aa r X i v : . [ h e p - ph ] D ec Prospects for neutrino oscillation parameters
Patrick Huber ∗ Center for Neutrino Physics, Virginia Tech, Blacksburg, USAE-mail: [email protected]
In this contribution we discuss the future of the global long-baseline neutrino oscillation pro-gram. The case is made that our current lack of understanding of neutrino-nucleus interactionsis a serious challenge which will need to be met with new experimental initiatives in neutrinoscattering.
Neutrino Oscillation Workshop4 - 11 September, 2016Otranto (Lecce, Italy) ∗ Speaker. c (cid:13) Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-ShareAlikeLicence. http://pos.sissa.it/ rospects
Patrick Huber
1. Introduction
The discovery of neutrino oscillation, awarded the 2015 Nobel prize in physics, is one of thegreat discoveries of our time. Apart from finding that neutrinos have mass, we also have beenblessed with large mixing angles in both solar and atmospheric neutrinos as well as a large valueof q . All of which are necessary ingredients to allow for the search of CP violation in neutrinooscillations, since the absolute size of CP effects is suppressed by the smallest of the mixing anglesand by the ratio of two mass squared differences. The study of CP violation in turn requiresthe ability to perform an appearance measurement, and given our lack of proper technology toefficiently create or detect n t , this implies the use of n e and n m . The need to involve n m in turnrequires to use energies in the 100’s of MeV and above and this in turn results in long baselines of100’s km and more.The event rate for n e → n m or n m → n e appearance in leading order is proportional to sin q and thus, the fact that q is large is good news: it allows to obtain a sufficiently large eventsample using conventional pion-decay, horn-focused neutrino beams, which we have been usingfor more than 3 decades. It still requires, however, to push this technology to its limits by usingMW-level proton beams and very large detectors of at least the size of Super-Kamiokande. Theselong-baseline experiment are characterized in terms of megawatts, kilotons and decades. This is anew scale of effort in neutrino physics and makes neutrino physics, in its sheer experimental scopeand scale, much more similar to traditional accelerator based science, like for instance the LHCprogram.Indeed, DUNE is expected to start data taking roughly a decade from now and to run for adecade and to absorb a very large fraction of the resources of its host country and the internationalneutrino community. Also for many scientists this will be the only major experiment during theircareer. The sheer scale of these new neutrino experiments also implies that failure to make a majordiscovery is not an option. For small-scale neutrino experiments like the ones in the Booster neu-trino beam at Fermilab or reactor neutrino experiments not finding anything or being preempted bysome other measurement is a completely acceptable outcome because there are many experimentsat this scale. As long as some of these experiments make discoveries or provide measurements, sci-ence and importantly, along with it, the careers of those conducting these experiments can progress.The bargain each scientists strikes with a generational experiment like the LHC is based on the con-viction that the science goal is worthy of a lifetime of struggle and the realization that this is the only way to achieve this particular science goal. Implied in this bargain is the understanding thatboth, the worthiness of the science goal and the unique ability to achieve it by this one specificexperiment, will endure over the decades it takes to carry out this program.For the LHC the science goal was the discovery of the Higgs boson and to find New Physics,either in the form of new degrees of freedom or by some other breakdown of the Standard Model.It also is understood that this would require particle collisions at unprecedented energies and afterthe global community had settled on the LHC is also was clear that this will be the only machinein this energy regime for the foreseeable future. The Higgs has been discovered and the search forNew Physics is ongoing, with a major upgrade of the machine underway – a roundabout success.For long-baseline oscillation experiments the science goal is the discovery of leptonic CP vi-olation and the search for New Physics by precisely testing the three-flavor oscillation framework.2 rospects Patrick Huber
However, this is where the similarity with the LHC program ends: there is no international consen-sus to pursue only one experiment and already on-going experiments, notably NOvA in the U.S.and T2K in Japan, will be making inroads into the very question of leptonic CP violation. Whileit seems very unlikely that the on-going experiments will achieve a 5 s discovery, a 3 s evidencefor leptonic CP violation may be conceivably obtained. Thus future long-baseline experiments willnot be entering a terra incognita but will be tasked with the exploration and mapping of terrainseen before. Therefore, precision measurements and a comprehensive set of physics goals is thereal scientific objective.
2. Future prospects
First tentative hints for leptonic CP violation became apparent earlier this year [1] and sincemore global fits have been performed [2] indicating a preference for a value of the leptonic CPphase around d ≃ − p /
2. This hint for CP violation currently is at 1–2 s level and thus may benothing more than a statistical fluctuation. On the other hand, both T2K and NOvA do consistentlyreport n e appearance rates at the upper end of the possible range, which is what is expected if d = − p / s rejection of leptonic CP conservation, assuming the current best-fit value isclose to the true value of d [3]. In combination with continued NOvA running, current NOvAdata represent only about 1/6 of the approved number of protons on target, and the precise deter-mination of q by Daya Bay [4] there will a be a good determination of the leptonic CP phaseby 2025. Also, the question whether q is maximal can be effectively addressed by this data inparticular when combined with atmospheric neutrino data, which Super-Kamiokande continues toaccumulate. Another crucial piece of information is the neutrino mass hierarchy and while currentdata seems to have no particular preference, future data from NOvA has the potential to answer thisquestion without any ambiguity. Then, of course, there is JUNO and possibly PINGU all trying toaddress the same question. It appears, therefore, likely that the question of the mass hierarchy willbe settled before too long.A reasonably good proxy for the actual physics reach of a long-baseline experiment is given bythe total number of events it accumulates in the n m → n e appearance channel. The time evolutionis shown in Fig. 1, assuming current best fit values: the next decade will see an order of magnitudeincrease in these numbers and the statistical errors will drop below 5%. DUNE has to run for about3 years to double the global event sample in the appearance channel and even at the end of itsplanned run it only will have roughly tripled the available global data. Also, by the time DUNEdata starts to dominate the global data set, this global data set will comprise about 1500 events, witha corresponding statistical accuracy of 2.6%. This excellent statistical accuracy has to be matchedby a corresponding and ideally somewhat smaller systematic uncertainty – otherwise the massiveinvestment in these experiments is wasted. 3 rospects Patrick Huber T o t a l s i gna l e v en t s s t a t. e rr o r Exps. Running 50% in neutrino mode
CD-R at our bfGLoBES 2016
T2KT2K IINOvAT2K(II)+NOvADUNE sin q =0.304 sin (2 q )=0.085 sin q =0.452 d CP =- p /2 D m =7.5x10 -5 eV D m =2.457x10 -3 eV Figure 1:
Shown is the time evolution of the total n m → n e appearance rate for the experiments named in thelegend. The assumption is that d = − p / / sqrtN , with N being the number of events shown on the left-hand vertical axisand thus, represents the statistical accuracy of the data set. The DUNE run plan is based on [5, 6, 7]. TheNOvA run plan is based on [8]. The T2K run plan is based on [3].
3. Road to precision
There are two challenges for a precision long-baseline experiment: determining the appearancerate and determining the neutrino energy. The rate uncertainty is driven by the knowledge ofthe beam flux and neutrino interaction cross section. Using a pion-decay, horn-focused beam,the flux knowledge is at best at the 10% level despite many years of efforts to improve the stateof the art by hadron production measurements. In realization of this limitation all long-baselineexperiments employ a near detector or a suite of near detectors. Following the argument of Ref. [9],the problem lies in the fact that the near detector, at best, measures the product of beam flux andcross section and thus to know either one, the other quantity is required; there are fewer observablesthan unknowns. In reactor experiments like Daya Bay the problem is solved by measuring theproduct of the identical cross section, in this case for inverse beta decay, and the beam flux, in thiscase ¯ n e , in the near and far detectors, which have the same physical dimensions and characteristicswithin a fraction of a percent. In inverse beta decay a unique and clean flavor tag is obtained andthere is no doubt about the underlying micro-physics, since neutron decay is very well understood.The source is a point source and thus the geometrical acceptance difference between near andfar detectors is given by the square of the ratio of baselines, which in turn can be measured withcentimeter precision. Thus, the oscillation probability can be extracted by simply taking the ratioof far to near detector data corrected for geometric acceptance.In a long-baseline experiments none of the conditions which allowed Daya Bay to succeedis met: cross sections in near and far detectors are different since the neutrino flavor and energy4 rospects Patrick Huber distribution is different. The beam flux and flavor composition seen by the near detector is differenteven in absence of oscillation because of the complicated acceptance. The near and far detector arenot the same size, do not use the same technology and sometimes not even the same target material.The near detector sees the decay pipe as line source whereas the far detector sees it as point source.Relating the observable quantities to the underlying micro-physics in an event requires a preciseunderstanding of the micro-physics, which we lack. Instead, we rely on Monte Carlo simulations,which have been tuned to existing data, for event identification. The limitations of this approach areexemplified by the fact that basically no two cross section measurements performed in a neutrinobeam ever seem to to agree with each other.Finally, the need to reconstruct the neutrino energy precisely is a design feature of experimentsusing a wide-band neutrino beam and all the physics benefits of covering a range of L / E -valuesrely on this reconstruction. Again, the lack of an understanding of the micro-physics of neutrino-nucleus interactions prevents accurate neutrino energy reconstruction because the observable sig-natures do not have an understood relation to neutrino energy. It has been shown that approximateschemes like the exploitation of quasi-elastic kinematics do not provide sufficient accuracy for thenext generation of experiments [10, 11] and also calorimetric methods have their limitations [12],in particular with respect to neutral secondary particles like neutrons. In particular, for the deter-mination of CP violation this is a major issue [13].Therefore, even an ideal near detector seems to be insufficient to resolve the systematics prob-lem in long-baseline experiments since it ultimately provides fewer observables than unknowns.The hope that the multitude of different event types, charged current quasi elastics, charged currentsingle pion etc. will provide a sufficient number of observables is naive: they only can constraineach other if we connect them with a micro-physical model, which we do not have. It is a fallacy toconfuse existing Monte Carlo event generators with an actual understanding of the micro-physicsas borne out by the great difficulty to reconcile any new measurement of exclusive neutrino crosssections with existing ones, to quote from the most recent MINERvA publication [14]: “ Unlike themeasurements of the individual processes (quasi- elastic, pion production) the total cross sectionmeasurements agree with the GENIE simulation and prior data to within their uncertainties [. . . ]”(emphasis added). Given that the MINERvA experiments in many ways represents the state ofthe art in neutrino scattering, this is discouraging, despite the success for inclusive cross sections:neutrino energy reconstruction desperately relies on an understanding of the exclusive interactionchannels.Usually when the situation seems hopeless on the experimental side, we turn to theory toprovide the needed answers. It is obvious that describing bound state multi-nucleon systems inthe ground state is a daunting task and so far can be only achieved by using phenomenologicalHamiltonians and for nuclei lighter than A=12. The problem at hand is however not to knowthe ground state (or the low-lying excitations) but the response to energy transfers well into theGeV-range, which requires a relativistic treatment. Once the hard scattering event has taken placewe also need to understand how the reaction products get out of the nucleus, a problem typicallysummarized under the term final-state interactions. Recently, Benhar [15] points out that manydifferent calculational approaches based on very different assumption seem to yield the same result,which is puzzling. The role of electron scattering data can not be overstated, since we can exploitfully exclusive kinematics to separate the various micro-physical contributions and any model of5 rospects
Patrick Huber neutrino interactions must reproduce electron scattering data. Fortunately, a program is underwayto obtain this crucial data for argon [16].
4. Summary
We are very fortunate in neutrino physics having found neutrino oscillation, and by associationthat neutrinos have a mass; also, we find large mixing angles, including a quite sizable value of q . Thus, the stage is set to study genuine three-flavor effects and to, hopefully, find leptonic CPviolation. There is a vibrant ongoing experimental effort, spearheaded by NOvA and T2K andwe have won approval for DUNE. The future of neutrino physics lies in precision studies of long-baseline neutrino oscillation and while a large q allows to accumulate significant event sampleswe also will need concomitant control of systematical uncertainties.The choice of pion-decay, horn-focused neutrino beams implies poor knowledge of the pri-mary neutrino flux, which combined with our lack of understanding of neutrino-nucleus interac-tions presents a challenge. In this note we reiterate the argument previously made in the literaturethat even a capable near detector complex can not meet this challenge. The state of theory is suchthat it is at best unclear whether theory can provide the missing answers. Let us assume, that therewill a be a breakthrough in theory providing a full model of neutrino-nucleus interaction includ-ing final state interactions. This model will not be based on the Standard Model Lagrangian orany other first-principles calculations; given the complexity of the problem it has to be based onphenomenological insights and appropriate approximations. Thus, even if we all agree that wehave the “right” model, we will need to test this model at the level of accuracy we intend to useit at. Therefore, the neutrino community needs to seriously think about an experimental neutrinoscattering program to accompany the long-baseline oscillation program, see for instance [17, 18]. Acknowledgments
This work was supported by the U.S. Department of Energy Office of Science under awardnumber DE-SC0009973.
References [1] A. Marrone, “Status and prospects of global analyses of neutrino mass-mixing parameters.” Talk atNeutrino 2016, Jul, 2016.[2] I. Esteban, M. C. Gonzalez-Garcia, M. Maltoni, I. Martinez-Soler and T. Schwetz,
Updated fit to threeneutrino mixing: exploring the accelerator-reactor complementarity , .[3] K. Abe et al., Proposal for an Extended Run of T2K to × POT , .[4] D AYA B AY collaboration, F. P. An et al., Measurement of electron antineutrino oscillation based on1230 days of operation of the Daya Bay experiment , .[5] DUNE collaboration, R. Acciarri et al., Long-Baseline Neutrino Facility (LBNF) and DeepUnderground Neutrino Experiment (DUNE) , .[6] DUNE collaboration, T. Alion et al., Experiment Simulation Configurations Used in DUNE CDR , . rospects Patrick Huber[7] C. J. Mossey, “News from LBNF.” Talk at DUNE collaboration meeting, May, 2016.[8] P. Vahle, “New results from NOvA.” Talk at Neutrino 2016, Jul, 2016.[9] P. Huber, M. Mezzetto and T. Schwetz,
On the impact of systematical uncertainties for the CPviolation measurement in superbeam experiments , JHEP (2008) 021, [ ].[10] P. Coloma, P. Huber, C.-M. Jen and C. Mariani, Neutrino-nucleus interaction models and their impacton oscillation analyses , Phys. Rev.
D89 (2014) 073015, [ ].[11] A. M. Ankowski, O. Benhar, P. Coloma, P. Huber, C.-M. Jen, C. Mariani et al.,
Comparison of thecalorimetric and kinematic methods of neutrino energy reconstruction in disappearance experiments , Phys. Rev.
D92 (2015) 073014, [ ].[12] A. M. Ankowski, P. Coloma, P. Huber, C. Mariani and E. Vagnoni,
Missing energy and themeasurement of the CP-violating phase in neutrino oscillations , Phys. Rev.
D92 (2015) 091301,[ ].[13] P. Coloma and P. Huber,
Impact of nuclear effects on the extraction of neutrino oscillationparameters , Phys. Rev. Lett. (2013) 221802, [ ].[14] MINER V A collaboration, J. Devan et al.,
Measurements of the Inclusive Neutrino and AntineutrinoCharged Current Cross Sections in MINERvA Using the Low- n Flux Method , Submitted to: Phys.Rev. D (2016) , [ ].[15] O. Benhar,
Interpretation of the Neutrino-Nucleus Cross Section , .[16] O. Benhar et al., Measurement of the Spectral Function of Ar through the ( e , e ′ p ) reaction , .[17] NU STORM collaboration, D. Adey et al., nuSTORM - Neutrinos from STORed Muons: Proposal tothe Fermilab PAC , .[18] NU PRISM collaboration, S. Bhadra et al.,
Letter of Intent to Construct a nuPRISM Detector in theJ-PARC Neutrino Beamline , ..