FERMILAB-PUB-22-726-T The impact of neutrino-nucleus interaction modeling on new physics searches

2025-05-06 0 0 1.52MB 24 页 10玖币

侵权投诉

FERMILAB-PUB-22-726-T

The impact of neutrino-nucleus interaction

modeling on new physics searches

Nina M. Coyle1Shirley Weishi Li2,3Pedro A. N. Machado2

1Department of Physics, University of Chicago, Chicago, IL 60637, USA

2Theoretical Physics Department, Fermilab, P.O. Box 500, Batavia, IL 60510, USA

3Department of Physics and Astronomy, University of California, Irvine, CA 92697, USA

E-mail: ninac@uchicago.edu,shirleyl@fnal.gov,pmachado@fnal.gov

Abstract: Accurate neutrino-nucleus interaction modeling is an essential requirement for the suc-

cess of the accelerator-based neutrino program. As no satisfactory description of cross sections exists,

experiments tune neutrino-nucleus interactions to data to mitigate mis-modeling. In this work, we

study how the interplay between near detector tuning and cross section mis-modeling aﬀects new

physics searches. We perform a realistic simulation of neutrino events and closely follow NOvA’s

tuning, the ﬁrst published of such procedures in a neutrino experiment. We analyze two illustrative

new physics scenarios, sterile neutrinos and light neutrinophilic scalars, presenting the relevant ex-

perimental signatures and the sensitivity regions with and without tuning. While the tuning does not

wash out sterile neutrino oscillation patterns, cross section mis-modeling can bias the experimental

sensitivity. In the case of light neutrinophilic scalars, variations in cross section models completely

dominate the sensitivity regardless of any tuning. Our ﬁndings reveal the critical need to improve

our theoretical understanding of neutrino-nucleus interactions, and to estimate the impact of tuning

on new physics searches. We urge neutrino experiments to follow NOvA’s example and publish the

details of their tuning procedure, and to develop strategies to more robustly account for cross section

uncertainties, which will expand the scope of their physics program.

arXiv:2210.03753v1 [hep-ph] 7 Oct 2022

Contents

1 Introduction 2

2 Near detector tuning 3

3 Sterile neutrinos 5

3.1 Model description 6

3.2 Analysis 6

4 Light neutrinophilic scalars 9

4.1 Model description 10

4.2 Analysis 11

5 Discussions and conclusions 13

A Tuning details 15

A.1 Sterile neutrinos 15

A.2 Light neutrinophilic scalars 16

– 1 –

1 Introduction

The neutrino sector is the least understood of the Standard Model (SM). Several questions about

neutrinos, of both experimental and theoretical character, remain open to this date, including the

conservation of lepton number, the mass mechanism, the mass ordering, leptonic charge-parity vio-

lation, and puzzling experimental results which remain to be understood, among others. One of the

reasons some of these questions remain unanswered lies in neutrinos’ small, weak-only interactions.

Studying these fermions requires special experimental setups, with intense sources and typically huge

detectors. As a consequence, the neutrino sector is widely regarded as a promising portal to beyond

the standard model (BSM) physics, while neutrino experiments are excellent tools to search for novel

BSM signatures.

Of particular interest are accelerator neutrino experiments, in which an intense neutrino beam is

produced by impinging protons on a target, and detectors are placed near and/or far from the neutrino

beam origin. While such a setup provides a rich environment to study standard and beyond standard

physics, it also comes with challenges. These experiments measure neutrinos through their interactions

on nuclei, which are quite complex and not fully understood processes. In the energy range between

a few hundred MeV to a few GeV, the kinematics span perturbative and non-perturbative regimes,

with the hadronic degrees of freedom being a mixture of nucleons, resonances, and partons [1–3]. In

addition, nuclear eﬀects can play an important role in the ﬁnal states of neutrino-nuclei interactions [4–

15]. Although there have been many recent studies on understanding neutrino-nucleus scattering cross

sections [16–26], we still do not have a sound theoretical framework to compute the cross sections or

estimate their uncertainties [27–29]. On top of that, the neutrino ﬂux is driven by meson production

and decay, which suﬀer from considerable QCD hadronization uncertainties [30–32].

A data-driven approach is typically sought to mitigate the large uncertainties associated with

the neutrino ﬂux and interaction models. Commonly, neutrino event spectra at near detectors (ND)

are used to calibrate the neutrino ﬂux and interaction cross section models. We will refer to this

procedure, which will be explained in detail later, as ND tuning. This information is then used in

the far detector (FD) to extract oscillation parameters and other neutrino properties. Although the

primary role of NDs is to provide these inputs for physics searches at the FDs, NDs themselves are

very special. Due to the proximity from the neutrino source and large size, they will typically have

a very high neutrino event rate, accruing millions of events over the lifetime of an experiment, as

opposed to FD statistics of Op1000qevents or less. Such large statistical data sets have an enormous

potential to probe new physics in the neutrino sector, particularly for models which may aﬀect the ND

more than the FD [33–63]. Because we do not have the perfect model of neutrino-nucleus interactions,

and the cross section and ﬂux modeling is tuned with the ND data, an obvious question is raised:

how much does the interplay between ND tuning and cross section mis-modeling interfere with new

physics searches?

This paper aims to study the interplay between cross section mis-modeling and searches for BSM

physics. To estimate the impact on BSM searches, we analyze a couple of illustrative scenarios,

presenting their signatures before and after tuning. Because the tuning procedure is not uniquely

deﬁned and its impact depends on the technical details, we follow the procedure described in a recent,

detailed publication of the NOvA collaboration [64] as closely as possible. We choose to follow NOvA’s

approach because they are the only collaboration that published the details of their tuning with codes

at the time of this work; we strongly encourage other experiments to follow NOvA’s example and

publish their tunings. For this work, we generate neutrino events with state-of-the-art tools and apply

NOvA’s tuning on an event-by-event basis. We identify and discuss how the smoking gun signatures

– 2 –

of new physics models are aﬀected. While some signatures, such as sterile neutrino oscillations, are

robust against tuning, mis-modeling of neutrino interactions can bias the experimental sensitivity. On

the other hand, the sensitivity to other types of BSM scenarios can be overwhelmed by cross section

mis-modeling.

This is the ﬁrst quantitative study exploring the impact of tuning on new physics searches. We

identify conceptual lessons that should be applicable to other tuning procedures, detector set ups,

analysis details, and BSM scenarios. For instance, we expect our ﬁndings to hold to some extent

even in BSM searches that do not employ a ND tuning but instead rely on a simultaneous ﬁt to near

and far detector data (see, e.g., Ref. [43]). Our goal is not to faithfully reproduce any experimental

analyses. We hope our results will encourage theorists and experimentalists to consider the impact

of ND tuning on new physics scenarios and to carefully estimate systematic uncertainties related to

neutrino ﬂuxes and cross sections.

2 Near detector tuning

Near detectors in long-baseline neutrino oscillation experiment are designed to mitigate the uncertain-

ties from both the neutrino beams and neutrino-nucleus cross sections. Conceptually, this procedure

may seem straightforward:

PpναÑνβ;Eνq “ C

dNFD

dEνNσβpEνq

dNND

dEνNσαpEνq

,(2.1)

where PpναÑνβ;Eνqis the oscillation probability the experiment is trying to measure, Cis a constant

accounting for detector sizes and distances to the source, the superscripts ND and FD denote near

and far detectors, and dNα,β {dEνare the event rates of να,β in terms of true neutrino energy. The

diﬀerences between σαand σβ, if α‰β, could in principle be computed theoretically [24]. The

oscillation probability can be directly derived from the unoscillated event rate dNα{dEνmeasured at

the ND and the oscillated event rate dNβ{dEνmeasured at the FD.

In reality, this “logical division” is not practical because one cannot directly measure neutrino

energy, the near and far detectors have diﬀerent systematics, and even the unoscillated neutrino

ﬂuxes are diﬀerent at the near and far detectors solely due to the diﬀerent solid angles. We can

appreciate this more concretely by studying the following equation,

NFD

NND

pErecoq “ CżdEν

dφFD

dEν

PpναÑνβ;EνqσβpEνqMFD

βpEν, Erecoq

żdEν

dφND

dEν

σαpEνqMND

αpEν, Erecoq

,(2.2)

where NFD{ND encode the reconstructed neutrino event spectra, dφFD{ND

α{dEνare the ﬂuxes at the

far and near detectors without oscillation, σα,βpEνqare the total cross sections, and MFD{ND

αare the

migration matrices. The challenge here is that experiments measure the left-hand side of Eq. (2.2)

and they need to infer the oscillation probability Pon the right-hand side. The main diﬃculty

is encoded in the term σαpEνqMαpEν, Erecoq, where the reconstruction of the true neutrino energy

depends on the details of neutrino-nucleus interaction, as well as detector responses to diﬀerent ﬁnal-

state particles. An obvious example is neutrons, for which both the modeling of neutrons produced

– 3 –

by neutrino interactions and the corresponding detector responses are relevant. Currently, predictions

on the number of outgoing nucleons in a neutrino-nucleus scattering event, as well as their energy and

isospin, diﬀer drastically among generators [18,29]. In addition, neutron detector responses suﬀer

from signiﬁcant uncertainties due to neutron propagation and event reconstruction [65–67].

Because of these complications, oscillation experiments adopt a near-detector tuning procedure.

Assuming SM physics, one can predict the measured neutrino energy spectrum in the ND with an

accelerator neutrino beam simulation, an event generator such as GENIE [68], NuWro [69], GiBUU [70],

or ACHILLES [71] that simulates neutrino-nucleus interaction cross sections and ﬁnal states, and a

detector simulation predicting the migration matrix. When the predicted and measured neutrino

spectra disagree, one can modify the cross section simulations until they match. Because of the

nature of such calculations and the complexities of these simulation packages, there is no unique,

agreed-upon way to tune the models. One can vary the model parameters, adopt alternative models,

or take a model-agnostic approach and add more degrees of freedom.

In addition, it is well appreciated that neutrino beams have sizable uncertainties and they interfere

with cross section uncertainties. Diﬀerent experiments also treat the ﬂux tuning and the cross section

tuning diﬀerently. NOvA, for instance, only uses hadronic production data, in-situ measurements of

horn position and current, beam parameters, etc., and MINERvA neutrino-electron scattering data

to tune their ﬂux prediction [72]. T2K, on the other hand, uses ND neutrino-nucleus scattering data

in addition to auxiliary data to tune their ﬂux and cross section models at the same time [73].

In this work, we follow the tuning procedure outlined by NOvA [74]. The NOvA experiment is a

long-baseline experiment comprised of two scintillator detectors (CH2) placed along a νµ{¯νµbeamline

produced by the NuMI facility at Fermilab: a near and a far detector placed 1 km and 810 km from

the beam source, respectively. The 14 kton FD observes the muon neutrino beam after long-baseline

oscillations, intended to measure oscillation parameters including ∆m2

32 and θ23. The smaller 0.3 kton

ND is nearly identical to the FD to minimize systematic uncertainties.

The ND tune procedure is detailed in Ref. [64]. Neutrino interactions with the material in the ND

are ﬁrst generated using GENIE v2.12.2 with the default models and parameters. Then, the following

changes are applied to the GENIE simulation based on auxiliary theoretical and experimental studies,

i.e., NOvA ND data is not used for this step:

•adjusting the value of axial mass, mA, from 0.99 to 1.04 GeV, based on recent re-analysis [16]

of neutrino-deuterium scattering data;

•modifying the momentum distributions of the initial nucleons for quasi-elastic scattering, based

on a MINERvA study [75];

•lowering the magnitude of neutrino-scattering in non-resonance pion production regime by 57%,

motivated by re-analysis of old bubble chamber data [76];

•suppressing delta resonance production in low-Q2region, motivated by measurements by Mini-

BooNE [77], MINOS [78], MINERvA [79,80], and T2K [81].

All these changes are to improve the baseline model in GENIE.

After these changes are applied to GENIE, there are still large discrepancies between the measured

neutrino spectrum in NOvA ND and the simulated spectrum. The last step of NOvA’s ND tune

is the crucial step of which we are studying the eﬀect. To understand how it works, let us ﬁrst

deﬁne the kinematics of neutrino interactions, then explain how kinematic variables are measured in

NOvA. Theoretically, an incoming neutrino with energy Eνproduces an outgoing lepton with energy

– 4 –

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

FERMILAB-PUB-22-726-TTheimpactofneutrino-nucleusinteractionmodelingonnewphysicssearchesNinaM.Coyle1ShirleyWeishiLi2;3PedroA.N.Machado21DepartmentofPhysics,UniversityofChicago,Chicago,IL60637,USA2TheoreticalPhysicsDepartment,Fermilab,P.O.Box500,Batavia,IL60510,USA3DepartmentofPhysicsandAstronomy,Univ...

展开>> 收起<<

FERMILAB-PUB-22-726-T The impact of neutrino-nucleus interaction modeling on new physics searches.pdf

共24页,预览5页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

FERMILAB-PUB-22-726-T The impact of neutrino-nucleus interaction modeling on new physics searches

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: