Peter Onyisi - “Top Quarks: The New Flavor”

all right welcome to my club I'd like to introduce assistant soon-to-be associate professor computer amuse me peter is a experimental particle physicist I'll talk to you about the app and some research at the Atlas experiment but he got his bachelor's at the University of Chicago or he was a winner of the ApS is after award and then he went to Cornell for his PhD or if he worked on the Cleo experiment studying branching fractions of charmed strange my song so play this business and after that he went to the University of Chicago as an Enrico Fermi fellow to do his postdoctoral work where among many things he was a significant contributor to the discovery of the Higgs boson and the WWE Channel um and since then he joined the faculty at the University of Texas Austin six years ago back in 2012 so quite personally and least have a number of things what I think he'll talk to us about paramount today he led the group that did measurements of the Higgs coupling to top quarks which is currently one of most interesting measured Higgs measurements that we can make right now and he's additionally the Atlas experiment data preparation coordinator and I just want to explain what that is because it's it's kind of an opaque term if you're not a member of the experiment but the David comes off the Atkins detector in a stream of zeros and ones from a very heterogeneous set of sub detectors and it's Peters job to make sure that that gets to the analyzers in a form that they can recognize and search for explosives it also requires that that we know how much data was taken and what conditions Rosia were taking everything so for the entire four thousand person collaboration Peters who tried to make sure that all works so without further ado what hand that's right all right thanks very much and thanks very much for having me so right up here I have the LHC page one which is a web page that anyone can access and it gives you information on the current status of the machines so in fact very recently I guess basically a week ago we had we had first beam in the machine for this year and in fact in principle on Thursday we might be getting first collisions although not good for physics one so we have to make sure that to filter them out for users so right now the LHC machine people are doing varieties of optics things and recommissioning the accelerator for this year oh it's it's a bit hard to do this okay so here comes the actual talk all right so I'm going to talk about top quarks as the new flavor and of course you know we've known about top for quite a while but you know since I have a bit of what we call a flavor physics background in my field I find that it's actually rather hard to escape it and so I've wound up studying you know quarks again except a much heavier one this time so this is not going to be you know full survey of top quark physics it's really actually going to focus on things that I personally have worked on and hopefully give you this little a little bit of a flavor of how we actually do things in this field for people who may not be familiar with with particle physics in general so I'll start off with a little bit of an overview of the standard model of particle physics so you know looking at the serve reductionist paradigm you know we we keep looking at things they're smaller and smaller trying to explain the properties of larger things in terms of a smaller subset of components so we start off with the other date everyday matter we realize oh well you know we have an awful lot of different kinds of materials in the world but they're all formed via chemical combinations of different elements we ask well why do we have the atoms that we have what explains the regularities of the periodic table we find that the period a periodic table arises from quantum mechanics of electronic configurations you have to put some number of electrons orbit you know in a potential okay what explains the kinds of potentials that you have that comes from nuclei and we find the nuclei we have this quantized behavior coming from having protons and neutrons in them integer numbers of those and then we find inside those that protons and neutrons themselves are composed of different numbers of more fundamental particles quarks and actually as far as we know as far as matter particles go this is where it ends so we haven't seen anything inside an electron to date and we've never seen anything inside a quark to date if it not for lack of trying people have certainly tried so we can actually summarize all the directly known fundamental particles again particles that we've not been able to divide in any way in this little cute pie chart and we can it's not enough to know what the content of the world is you also want to know how they interact with each other so you want to the dynamics of it so you have a Lagrangian here and you can actually there's an error in this so this is something you can get from CERN and there's actually a flaw in it so if you spot the flaw you can pat yourself on the back so the question that we're really trying to ask and answer is is this a complete set of material and forces and we actually know it can't really be but this is what we know of and is this Lagrangian which again very very simplified expression of it is this what there is to know about the dynamics of these particles and again we have no reason to believe that's not the case so all ordinary matter is actually just composed of these so now looking back at that first slide I showed we have electrons and then the two quarks inside the protons and neutrons the up and down quarks and then it turns out that for reasons unknown there are heavier versions of these things so things were the same kind of dynamics the same charges under different forces but heavier for some reason so you can also sort of tell the point at which people stopped trying to come up with Greek names for things and just switch to English so we have heavier partners of electrons called me ons and Tau's and then we have up charm and top and then four down strange and bottom so these have a complicated pattern which I'll come back to a bit later of exactly of their relationships with each other and no one really knows where this pattern comes from on top of that we've got these heart to see neutrinos so 100 trillion neutrinos from the Sun pass through you every second you don't notice so that tells you a little bit about how a rare their interactions are and studying the properties of these which in some sense really are some of the least well-known particles that we have established the exist that we don't actually know what their masses are for example as opposed to everything else so this is a major part of the u.s. particle physics program for the future then one of the major realizations is that not only is matter can of particles but in fact the forces between particles are also part of our also particles so we have the particle of light or electromagnet ISM in general the photon and you know given as we are visual animals you know a lot of our culture is encoded in electromagnetic interactions we have the glue on this is the thing that keeps these up and down quarks actually bound into these objects protons and neutrons and not just those but gives us other kinds of configurations of these quarks the W and Z which are a bit more obscure the most direct relevance they have to us they mediate radioactive beta decay the W does and allows the Sun to shine so the way the Sun turns four protons into a helium nucleus involves the W and then in the middle of this of course since I like to do Higgs physics it would put the Higgs in the center obviously this is so this particle gives mass to the others in this picture so looking at the dynamics here we can actually sort of summarize what's going on these different terms so we have a term corresponding to the force particles moving and again this is highly summarized this is cramming you know many many lines of the thing together we have the matter particle motion and it's interaction with forces and they're very tightly coupled together in terms of you don't have an awful lot of flexibility in how this term must be once you've chosen certain underlying mathematical symmetries we have the matter particle interactions with the higgs field and this is much more free in fact we have the Higgs field motion and interaction with the forces and then we have the Higgs fields self interaction so you see the Higgs actually shows up a whole bunch of places in here so this picture is not at all complete there are a lot of things that we know are not there so the big thing is I have said they look at these forces well I haven't mentioned gravity and since you're all sitting down I think you believe gravity exists so there's no verified theory of quantum gravity there's no dark matter so if you our dark energy so if you look at the universe you do cosmology mentions you ask how much stuff is there how much energy is there in various kind of various components and you use general relativity you go and measure it counts things actually you find that you don't have enough you certainly don't have enough matter of our kind to explain the amount of matter that we conclude must be in the universe and we say we also know that neutrinos don't have the properties that you would need to have for a valuable dark matter things so it's not just that we have a lot of neutrinos and I'm not just baryons so basically we know there must be some kind of dark thing out there we also know I mean even worse than that is that the majority of the energy budget of the universe at this time comes from something that isn't matter at all it has the property that is the universe expands there's more of it so the universe is being inflated by this invisible source of energy you don't know what that is in this model as is there's not enough matter it doesn't predict you know if you start with the universe that doesn't have that has no net amount of matter vs. antimatter you can't generate enough of any symmetry to explain the how much matter there is now as opposed to that T matter there's no neutrino masses in this model although in principle most theorists would say it's an easy bolt on so but we we know that there are degrees of freedom beyond the standard model we know neutrinos have masses and also a somewhat more theoretical concern naturalness in that the parameters of the model regarding the Higgs field seem to be very highly fine-tuned in a way that doesn't seem sensible so I'll say a little bit about try and give you this idea of how the Higgs field works in terms of giving masses to things and the top quark being the most heavy particle that we know of this is very strong interaction between the two of them so this spontaneous symmetry breaking if you ask well we know what's the point of this whole thing this is basically when we talk about electroweak physics is what we're talking about what the fundamental point is if you look at the weak interaction so the thing with the w and z that's a force so their charges involved the same way that you know what determines electric force well you've got electric charges the weak force particles that experience the weak force have a weak charge the thing is if you actually ask what would happen if I put a mass term in the Lagrangian just bare you find that this would actually violate conservation of this charge and that's a bad thing we do not like of a condom you have a violation of charge conservation so what we do is we have a fundamental Lagrangian that is invariant that doesn't violate charge but we create a vacuum that does so this is a serf same principle that you have a ferromagnetic material it has the ground state of it has all these domains where you have low enough temperatures domains where spins are aligned and you have a net field and it picks out a preferred direction even though the fundamental laws don't have any preferred direction in them but the ground state has a preferred direction same here essentially if you look at the potential that we posit for the this Higgs field and of course this is a this is a sub position this is not necessarily the full form of the potential that could be more than one Higgs field etc etc but this is a simplest model then what you find is that the symmetric situation where there is no Higgs field is actually a local maximum of the potential and that if you want a minimum of the energy you actually want to be at some value that's not zero so this drives the so-called vacuum expectation value in which you actually have the Higgs field is a lower energy if you have a non zero field given the way this is coupled to the fermions this vacuum expectation value winds up generating a mass term in the Lagrangian but it also it does this at the same time as it generates an interaction of the fermions and some oscillations around this minimum so the oscillations around this minimum or what we call the Higgs boson and this says well the mass of the particle which is determined is this constant Y which we call you Kayla coupling that mass here relates is not only telling the masses that that Y tells you not only the mass of the particle but also the strength of the interaction between the particle and the Higgs field so this is how the top court gets its mass and this is why the top quark is expected to interact with the Higgs boson um so there's this wonderful paper which is actually just past its 50th anniversary so it's called a model of leptons by Steve Weinberg and if you read it to serve read it and say ok we're done you know it's a the this is the answer to everything so he doesn't have works in it which is a which ok at the time there was not there were good reasons for not having them there but as far as the leptons are concerned that's correct so when I say flavor physics I really mean against studying the matter sector and it's in its interactions and particular I'm mostly going to talk about this particular thing here which is the interactions of the matter with matter with the Higgs field so one more bit of fun theory here and this is actually is driven a lot of work including you know the people here who worked on the bar so you know all of this so this issue here that happens and it happens for both leptons and quarks of both electrons muons Taos neutrinos and the quarks is a mismatch between the mass and weak I can states so I figured if you can think of back at this this coffee cup you basically have matrix matrices here you can diagonalize but there's no particular reason relating the different generations of quarks but there's no particular reason that this diagonalization has to have the same form for both these two terms and in fact they don't so the Higgs interactions and the weak interactions why end up picking out different basis states so you have this sort of multi dimensional states of lepton and quark flavors and you pick out specific preferred directions in them but those preferred directions need not be the same for the masses for the things there are mass eigenstates and the things that interact with the weak the leak force so for the standard model quartz we use mass basis so we actually have things where we can go and more or less measure the mass of them and so it's natural to say well iodide the top quark is this thing of charge plus two-thirds and mass 170 2.5 GeV so this means that the mass terms in this case conserved flavor because we define that the flavor is the are the eigen basis of the mass the photon Z and Higgs interactions conserve flavor that means they don't change one kind of particle into another so if I have an Allah if I have a top quark it emits a photon it's still a top quark will always be so this means that there is in our jargon there's no flavor changing neutral currents so there's nothing that doesn't involve a W that changes the kind of what a particle is and the W interactions can actually change particles to another and this mismatch of the basis produces a matrix which is unitary and tells us basically about the overlap of these bases leptons in fact use the weak interaction basis so neutrino oscillations are discussed in terms of that they're in there are conceptually the same thing excited be the processes of neutrino oscillations and of the c-care matrix are actually conceptually the same thing just expressed in slightly different ways so a brief taste of the sort of thing that it has been done in quark flavor physics which is actually still I mean for non top quarks which is actually still a really interesting field in many ways a very very brief taste here so one thing of course to do is to take this matrix and actually prove it and try and measure all the components and test the unitarity and that's done with this so-called unitarity triangle amongst other things because if you say you find some measurement that should agree with the consensus of all the other experiments and find it disagrees that might be evidence that something is contaminating your measurement some kind of new physics so you would have something inconsistent with a global fit so what this actually represents is basically three numbers that should add up to that true that'll end up adding up to zero which is essentially the same as saying that these two columns of this thing should be orthonormal to each other so they should have if you take the inner product they should be zero so that's what that comes from another thing you can try and do is test the universality of lepton couplings so in the standard model except for the Higgs the leptons are treated the same way by the weak interactions so if I look at a process where I said this this particle is a be quark here it turns into a charm quark again it can only do that by emitting a W in the standard model then I should be able to predict that W doesn't really care whether it decays to the top s'alright to a tau of muon or an electron for a that those things are all the same but if I had say some kind of extra particle that doesn't that distinguishes tau from muon from electron which for example could be some extended Higgs boson then this would violate the expectation from the standard model and in fact there is a hint so far that this may be happening in that there is a prediction of the standard model in these these ratios here of various decays which should be here and the combined result from all the experiments is over here and this is actually quite a significant discrepancy although no single experiment is yet shown conclusively that there actually is a discrepancy so but this is something to look out for so this is actually tree right so right so the the the fact is actually people have to explain how this is consistent with everything else given that it should be a fairly important process if it's true you know yeah so you don't even have to go to loops for this sort of thing okay so to come back to the top so I mentioned that there is this relationship between the top and the Higgs and this is naturalness problem so the top quark is very heavy so it is the heaviest in fundamental particle that means the top Higgs interaction is very strong and that means that you know you have to actually consider quantum Corrections you can't just take an ordering of first-order in perturbation theory you have to consider higher orders in perturbation theory when computing things and you get corrections to the Higgs parameters and so something like this where the Higgs boson turns into a top pair and then they turn back into expose on this can happen on short timescales no problem and then you find well if I assume that there's you know I I have new physics that cuts things off at some high momentum scale of this top pintura this this top loop because I in principle can have arbitrarily high momentum tops in here what I find is that this is correction to one of the higgs mass parameters that actually squares quadrat actually scales quadratically with that cut-off choice so if I chose that to be say the quantum gravity scale this would be 10 to the 32 times the observed value and you have to somehow cancel 10 to the 32 a gift n to the 32 to get one so you really need an extreme cancellation of this bear parameter now I mean fundamentally there's an there's no fundamental problem with that but it seems extremely arbitrary you would like to have some motivation for that so this motivates looking at new physics models which cancel the correction in some natural way or lower the cutoff or both so supersymmetry famously introduces particles where you would have a plus sign for a top you would have a minus sign for these particles and so naturally these things would cancel extra dimensions for example can come in and reduce the quantum gravity scale so the cutoff goes down goes down composite Higgs would say well again at some point the Higgs becomes more a fuzzy thing and these correct these computations no longer apply so there there's a lot of work that's been done in this direction motivated by this problem in some ways this is really driven most of the work that's been done at the LHC is trying to find solutions to this but the bottom line here is that Higgs properties are enormous ly affected by the top quark interactions and then here's what I like to call the Doomsday plot which asks whether so if you took the standard model to be correct you say okay I forget about all this naturalness stuff I just take what it is but then I do the just the quantum Corrections under those parameters to the Higgs potential what you find oK we've got we're here at some relatively small but nonzero value of the Higgs field it's our vacuum and then you look at what happens the potential it grows and then it actually turns down again so with our current best knowledge basically depends on the top mass or the top Higgs interaction because that's what gives you the top mass and the Higgs mass and you find that actually this turns down and for our current known values gives a second minimum that is lower in energy than the one where yet now if you know what this kind of thing implies that implies there's some probability to tunnel between these two states if this minimum were significantly lower than where it is we would have done that already so we wouldn't be here to talk about it in fact we kind of like that situation where that to be true because then we could say well something must be stopping it because we're here but unfortunately that we are in this method well fortunately or unfortunately the current parameters that we've measured put us in this metastable region and you can see that a small change of few GeV a Giga electron volts on the top mass would actually move you from metastability to stability so measuring the top mass which again is basically measuring the top Higgs attraction is clearly a priority in order to understand whether this might mean something okay so questions for the Higgs sector that we're trying to resolve basically the standard model has the simplest possible higgs sector it's one fundamental doublet of complex-valued field so for degrees of freedom is this we don't know if this is actually you know it's the simplest model is correct maybe there might be extra doublet or triplet or something we don't know that these particles are in fact fundamental because all other known spinless particles are in fact composite and we don't actually know that the symmetry breaking potential is what we guess it is so what you'd expect to see things like modifications of the interaction strength if the Higgs is actually a composite particle a composite field if you had more degrees of freedom you'd have more particles in the sector but also you could expect modifications of the interaction strength and these flavor-changing neutral currents to appear and if the new interactions we don't expect and if the potential wasn't a quartic actually self interactions may be modified so i'm going to talk about things in blue modification of the top Higgs interactions of measuring that directly and favored changing neutral currents searching for those so to move to the experiment a fun side okay so here is a representation of where the LHC runs underground it's actually sort of order 100 meters underground so you don't see very much of it on the surface so here is the runway of Geneva Airport for comparison here is the Atlas experiment where I work it is conveniently close to a tram line here are is where the essential equivalent of atlas our colleagues at the CMS experiment are they are not close to a tram line so this thing this ring 27 kilometers circumference more or less actually has other experiments on it as well beyond appleís my experiment at CMS the source sister experiment to Atlas so at least is focused on heavy iron collisions and LHC be actually it is very interesting business belief lower mass work physics with a different kind of detector designs so they're very complementary to us so we collide protons that have kinetic energy over 6500 times the rest mass so basically all the energy of these things is in their motion if his first operational in 2009 we saw the Higgs in 20 twelve and in 2015 we got to near the plant energy of the machine so here is a person biking between points so this is actually if there's only so many ways to get into the thing and if you're trying to get to an intermediate point you take a bike so this is very carefully posed at the French Swiss border underground this is what the experiment itself what would look like if you could actually see it all at once because the geometry of the underground cavern means you can't really so this is Sierras here people for scale spending where they shouldn't be about 25 meters tall 44 meters long it's a very big thing although most of that it's actually air so we have a fairly compact inner part of the detector and then a large volume with magnetic fields in order to allow enough lever arm for measuring the particles trajectory bending in the magnetic fields so this design a lot of different kinds of detectors where you arrange them in the sort of onion shell pattern where you do try and do non-invasive measurements first and then more invasive measurements later to try and collect as much information as you can about the trajectories and the properties of all the particles involved so since computing is the thing I worry about so here's a here's some facts on that the early data output of the experiments is order 30 petabytes and we show no signs of reducing that so to actually turns out that the major challenge is not actually reconstructing the data that you can actually do at CERN the major challenge turns out to be simulating what we would see so it prom is of course the universe is a very good quantum computer on its own and we're trying to do a lot of simulations of quantum mechanics ourselves but it's difficult on a digital computer and so and then you want to actually see how the detector would experience having a certain kind of interaction occur inside it so that takes the vast majority of the CPU resources these are run on the distributed worldwide LHC computing grid which as the last time I checked is 170 sites over the world and about 260,000 real pretty much all the time okay so the general principle of what we do we take these protons we collide them stuff comes out well something happens first but that's something you know if we make a Higgs that thing has a lifetime about 10 to the minus 22 seconds if we make a top quark that's at five times ten to the minus twenty five we don't see those things directly what we see is none of these things they decay into something else and those things decay and by the time you get to stuff that has lifetimes before microseconds or later then we can actually have it move far enough for us to actually see things so we get some particles out and in the end some things are actually detected and then it's a bit of interesting jigsaw puzzle to figure out what is consistent with having produced what we saw so we're looking for the long-lived particles that come from from these collisions so and talk about you know what's what why is this interesting you know what what is qualitatively new about this in terms of making top a new flavor well only two accelerators on earth have ever had enough energy to make top quarks or Higgs bosons and if you look at service this bar chart here essentially all top quarks that have ever been made on earth have been made at the LHC so we have roughly 17 million of them have produced roughly 17 million of them and the amount that was a number produced at the Tevatron much much much smaller so you know once you have that kind of numbers you can actually begin to do lots of really detailed investigations that you couldn't do when every top quark was we're supremely precious so this is what top quark pair production looks like so I mentioned you know you all you see is that the remnants that live long enough so what's happened is we made a tea in a tea bar so top in a knotty top the top is decayed to a W boson and a B which is what it does essentially all the time the anti top is gone to W and B bar and in this case the W is decayed to me on a new neutrino and this would abuse got to an electron the neutrino so what we see in the final state is a muon here an electron here and these are recognized by the different patterns of energy deposition in different detectors where they are able how far they're able to penetrate and then we have these two bunches of particles that contain quark so these hadrons which we identify as having arisen from b quark so we never seek works on their own we only see them bunched up in other part with other partnering with other quarks but we're able to tell that these actually came from bees and how are we able to do that well it turns out that the B quark has lives just long enough for us to be able to see that it flies a macroscopic distance so if we look at the kind of the particles coming out from the collision so this is the primary vertex this is where most of the particles are coming from we see that we have a bunch of particles here that when reconstructed where they came from actually come from a point that's displaced so this fact that something is flown from this collision from this a primary interaction point to here before decaying till is gives us something very consistent with the the idea that that was actually the time it took for a B quark to decay so we collect data at fast clip at the design luminosity which is expressed in these weird units of area per second we actually produce h top pairs per second and we exceed that so this year we expect to maximal run it twice that so we routinely run higher than the designs so this is data accumulated per year in against some units don't worry about it but higher is better and the best year we've had for LHC operations was in fact last year and we're hoping to do as well or better this year and something we've been doing at UT is actually validating that the luma that the luminosity measurement is stable with time using physics processes that we can actually control that's fun so while collecting data were actually sensitive to about 93% of the delivered data we can't actually record the 40 mega Hertz of collisions that occur we have to reduce that to about 1 or 1 1 1 and a half kilohertz so we have triggers of various kinds and if you want to know more about the masks Lauren that reduce that perform this data reduction in real time so has to be done immediately account yet Pharma da fall into computers and once you require that all sub detectors are all working together well that's typically about mid-90s present efficiency so this is last year 93 to 95 percent depending on exactly what's up detectors you needed so this is sort of telling you we've got about three quarters of the data recorded that are used for most of the most of our physics studies ok so you might ask what limits your ability to do this kind of science well some things are just very rare we don't get to control what happens in collisions because it's it's quantum mechanics we have some probability so the things just don't happen very often or they can be very hard to distinguish from other things that are more common so you have a game of statistics you need to understand the detector many calibrations need to be done it's it's you because you don't actually have the full detector available to test with test beams and you know the things age so you need to actually understand that the performance of the detector while the detector is running and the simulations are never perfect so there's some mismatch between what we expect might happen and what actually does happen and we have to account for that and then there's also the fact that the theoretical predictions from perturbation theory there's funk and uncertainties we don't go to infinite order in perturbation theory and our many effects are non perturbative that we encapsulate in models which we know aren't perfect so those are also tuned but also have uncertainties associated so these are it's kind of a list of things that prevent us from seeing everything immediately so we've done measurements of benchmark processes so covering something like 10 orders of magnitude and production cross section production rate from proton proton to anything to down here proton proton to a pair of Z those ons and these are measurements at different energies the points and the bands are theoretical predictions and things look pretty good there so at least we have some belief that we can use we can have theoretical predictions for processes at the LHC so something a couple of things that we worked on aut so proton structure and you actually have to understand what's going on in these collisions so one thing is say well you know you're colliding parts you're collecting things inside the proton so part ons quarks and gluons so do you actually understand that behavior because that's one of the non perturbative things I mentioned so you can couple high order perturbative calculations to the non perturbative models of the proton structure itself and see whether they agree with data so here we have data as recorded here looking at production of W bosons W pluses and looking at essentially the angle at which they're produced and the data are the black points with the green uncertainty bars and different models of the proton structure actually give us different bands here so there's a purplish and there's blue and all these other colors and what that tells you is actually these things are systematically different from each other and we are able to constrain them so one thing we found in fact was that expectations for the amount of strange quark in the proton were incompatible with data and so once we included that we found that we really had to raise our understanding of the fraction of the proton that is actually carrying strange quarks and we're consistent with this being quite similar to the C quark contribution of up and down another thing we can ask is well how often when I collide these protons to actually collide how do I have to have two collisions so you and normally we think of only one pair of things colliding and then producing stuff but you could perfectly well have two different sets of things colliding the fact that you have a proton collision in the first place means that there's other stuff in close proximity already so you raise the probability of something else happening and the reason this is interesting is because if you're looking for something rare where you produce x and y but you could actually just instead have one collision that produced x and another collision that produced y then you could have a actually quite common process mimic a very rare one and so the way we looked for this there's actually a bunch of analyses but our particular one we looked at the angular just the angular difference between x and y if they are produced in a single collision you expect momentum to be conserved and they'll be back-to-back if you have instead this double part on scattering they should be uncorrelated with each other and so we find in fact back-to-back so this is pi difference between the angles and then also some contribution that is quite consistent with with this flat sort of uncorrelated bit so Atlas has done a bunch of measurements and what we see is actually consistent with models that have been proposed but we really do need to understand these numbers a bit more ok so moving on to the Higgs so searching for the top picks coupling which is a very hot topic so what we're looking for here is top quarks emitting Higgs bosons so ok so this statement I'll come back to in a bit I I think a CMS says that yes they have seen it at 5 Sigma but ok so what does that mean that means a p-value less than 3 times 10 to the minus 7 so this is sort of our standard for observing things in our field so we're trying to get a handle on this thing so the search strategy here is determined by what the Higgs turns into so the Higgs has a bunch of things that can decay into so looking at the look at this vertical dotted line this is a probability of various decays so that would be the probability to be P bar so here's one that's like 50 something percent then there's ww like that and then there's various other modes so actually mostly I'm still do W W as Lauren mentioned for a discovery I did WWI still dww find it's it's a second most common mode said it's quite quite useful so we look and look it actually a whole bunch of modes not one single mode is going to give you an observation or not you have to combine a whole bunch of different you know you look at a whole bunch of different places and all of them say well it's somewhat more consistent with Higgs being there being there and then once you put them all together they say okay well you couldn't be wrong in all these different places at once so the hick should be there so in particular my group worked on the Tri leptons signature so one of the problems with t th this is a TT bar plus six production is it's rare so at the Zion luminosity you produce one in every two hundred seconds and then you're looking for a specific pattern of decay that is 0.7 percent of that so if you were running at a hundred percent duty cycle you get only three of these produced per day and then we actually are only able to find about 13% of them for various experimental reasons so you see you don't get very many of them you have to run full out for quite a while in order to collect enough events to just say you've seen something well you see expect certain patterns in the detector so in this particular case we've made we've chosen to have three electrons or muons or some combination we don't expect any correlation of the whether their electrons or muons we expect for jets of hadrons that's basically we have four quarks and two of them are B quarks so expect to have that displaced vertex that I showed and then there's various things arising from the angular momentum you go and play with the collect Gordon coefficients and you'll find that you expect that the leptons should be clear to compare the lepton should be close to each other and then you since you don't see the neutrinos you have some momentum imbalance in the event so there are some backgrounds I will go into them in great detail but these are just other things that can happen so instead of a Higgs we have a Z boson or a photon so that actually gives you the same kind of final state except okay here you expect to see a peak at the Z mass for the mass of the leptons and in fact we do so that's that's we see this process and we can normalize that process from data we also have situations where a lepton comes out of a b DK or a photon converts into an electron-positron pair so this we have to determine directly from data through some complex things you can ask me about so since since I would like to highlight the work of my student instead of sharing the normal analysis I show the alternate analysis that we did so this first paper on run to data is public infected we just got the notification it was published today what we did in our case was instead of running so the nominal analysis used is very powerful multivariate analysis techniques it runs boosted decision trees on boosted decision trees and then phones everything in five dimensional space and all that good stuff and it has good sensitivity we actually get almost as good sensitivity by choosing three or four variables and then chopping things up in those so if you are close to the screen maybe you can actually see what the selections are but the point is as you go from left to right the red part here which is the expected contribution from the from GTR Higgs production gross we get a more higher signal to background on the right and you can see the data points are consistent with with the the adding the red and they're less consistent if you take them out so if you actually combine all these things together none of the channels themselves again these are a whole bunch of different decays that are looked at none of them by itself is able to conclude anything super significant but once you combine them all together you get a measurement compared to the ratio to the standard model of 1.6 plus 0.5 minus 0.4 okay so that's still not statistically significant by standard but it's getting there if you then take that and combine that with other measurements that were made what you find combined is 4.2 Sigma so again on this thing it's it's you know not observation but but very good and we expect three point eight so that's going it's is consistent with you know we're not somehow giving a much much stronger evidence than we would expect to for the standard model so if you combine everything it's 1.2 times the standard model plus or minus 0.3 so perfectly complete compatible with standard model if you look at the systematic uncertainties we at this point we're really worried about how well how good our measurement is and not just you know whether we can see it or not you find that quite a few of these are actually theory uncertainties so we have significant experimental uncertainties as well but we do need to basically understand some of the theoretical modeling of these backgrounds and the cross-sections for that I so here's the asterisk yesterday CMS put out a claim of a 5 plus sigma observation and you can ask me about that later okay so the going back so we haven't actually repeated this exercise but we did it for run one for earlier data you can combine all the different Higgs analyses together and ask parameterize the Higgs interactions in different ways and do this comprehensive analysis see how much you scale the standard model couplings by how much you're allowed to scale them by and if you look at the top Higgs interaction which is this row here we are come back combined Atlas and CMS results compatible we were compatible it to Sigma we're going to be much more compatible with the standard model now once we because both of us have measured stuff compatible with it now but this is actually you're able to pick apart couplings to Z bosons W bosons top tau B glue on so this is actually I mean you can even say that the undetected decay fraction is less than 34% so we're not missing more than the third if Higgs decays but the point is that really there's still a lot of room for modifications this is a very log large scale and so you can easily sneak 10% in there and that's quite you know reasonable discrepancy and we would not have seen it yet 20% you see okay so a brief thing about these flavor-changing neutral currents so as these aren't effect in the standard model they're they're at an extremely small rate 10 to the minus 15 I think so you're just not gonna see them but if you have more complex six sectors than just the one of the standard model you could in fact modify you could create this coupling so this would cause the top quark to decay into a charm quark and Higgs boson so we can go look for that so we've done math we actually reinterpreted the results that I showed earlier so with try leptin and to lepton events and what we found was consistent with no signal but also consistent with some signal so this is something to to look forward to look at again and I will just say there's actually updates for run too and we have a paper that should be coming out very very soon on this is stay tuned ok so future so just a few things on the LHC program so we're currently near the design energy and we don't expect any massive increases of the energy in the near future so we'd like to go to the design which is 14 TeV we already surpassed the design luminosity so that's good we although the design luminosity was only a for a stepping stone we expect it to do better so through the mid 2030s what were the plan is basically that you have the velocity increases this is a bit out of date so this actually somehow I didn't get the latest version of this slide so this actually is now 30 times because we took enough data last year so we're going to get we need major accelerator and detector upgrades for this and lots of people are involved with those and after that perhaps we get an energy increase with stronger magnets you know so this is a going way off the end of this chart in terms of plan so it's not just accelerator you get collect more data just run things and do things the way you've always done them there are lots of improvements of other kinds to come in particular analysis techniques we've actually had a very fruitful collaborations between theory and experiment for this generally I think final LHC results are probably better that we currently predict because we'll will just be better at doing analyses there are a whole bunch of ideas that people are working on to improve sensitivity to do better to reduce systematics I my group has some interest in the non non shaded things so some new opportunities as mentioned the top mass and the top you call a coupling are intimately related so we can also do top mass measurements is you get more data you can try and use different techniques so this is one that we've been studying here for people in the know it's using J size as a proxy for B Jets but by collecting lots of data you can determine the difference between this red gaussian ish thing and this blue gaussian distinct and that corresponds to 5gv difference of the top mass but the point is we can actually get you know that's is really systematic assists our statistics limited and by the end of this year we're actually going to be competitive statistically with some other normal measurements of the top mass so this is something again we it's enabled by taking more data for far-future LHC is really the only place we're going to be able to study topics interactions for a long time there are many statistically limited topics even with HOH see you know in the 2030s when I get new colliders that can improve give us better measurements so for example if we have this thing called the international linear collider we can imagine reducing the top mass and certainty down to 40 M visa or over an order of magnitude improvement which would be great so here from the Japanese effort on this is a celebrity and Lynn Evans who is head of the ILC project so here a bunch of people that I've had the privilege to work with doing this research and here is a place in Texas so so thanks for your attention yes I mentioned to my discrepancies that cinema one was the CSEA this range towards being more so I wondered what yeah okay so that's a good question so you're talking about this I assume so what what so this the the issue here is as far as I know and someone who knows more about lattice couldn't correct me in the problems you're dealing with details of proton you know of protons addressed and this is this is very difficult to do so I don't think there's actually I've been to show predictions of what this ought to be so it's basically been assumed in done in fits to earlier datasets which had significantly more uncertainties that you would have less strange that it's heavier therefore there's a strange depression and some numbers got put in as spaced lines but they've never really been tested again with with the data we have now the problem is that these these things they have physics in them but they always start off from some model that you put in of what the what the densities are and if that initial model is wrong all the extra physics you do to it will not fix it I I don't believe so no this is really you have to go look at it the flavor violation yeah savor yeah universality of I was okay so the easiest way so one thing is just to put in this this Higgs this charge Higgs so if you have more than one you have a more complicated Higgs sector than just one doublet like we have in the standard model then naturally you'll get an object out like this and it will interact differently with towers with you with electrons unlike the W so you could introduce that and it would change this interaction the problem is it turns out that this is incompatible with other things so the favorite explanation I believe at this point is left to quarks so essentially turn it on its side and attend this interaction Linda's idea well so theorists are very good at coming up with ways to hide those large numbers so you don't actually wind up with I mean once once you've once you've lived with this then you you do the cancellation from the beginning and then you don't have these gigantically large numbers appearing in your actual calculations so you're safe there but yes it's a visa so that this is more of a conceptual thing than actual impediment to calculating the rest of it precision is yeah we don't usually have precision problems per se but we do have lots of fun things where people you know overflow counters without expecting it because they didn't really think about it sure okay so the latest cm so the result that we have I showed here is the is using only the 13 tv data of the 2015-2016 data and so we're adding 2017 obviously so CMS did so we also had earlier data at 7:00 and 8:00 TV which we did the analysis on so in that case we wound up with actually a fairly significant excess over the standard mami knots not significant statistically but you know it was large so what we knew was if we combined this result with the earlier results we would actually get a very large number here with not much improvement on the expected because the earlier data really was worth much less than the data we're taking in terms of what you would expect to get out of it but because we had enough word fluctuation there I believe we believe it was an upward fluctuation we would be sort of claiming 5 Sigma even though we didn't expect the experiment to do that well so the CMS so CMS has run to results are essentially the same as ours they also had an upward fluctuation in run one and so their claim of observation is precisely that they combine the two data sets but in that case they they're taking advantage of the fact that they already had what in this assumption must then have been an upward fluctuation in the earlier data so it's a little bit at that point it becomes a philosopher philosophical argument of what is 5 Sigma which so I think we can say we've seen it but okay you know it's a who got to 5 Sigma transpose yeah the expected is 4.2 so so they had a significant fight the problem is you know say that there was some correlated effect that was visible and run one data for both Atlas and CMS and got swamped by you know stuff in in run two then we might have been mistaking something else for t th and run one and run two we've seen t th so when you're doing that combination you might actually be calling extra significance for t th production something in fact was some other process and that would be unfortunate but again it's a bit of a philosophical argument to be Fred this is the least powerful way of doing it but so one of the famous issues is same sign W production so if you produce W plus W plus or W - W - so then you you've got the you don't have charge total charge zero which okay so you have to arrange for some interesting initial states to get that to happen so that's quite rare in the standard model and could be used as a probe of how the W bosons interact with the experiment with each other the problem is that you could produce two W bosons and different scatterings with the same sign and so so it's believed that your similar kind of kinematic effects here that they just don't quite look the same allow you to select for W plus W plus produced like this instead of like this so they're actually not so worried about it now but they realized oh no actually this might be something we have to worry about and in just in general if you're really doing rare studies you should actually at least stop and ask yourself whether this could happen certainly when we started this particular analysis with this w ad the j sigh we didn't expect this to show you know we did the calculation we said oh actually this isn't like a third of our signal we didn't expect that I actually expected zero so the fact that we saw anything here at all was actually very interesting but but then okay on top of that we we saw this contribution thanks again thank you I don't know

Channel: Stanford Physics

Views: 2,509

Rating: 5 out of 5

Keywords: stanford, physics, colloquium, peter onyisi

Id: Di0Uyj6gPeo

Length: 58min 15sec (3495 seconds)

Published: Wed Apr 11 2018