true lepton and hadron energies, minus the true neutrino energy. So this is basically correlated with the energy that's not calibrated. Calorometrically available in our detector.

230
00:40:19.590 --> 00:40:25.429
STFC-RAL-CR03-RAL-R61-2.01: Because, yeah, this is a useful quantity note, but it tells us about how much of the neutrino energy we might

231
00:40:26.220 --> 00:40:37.439
STFC-RAL-CR03-RAL-R61-2.01: So if I make the… so if I make the event rate in terms of these parameters, so each, each, square rectangle that you can see is one of these parameters.

232
00:40:37.590 --> 00:40:43.760
STFC-RAL-CR03-RAL-R61-2.01: I've chosen the… chosen them so that they… the parameters and the binning kind of follows the features of the event, right?

233
00:40:43.910 --> 00:40:46.800
STFC-RAL-CR03-RAL-R61-2.01: Melissa's ended up with 600,

234
00:40:47.480 --> 00:40:50.700
STFC-RAL-CR03-RAL-R61-2.01: Cross-section parameters in total that are model independent.

235
00:40:50.750 --> 00:41:08.279
STFC-RAL-CR03-RAL-R61-2.01: And this is what they look like in the 1D project, because this is what each looks like. We have chosen the binning so that it looks and represents the true neutrino energy spectrum quite well. And then the biased energy and the peaks that you're seeing come from the pion production, and of course the elastic peak from the interaction.

236
00:41:11.930 --> 00:41:27.839
STFC-RAL-CR03-RAL-R61-2.01: So how do these new parameters impact the truth to read here mapping? Well, you can see that actually their effects are much more localized. So before, with the… the traditional… the current set of cross-section parameters, they were affecting the entire truth-to-reader spectrum.

237
00:41:27.910 --> 00:41:40.700
STFC-RAL-CR03-RAL-R61-2.01: Whereas these, again, are the one sigma shifts, from the nominal… doing the nominal and the one sigma shift of the parameter. You can see that they are making much more small local changes, which are these red blocks that you're seeing here.

238
00:41:40.700 --> 00:41:48.639
STFC-RAL-CR03-RAL-R61-2.01: You can see some of them affect higher E2 energies, low EU bias, and some of them are affecting up to 1 million regions, so…

239
00:41:48.730 --> 00:41:51.970
STFC-RAL-CR03-RAL-R61-2.01: about, like, 3.5 to 4 GEV.

240
00:41:53.490 --> 00:41:59.420
STFC-RAL-CR03-RAL-R61-2.01: And these are making, yeah, able to make localized changes to the Puerto Rico mapping.

241
00:41:59.530 --> 00:42:08.990
STFC-RAL-CR03-RAL-R61-2.01: So now, in theory, we have a way to describe the local discrepancies between the data in Monte Carlo in a way that the theoretical uncertainties we have are not able to do.

242
00:42:11.180 --> 00:42:24.480
STFC-RAL-CR03-RAL-R61-2.01: So, I've shown what they look like and how they're impacting the true variables, but what we actually measure and our data in is reconstructed variables, which correlate with the energy that we've measured in our sector.

243
00:42:24.660 --> 00:42:36.989
STFC-RAL-CR03-RAL-R61-2.01: So each of these parameters are controlling and making different variations to the reconstructed neutrino energy. So you can see the effect is smeared out. So, for example, this parameter here, which is about

244
00:42:37.100 --> 00:42:41.860
STFC-RAL-CR03-RAL-R61-2.01: This one, which is just below 3GB, is affecting events

245
00:42:42.230 --> 00:42:47.669
STFC-RAL-CR03-RAL-R61-2.01: In a similar region, but the effect is smeared out, and you see some effect at lower energies as well.

246
00:42:48.830 --> 00:42:54.360
STFC-RAL-CR03-RAL-R61-2.01: So the effect in the observable space is broader, and you see a more, sort of, diluted change.

247
00:42:55.170 --> 00:42:59.509
STFC-RAL-CR03-RAL-R61-2.01: So, yeah, so the change you see in local space is different to the reconstructed.

248
00:43:01.980 --> 00:43:15.439
STFC-RAL-CR03-RAL-R61-2.01: Okay, so all… all the, plots I'm about to show, were from my, from my PhD analysis, were conducted in MAC3. Mac3 is a Bayesian, fitting framework for neutrino oscillation analyses.

249
00:43:15.440 --> 00:43:25.279
STFC-RAL-CR03-RAL-R61-2.01: It's been going quite a while. It was originally developed for T2K, and, yeah, all the plots used in this analysis, use MAC3 and 18 inference.

250
00:43:28.980 --> 00:43:36.150
STFC-RAL-CR03-RAL-R61-2.01: Okay, so with the new parameterization that I'm using, how does that affect the true neutrino energy uncertainty?

251
00:43:36.160 --> 00:43:52.000
STFC-RAL-CR03-RAL-R61-2.01: So this is a plot from a few slides ago, so just as a reminder, this is the uncertainty bound you get on the true neutrino energy, where you use the traditional cross-section parameters from theory, and you get the less than 1% uncertainty across basically the entire energy.

252
00:43:52.000 --> 00:43:57.119
STFC-RAL-CR03-RAL-R61-2.01: So you've got no residual cross-section uncertainty to propagate to the plant detection.

253
00:43:58.010 --> 00:44:02.069
STFC-RAL-CR03-RAL-R61-2.01: However, if you use a new approach to the new 600 parameters.

254
00:44:02.430 --> 00:44:06.620
STFC-RAL-CR03-RAL-R61-2.01: You can see now the uncertainty that you're getting on the true neutrino energy sector.

255
00:44:06.910 --> 00:44:24.469
STFC-RAL-CR03-RAL-R61-2.01: goes, yeah, goes between about 4% and 11%. So this is a significant residual uncertainty in each room that we can propagate the file detector, and also means that any, any, uncertainties that we have in our model can be absorbed by some… by those uncertain band.

256
00:44:25.700 --> 00:44:36.689
STFC-RAL-CR03-RAL-R61-2.01: And this is a much more realistic uncertainty, so 4% to 11%, compared to the theoretical models. So this is a more realistic, uncertainty for retroot, than we currently get.

257
00:44:39.560 --> 00:44:41.010
STFC-RAL-CR03-RAL-R61-2.01: So,

258
00:44:41.470 --> 00:44:53.049
STFC-RAL-CR03-RAL-R61-2.01: I've talked about Zoom Prism quite a lot in this talk already. So how do we use, how can we use, Zoom Prism, with this… this new approach for cross-section systematics?

259
00:44:53.180 --> 00:45:10.120
STFC-RAL-CR03-RAL-R61-2.01: So these plots here are comparisons between the posterior standard deviation, so the constraint you get post-fit on the parameters, for a fit when you're just using on-axis data, so neutrinos that have been measured directly in front of the beam axis.

260
00:45:10.690 --> 00:45:15.050
STFC-RAL-CR03-RAL-R61-2.01: Compared to a fit where you take 50% that have been measured on axis.

261
00:45:15.160 --> 00:45:23.019
STFC-RAL-CR03-RAL-R61-2.01: And the other 50% neutrinos that have been split between the other off-axis positions, so I've almost, like, 50-50 on-axis versus off-axis.

262
00:45:23.400 --> 00:45:41.740
STFC-RAL-CR03-RAL-R61-2.01: So hopefully you can see on the screen here, so this, this bottom, the one that's on-axis only, is significantly more yellow and green, so this corresponds to a bigger posterior uncertainty compared to the off… on passive axis, which the uncertainty's a lot more reduced, so about 20% on.

263
00:45:42.560 --> 00:45:57.879
STFC-RAL-CR03-RAL-R61-2.01: And this covers… so it's the same number of events, but this extra constraint comes directly from the fact that by using prison data, we're able to reconstruct… we're able to get different neutrino fluxes, which allows us to get a tide constraint on the truth to Rico mapping.

264
00:45:57.940 --> 00:46:02.589
STFC-RAL-CR03-RAL-R61-2.01: And thereby, getting a better constraint on the cross-sectional basis.

265
00:46:03.390 --> 00:46:07.720
STFC-RAL-CR03-RAL-R61-2.01: So, yeah, this method, we get a bigger uncertainty,

266
00:46:08.170 --> 00:46:14.250
STFC-RAL-CR03-RAL-R61-2.01: From the parameters, and then constrain it and get a more realistic concept using the off-axis data as well.

267
00:46:16.210 --> 00:46:29.229
STFC-RAL-CR03-RAL-R61-2.01: So, I've talked about, so we want to propagate this uncertainty to the far detector, so these plots here are comparisons. We've taken the uncertainty to get to the near detector, and then made a prediction out of the far detector's going through.

268
00:46:29.660 --> 00:46:44.960
STFC-RAL-CR03-RAL-R61-2.01: So you can see this is oscillated. You can see that the residual uncertainty, again, is about 8-17%, which falls to about 3-10% with off-axis, but this is at about the same level as the near-detector prediction. So all that uncertainty,

269
00:46:45.490 --> 00:46:54.900
STFC-RAL-CR03-RAL-R61-2.01: The realistic error bound you get and learn at the near detector, you're able to, transfer that information to the star detector, and this means you can,

270
00:46:55.220 --> 00:47:02.349
STFC-RAL-CR03-RAL-R61-2.01: extract your oscillation parameters, more confidently, with a better confidence on the precision that you can record them to.

271
00:47:05.060 --> 00:47:25.020
STFC-RAL-CR03-RAL-R61-2.01: So, how well do we learn the true story of mapping in a fit? So this histogram here is a ratio of the near, the far, so yeah, it's a ratio of the true to reconstructed, posterior uncertainty, for the fit, for the on-plus off-axis fit divided by the on-axis fit.

272
00:47:26.100 --> 00:47:44.400
STFC-RAL-CR03-RAL-R61-2.01: So, where you can see blue, what that means is that the on plus off axis is giving a tighter constraint in that region of truth to Rico mapping compared to the on-axis only. So you can see that it's pretty much blue everywhere, with a couple of red, red bins, which are regions where there are less statistics anyway.

273
00:47:45.390 --> 00:47:59.210
STFC-RAL-CR03-RAL-R61-2.01: So, yeah, this is showing that the off-axis… including the off-axis data allows you to pin down the trace to Rico mapping, with a tight constraint for the near detector, so we… we're learning that uncertainty envelope, better.

274
00:48:01.550 --> 00:48:16.190
STFC-RAL-CR03-RAL-R61-2.01: And then we can take that constraint and then extract it to the bar detector. And again, you can see a very similar story here. Map is mostly blue, so this is showing that the improvement you get when you include off-axis data is quite significant, across all energies.

275
00:48:17.380 --> 00:48:27.070
STFC-RAL-CR03-RAL-R61-2.01: So with the off-axis data, you learn the file detection intruder Ricoh mapping, more precisely than we would have done if we just had, events on axis.

276
00:48:30.880 --> 00:48:31.990
STFC-RAL-CR03-RAL-R61-2.01: So…

277
00:48:32.170 --> 00:48:38.240
STFC-RAL-CR03-RAL-R61-2.01: the plots I've shown so far, I've said that I've spent 50% of the time on axis and 50% time

278
00:48:38.430 --> 00:48:47.090
STFC-RAL-CR03-RAL-R61-2.01: Off-axis, and split the time equally between every position, between 1 meter and 28 meters.

279
00:48:48.040 --> 00:49:00.460
STFC-RAL-CR03-RAL-R61-2.01: But we don't have to do that. So the time spent on access, relative to off-axis, I'm going to refer to as the run plan. And actually, the run plan is important for, optimizing gene sensitivity.

280
00:49:00.880 --> 00:49:08.890
STFC-RAL-CR03-RAL-R61-2.01: So, I've investigated some different potential run-pound scenarios, for PRISM, Out of competitors.

281
00:49:09.810 --> 00:49:11.760
STFC-RAL-CR03-RAL-R61-2.01: So, kind of…

282
00:49:11.950 --> 00:49:23.869
STFC-RAL-CR03-RAL-R61-2.01: before… before doing these studies, I kind of had two things in mind. So, more time on axis. So, more time on axis means you have higher statistics, because that's where most of the neutrino events are, so you get more constraining power.

283
00:49:25.100 --> 00:49:29.919
STFC-RAL-CR03-RAL-R61-2.01: So what this means is if you spent too much time on taxes, you would lose some statistics.

284
00:49:30.360 --> 00:49:43.870
STFC-RAL-CR03-RAL-R61-2.01: However, more time off access means that you get to sample, more events at lower energies, as we saw in TRIO Prison Flux, a few slides ago, and you get a better constraint on the trees to reca mapping.

285
00:49:44.880 --> 00:49:51.669
STFC-RAL-CR03-RAL-R61-2.01: So how we split this on axis versus off-axis is a sensitivity, study that I did.

286
00:49:53.140 --> 00:50:09.859
STFC-RAL-CR03-RAL-R61-2.01: So you can compare the run plant, at the near detector, so these coloured lines are a little bit hard to see. So blue is on-axis only, and then the yellow, green, orange, and pink are different, run plant scenarios, so different proportions of time spent on-axis and off-axis.

287
00:50:09.940 --> 00:50:18.820
STFC-RAL-CR03-RAL-R61-2.01: I think what's really clear, first of all, is that the uncertainty for on-axis only is much bigger than any time, any of the ones that include some off-axis data.

288
00:50:19.690 --> 00:50:28.640
STFC-RAL-CR03-RAL-R61-2.01: So this is showing that, yeah, spend cutoff access massively benefits the constraint, you get on your E2 posterior and the NID detector.

289
00:50:30.040 --> 00:50:37.979
STFC-RAL-CR03-RAL-R61-2.01: The exact proportions possibly don't matter, but if you look at the pink line, which is spent… is 50% on-axis.

290
00:50:38.050 --> 00:50:52.610
STFC-RAL-CR03-RAL-R61-2.01: and 50% of the time off-axis, but I'm only going up to 12 meters off-axis, you can see that the uncertainty in the pig band is quite a lot bigger than the uncertainty, from the other brand plans, which go to the full 28.5 meters off.

291
00:50:52.680 --> 00:50:58.470
STFC-RAL-CR03-RAL-R61-2.01: axis. So this is showing that spending time further off axis is really, really important, if we want to constrain the tree.

292
00:50:58.960 --> 00:51:00.520
STFC-RAL-CR03-RAL-R61-2.01: Take little energies.

293
00:51:02.210 --> 00:51:21.719
STFC-RAL-CR03-RAL-R61-2.01: So we can also do a run plan comparison file detector, so I've included this for the true, E-True, and also Eureko, because we have a non-zero negligence on Eureko at the file detector, because we've got fewer… fewer events. I think it's slightly more clear to see the differences between the run plans and file detector.

294
00:51:22.450 --> 00:51:30.470
STFC-RAL-CR03-RAL-R61-2.01: Again, you can see at lower energies, pink, pink uncertainty, that you get worse constraint, at lower energies.

295
00:51:31.090 --> 00:51:34.330
STFC-RAL-CR03-RAL-R61-2.01: And, yeah, so, sort of…

296
00:51:34.750 --> 00:51:38.890
STFC-RAL-CR03-RAL-R61-2.01: Yeah, so yeah, they're quite similar, but you can see,

297
00:51:39.290 --> 00:51:58.570
STFC-RAL-CR03-RAL-R61-2.01: interesting thing with the 75% on-axis and 25% off-axis. You can see the green line is quite similar to the blue, in the middle energies range. I think what this is showing is that if you're going to spend time off-axis, you want… you need to spend a decent amount of time off-axis in order to see the benefit against rate.

298
00:51:59.100 --> 00:52:06.369
STFC-RAL-CR03-RAL-R61-2.01: So these, yeah, these studies are kind of work in progress, but, yeah, they're important for trying to decide what

299
00:52:06.490 --> 00:52:13.620
STFC-RAL-CR03-RAL-R61-2.01: what proportion of time you should spend on access versus off-axis, which needs to be done before June starts taking data.

300
00:52:16.870 --> 00:52:27.790
STFC-RAL-CR03-RAL-R61-2.01: Okay, so the final thing, bit of work I did for my thesis was a fake data study. So as I mentioned it quite a few times, we know the cross-section model's got inaccurate statistics.

301
00:52:28.000 --> 00:52:36.129
STFC-RAL-CR03-RAL-R61-2.01: So what we want to do is see if the new approach is able to fit data that has an altered cross-section model in the Monte Carlo.

302
00:52:37.290 --> 00:52:41.620
STFC-RAL-CR03-RAL-R61-2.01: So what we want to do is see if the new approach has enough freedom that the uncertainty band

303
00:52:41.820 --> 00:52:44.369
STFC-RAL-CR03-RAL-R61-2.01: Contain this modeling.

304
00:52:44.880 --> 00:52:54.439
STFC-RAL-CR03-RAL-R61-2.01: So, the… so, yeah, I've got this little diagram of what I did. So, we have the nominal Monte Carlo, which has the simulation using the FIT, so our nominal cross-section model.

305
00:52:54.940 --> 00:52:57.440
STFC-RAL-CR03-RAL-R61-2.01: It comes from the interaction models we have currently.

306
00:52:58.370 --> 00:53:06.099
STFC-RAL-CR03-RAL-R61-2.01: So I've got some data where this will be in a world where 20% of the proton energy is missing. So this could be a scenario where

307
00:53:06.190 --> 00:53:20.110
STFC-RAL-CR03-RAL-R61-2.01: nature has produced more neutrons than we were expecting, so 20% more neutrons, and what that would result in is that the reconstructed neutrino energy we measure would actually be reduced, so it's been reduced by 20% of the proton energy.

308
00:53:21.200 --> 00:53:25.610
STFC-RAL-CR03-RAL-R61-2.01: So what we've done with this data is basically altered the reconstructed neutrino energy.

309
00:53:26.530 --> 00:53:42.389
STFC-RAL-CR03-RAL-R61-2.01: So, before I run a fit, the questions I want to ask the answer is, is the fit still able to recover the true neutrino energy correctly, despite the fact we've buried the recove? And does the uncertainty envelope that we get capture the discrepancy we see between the best fit and the prediction?

310
00:53:44.960 --> 00:53:59.940
STFC-RAL-CR03-RAL-R61-2.01: So, yeah. So, and then the other question that I want to ask is, if there is any bias, does having off-axis samples and being able to learn the truth for UK mapping better, mean we can remove some of the bias, to encapsulate discrepancies better as well?

311
00:54:02.350 --> 00:54:10.410
STFC-RAL-CR03-RAL-R61-2.01: So these are the fit results I got for the on-axis only. So the plot on the left here, you can see the reconstructed neutrino energy.

312
00:54:11.350 --> 00:54:28.739
STFC-RAL-CR03-RAL-R61-2.01: So this is what I was talking about when I said that you remove 20% of the proton energy. So you can see the nominal shape of the reconstructured neutrino energy in yellow, and the… when you remove 20% of the proton energy, you see the events shift to lower energies. So the shape changes, and you get this bump… bump here.

313
00:54:29.380 --> 00:54:41.169
STFC-RAL-CR03-RAL-R61-2.01: So you can see that the best fit result from the fit is actually in the blue line, which is on top of the, fake data. So what this is showing is that we're reconstructing the neutrino energy, perfectly.

314
00:54:41.750 --> 00:54:45.440
STFC-RAL-CR03-RAL-R61-2.01: And this is because we have high statistics, and we're able to do that.

315
00:54:46.110 --> 00:54:57.890
STFC-RAL-CR03-RAL-R61-2.01: The true neutrino energy, so what we should have… what the best fit should have recovered, if there was… if it was able to, was the nominal result, because this is what the true neutrino energy spectrum looks like, and that hasn't changed.

316
00:54:59.590 --> 00:55:07.500
STFC-RAL-CR03-RAL-R61-2.01: For the uncertainty band, you can see it, doesn't cover the yellow band, even though you can see that the same thing, and it would be best fit.

317
00:55:07.690 --> 00:55:11.460
STFC-RAL-CR03-RAL-R61-2.01: Quite carefully, to the result you were expecting.

318
00:55:11.900 --> 00:55:20.980
STFC-RAL-CR03-RAL-R61-2.01: So the bias, the difference between the Montecraft, between the prediction and the best fit is shown here. So you can see that, particularly in low neutrino energies, there's quite a big difference.

319
00:55:21.330 --> 00:55:25.399
STFC-RAL-CR03-RAL-R61-2.01: sort of middle around 2GE is not too bad, and then you get this,

320
00:55:25.600 --> 00:55:29.730
STFC-RAL-CR03-RAL-R61-2.01: PK when there's quite a big difference, and see this in math grade as well.

321
00:55:31.750 --> 00:55:46.840
STFC-RAL-CR03-RAL-R61-2.01: So, this is showing that the E true best fit is incorrect at some energies. So there is a bias in the true neutrino energy, and, yeah, fitting the Ricoh spectra does not mean we're going to get an E true spectra correctly.

322
00:55:49.420 --> 00:56:00.630
STFC-RAL-CR03-RAL-R61-2.01: So, what happens if we add some off-axis data? So, can learning truth to Rico mapping better reduce the bias we see, and then we can reconstruct the true neutrino energy better?

323
00:56:01.600 --> 00:56:16.329
STFC-RAL-CR03-RAL-R61-2.01: So, this is… this is very much work in progress, but you can see, hopefully, on the on-axis sample, that the difference between the best fit and the… what we were predicting for the true neutrino NG looks much more similar, so the bias

324
00:56:16.450 --> 00:56:22.660
STFC-RAL-CR03-RAL-R61-2.01: Okay, it's still there at the very lowest energies, but you can see that it's much closer… they're much closer to 1.

325
00:56:23.150 --> 00:56:24.719
STFC-RAL-CR03-RAL-R61-2.01: Across most energies.

326
00:56:25.070 --> 00:56:41.160
STFC-RAL-CR03-RAL-R61-2.01: But you can see there's still some differences, so particularly at… so if I make the same plot at 8 meters and the same plot at 20 meters, you can see that in various regions, there's quite a big difference between the best fit and, the nominal values.

327
00:56:42.620 --> 00:56:56.399
STFC-RAL-CR03-RAL-R61-2.01: So what this is basically showing is that this model can't currently, be trusted, to predict, the neutrino energy at all the off-axis positions, which means we currently don't have a result we can propagate for the bar detector.

328
00:56:56.810 --> 00:57:05.769
STFC-RAL-CR03-RAL-R61-2.01: So, there might… there's various reasons why this might be. So it might be that the current parameterization of eternity bias is not sufficient for this pegged study.

329
00:57:06.370 --> 00:57:11.700
STFC-RAL-CR03-RAL-R61-2.01: It could be to me to do with the way that I've been parameterized,

330
00:57:12.130 --> 00:57:22.670
STFC-RAL-CR03-RAL-R61-2.01: these variables, that's something that's under investigation currently. It may be that I need to look at an alternative parameterization in order to fit,

331
00:57:22.790 --> 00:57:24.320
STFC-RAL-CR03-RAL-R61-2.01: This payday study.

332
00:57:24.850 --> 00:57:27.729
STFC-RAL-CR03-RAL-R61-2.01: But yeah, this is… this is kind of what I can regress.

333
00:57:29.120 --> 00:57:38.279
STFC-RAL-CR03-RAL-R61-2.01: Yeah, so, summary slide. So June is going to be systematics limited, which means we need to, control the cross-section uncertainties,

334
00:57:38.710 --> 00:57:42.430
STFC-RAL-CR03-RAL-R61-2.01: Better, because all the models, disagree.

335
00:57:43.380 --> 00:57:54.930
STFC-RAL-CR03-RAL-R61-2.01: So, we need to… to do an oscillation analysis, we need to understand the truth to reca mapping, and the current cross-section models, we have to do that are too restrictive to do that, with any degree of confidence.

336
00:57:55.950 --> 00:58:04.730
STFC-RAL-CR03-RAL-R61-2.01: So, there's different new… I should demonstrate a new model agnostic approach where we might be able to do this, which involves introducing more parameters.

337
00:58:04.930 --> 00:58:10.819
STFC-RAL-CR03-RAL-R61-2.01: That are model, yeah, model-independent, and based on interaction kinematics.

338
00:58:11.660 --> 00:58:17.389
STFC-RAL-CR03-RAL-R61-2.01: And when we do this, we get a much more honest residual uncertainty, which can propagate the bi-sector.

339
00:58:17.710 --> 00:58:26.139
STFC-RAL-CR03-RAL-R61-2.01: Zoom Prism allows us to learn the truth to reca mapping more precisely, and we get a better constraint of the near detector and file detector when we do that.

340
00:58:27.220 --> 00:58:36.699
STFC-RAL-CR03-RAL-R61-2.01: This technique can also be useful for deciding how much time you want to spend on-axis compared to off-axis, before June takes data in order to optimize sensitivity.

341
00:58:36.810 --> 00:58:39.819
STFC-RAL-CR03-RAL-R61-2.01: And fake data study,

342
00:58:40.300 --> 00:58:49.249
STFC-RAL-CR03-RAL-R61-2.01: So the on-axis, fits the RICO, but we're still seeing a bias in the true neutrino energy. The PRISM is able to produce, but we're not quite there yet.

343
00:58:50.640 --> 00:58:54.389
STFC-RAL-CR03-RAL-R61-2.01: Yeah, thank you. Does anyone have any questions?

344
00:59:00.700 --> 00:59:03.419
STFC-RAL-CR03-RAL-R61-2.01: Thank you very much, that was a great help.

345
00:59:03.560 --> 00:59:08.879
STFC-RAL-CR03-RAL-R61-2.01: Amit, are there any questions here in the room or on Zoom?

346
00:59:11.320 --> 00:59:23.189
STFC-RAL-CR03-RAL-R61-2.01: Yes. Oh, I have several. You've eloquently made the case for having the off-axis data, but you're going to spend beam time collecting it.

347
00:59:23.760 --> 00:59:30.690
STFC-RAL-CR03-RAL-R61-2.01: beam time is expensive. Have you considered making a second copy of the neo-detector? I think that would be even more expensive.

348
00:59:31.030 --> 00:59:33.809
STFC-RAL-CR03-RAL-R61-2.01: I don't know. We're talking about 10 years running.

349
00:59:34.850 --> 00:59:42.589
STFC-RAL-CR03-RAL-R61-2.01: I guess it's something to consider. I mean, I think, yeah, the volumes of liquid are what we've got currently, but it's expensive and expensive technology.

350
00:59:44.760 --> 01:00:04.760
STFC-RAL-CR03-RAL-R61-2.01: So, yeah, but yeah, really interesting thing to consider making a, yeah, larger… larger nude detector. But you want 28 meters away, why not put another one 28 meters? I mean, don't make it 28 meters across, so that's big. But you could go further off axis, and going further off axis does mean you would get even more constraint,

351
01:00:05.380 --> 01:00:18.749
STFC-RAL-CR03-RAL-R61-2.01: But, yeah, the near-detector hole can only be so large, so I think 28 meters was kind of decided as the optimum sort of compromise, but yet enough constraint for axis was not being too expensive.

352
01:00:19.170 --> 01:00:38.509
STFC-RAL-CR03-RAL-R61-2.01: The thing with slower statistics is the need to become systematically limited about a month after we turn it on, so moving, it doesn't matter. Okay, so the mean time is not really an issue? No, it's for the 5 tech. In fact, the neat text probably won't be on 2 or 3 years after we start taking 5T data anyway. You were talking sort of run-plan optimization, but it sounds like you can collect as much data as you want.

353
01:00:41.030 --> 01:00:43.379
STFC-RAL-CR03-RAL-R61-2.01: Well, that's for early results.

354
01:00:44.680 --> 01:00:47.420
STFC-RAL-CR03-RAL-R61-2.01: Yeah, you said several, so please go on. Yes.

355
01:00:47.850 --> 01:00:53.150
STFC-RAL-CR03-RAL-R61-2.01: Page 13, you showed an equation that you were claiming to be solving to…

356
01:00:53.260 --> 01:00:55.740
STFC-RAL-CR03-RAL-R61-2.01: Work out what you're looking at.

357
01:00:56.050 --> 01:01:13.650
STFC-RAL-CR03-RAL-R61-2.01: And I was wondering if part of your problem might be here. You've got an efficiency for the near-detector reconstruction that's a function of the true neutrino energy, but is it not actually going to be as much a function of the RICO energy if you lose neutrons and scatter energy out of your beam, but these are correlated?

358
01:01:13.860 --> 01:01:21.110
STFC-RAL-CR03-RAL-R61-2.01: Presumably that's actually taken care of, is it? That is, that's taken care of by the energy smearing, between the interaction bottle and the detector.

359
01:01:21.240 --> 01:01:24.059
STFC-RAL-CR03-RAL-R61-2.01: Well, that's written as an efficiency,

360
01:01:25.200 --> 01:01:36.599
STFC-RAL-CR03-RAL-R61-2.01: Sorry, so is the energy smearing taking place as an efficiency, or by moving the energies of the reconstructed events? Maybe I've misunderstood what you're doing here. So it's… it is taken as an efficiency.

361
01:01:38.690 --> 01:01:42.509
STFC-RAL-CR03-RAL-R61-2.01: Don't you want to do something that corrects the energy, rather than just…

362
01:01:42.750 --> 01:01:47.929
STFC-RAL-CR03-RAL-R61-2.01: I think when you do both, you can measure it as an efficiency and also correct as well.

363
01:01:48.520 --> 01:01:55.200
STFC-RAL-CR03-RAL-R61-2.01: But yeah, the detector efficiency and the energy smearing this equation are supposed, I think, to account for both.

364
01:01:55.600 --> 01:01:59.739
STFC-RAL-CR03-RAL-R61-2.01: Okay We'd be… He says, not quite convinced how that works.

365
01:02:02.580 --> 01:02:09.459
STFC-RAL-CR03-RAL-R61-2.01: All right, so then you've got 600 bins. How did you fix on 600 bins, and to what extent did the results depend upon that choice?

366
01:02:09.730 --> 01:02:26.629
STFC-RAL-CR03-RAL-R61-2.01: So 600 bins, so I kind of needed enough to get a residual uncertainty. The problem when you're doing, Markov chain Monte Carlo is that the number of parameters you have in your fit, sort of scales with convergence, so the more parameters you have, the more difficult it is

367
01:02:26.870 --> 01:02:31.159
STFC-RAL-CR03-RAL-R61-2.01: To get it to conversion, because the longer the chain that you need in order to do that.

368
01:02:31.860 --> 01:02:47.070
STFC-RAL-CR03-RAL-R61-2.01: So, I knew that I needed to choose my billing such that it made sense with the event rate spectra and the variables that I was using, so that was one consideration. But also, yeah, choosing the number of parameters.

369
01:02:47.320 --> 01:02:48.290
STFC-RAL-CR03-RAL-R61-2.01: So…

370
01:02:48.660 --> 01:02:56.260
STFC-RAL-CR03-RAL-R61-2.01: kind of the number of parameters. I did do various studies where I varied the number of parameters slightly to see what effect that had on the uncertainty.

371
01:02:56.370 --> 01:02:59.550
STFC-RAL-CR03-RAL-R61-2.01: And, yeah, when… when I could see that it…

372
01:02:59.700 --> 01:03:04.820
STFC-RAL-CR03-RAL-R61-2.01: exact number of parameters wasn't important anymore, that's kind of where I decided to place a cut.

373
01:03:05.280 --> 01:03:11.400
STFC-RAL-CR03-RAL-R61-2.01: So it gets to a point where it doesn't matter. If you double the number of bins, you would not increase the uncertainty by Route 2.

374
01:03:11.570 --> 01:03:19.520
STFC-RAL-CR03-RAL-R61-2.01: So, it's not, it's not an exact, yeah, it's not an exact, if you double the number of bins, you double the uncertainty.

375
01:03:20.050 --> 01:03:33.390
STFC-RAL-CR03-RAL-R61-2.01: Because this is something, at least in Atlas, has been an issue that you bid more and more finding, because you want to represent detail, and people are often going to do some sort of DNN or something to learn the shape, rather than to a binned approach. Is that something you've considered?

376
01:03:36.190 --> 01:03:47.340
STFC-RAL-CR03-RAL-R61-2.01: Yeah, a little bit. I think that's, yeah, kind of beyond, sort of what I was doing, but I think, yeah, the number of bins that I have in the analysis is definitely, yeah, definitely, I think 600 is probably not enough, and yeah, using a different approach.

377
01:03:47.470 --> 01:03:51.470
STFC-RAL-CR03-RAL-R61-2.01: Eventually, is, is another avenue you could go down.

378
01:03:54.880 --> 01:03:56.110
STFC-RAL-CR03-RAL-R61-2.01: That's most of my question.

379
01:03:57.250 --> 01:04:13.369
STFC-RAL-CR03-RAL-R61-2.01: One suggestion. You introduced the, hydronic energy about page 20 or somewhere, 14 or 15, but I think you meant the hydronic charged energy, or hydronic energy excluding neutrons, or something. That confused me for many pages. Thank you.

380
01:04:16.230 --> 01:04:17.170
STFC-RAL-CR03-RAL-R61-2.01: What's that?

381
01:04:20.900 --> 01:04:22.879
STFC-RAL-CR03-RAL-R61-2.01: Any other comments?

382
01:04:26.780 --> 01:04:37.520
STFC-RAL-CR03-RAL-R61-2.01: Yes. So you, you alluded, especially with the, the run plan stuff, about what if you only went to 12 or something majors, you said?

383
01:04:37.660 --> 01:04:43.500
STFC-RAL-CR03-RAL-R61-2.01: Given… globally funding and potentially funding situations in the US.

384
01:04:43.780 --> 01:04:57.559
STFC-RAL-CR03-RAL-R61-2.01: what's the… you have alluded to, you know, what if you don't go so far? What's the sort of timeline and the music as to funding for… can you go to… will you still get to go to the full 20 meters?

385
01:04:57.680 --> 01:04:59.659
STFC-RAL-CR03-RAL-R61-2.01: Or is it going to be cuts to 12?

386
01:04:59.910 --> 01:05:11.529
STFC-RAL-CR03-RAL-R61-2.01: Yeah, so this rumplow was not designed because it's going to be cut to 12 meters. I think, yeah, hopefully these studies show that you need to use the full scope of 28 meters,

387
01:05:11.560 --> 01:05:25.630
STFC-RAL-CR03-RAL-R61-2.01: And I don't think there's any plan to change that. Yeah, that run plan was kind of designed, because obviously within the first few years of data taking, or the first, you know, few months of data taking, you might not be able to move the detector to all off-axis positions.

388
01:05:26.140 --> 01:05:36.179
STFC-RAL-CR03-RAL-R61-2.01: you know, it takes time and energy to move the detector, so, you know, what would you… what constraint can you already get, even if you haven't taken the near detector as far off access?

389
01:05:36.310 --> 01:05:37.950
STFC-RAL-CR03-RAL-R61-2.01: As you're capable of doing.

390
01:05:39.770 --> 01:05:42.049
STFC-RAL-CR03-RAL-R61-2.01: Wow, that's for how long is Connect.

391
01:05:43.580 --> 01:05:51.979
STFC-RAL-CR03-RAL-R61-2.01: I'd like to apologize about my comment on page 20. If I hadn't been sitting at the back so other people's heads would be writing, I'd have seen them on the bottom.

392
01:05:52.130 --> 01:05:56.070
STFC-RAL-CR03-RAL-R61-2.01: Okay,

393
01:06:00.840 --> 01:06:02.750
STFC-RAL-CR03-RAL-R61-2.01: On Zoom?

394
01:06:04.740 --> 01:06:08.010
STFC-RAL-CR03-RAL-R61-2.01: Yeah, please speak on Zoom if you have a question.

395
01:06:09.430 --> 01:06:19.560
STFC-RAL-CR03-RAL-R61-2.01: Let's see… Okay There are no more comments, and let's thank Harvest again. Thank you.

396
01:06:27.440 --> 01:06:32.420
STFC-RAL-CR03-RAL-R61-2.01: Thanks also coming to close the seminar.

