This document provides an overview of the key steps involved in designing and conducting neutrino oscillation experiments. It discusses how experimental neutrino physicists formulate hypotheses about neutrino oscillations, design experiments to test these hypotheses, build and operate detectors to collect data, and make statements based on the observed data. Specific examples from T2K and NOvA are used to illustrate how these experiments addressed challenges like creating neutrino beams, choosing detector locations, and identifying electron neutrinos emerging from muon neutrino beams. The document aims to provide theoretical physics students a practical guide for thinking about neutrino experiments.
An automated and user-friendly optical tweezers for biomolecular investigat...Dr. Pranav Rathi
An automated and user-friendly optical tweezers for biomolecular investigations; a versatile, automated, fast, precise and user friendly optical tweezers capable of doing verity of biomolecular experiments.
The Large Hadron Collider at CERN is the world's most largest machine, having stored magnet energy in excess of 5 Giga-Joules, and energy stored in each circulating beam in excess of 360 Mega-Joules.
A machine protection system has been implemented to protect LHC against the risks related to stored energies, whilst allowing the machine to operate close to the physical limits.
An automated and user-friendly optical tweezers for biomolecular investigat...Dr. Pranav Rathi
An automated and user-friendly optical tweezers for biomolecular investigations; a versatile, automated, fast, precise and user friendly optical tweezers capable of doing verity of biomolecular experiments.
The Large Hadron Collider at CERN is the world's most largest machine, having stored magnet energy in excess of 5 Giga-Joules, and energy stored in each circulating beam in excess of 360 Mega-Joules.
A machine protection system has been implemented to protect LHC against the risks related to stored energies, whilst allowing the machine to operate close to the physical limits.
Dentro del marco actual del software offline de ATLAS (Athena ), dos tipos diferentes de algoritmos de clusterización son utilizados. La reconstrucción de la deposición electromagnética que proporciona una medida óptima de electrones y fotones se realiza mediante un algoritmo que busca el centro de dicha deposición. Este algoritmo es un cono alrededor del centro de gravedad de una celda y forma en una ventana de 3x5 celdas (para fotones no convertidos) o 3x7 celdas (para electrones y fotones convertidos). En el caso de la reconstrucción de las cascadas hadrónicas, las celdas cercanas al centro con deposiciones de energía por encima de un cierto umbral son añadidas al cluster. Este algoritmo es conocido como algoritmo topológico.
Atomic level manipulation of matter using Scanning Transmission Electron Micr...Ondrej Dyck
Discussion of developments surrounding the transformation of the scanning transmission electron microscope from an imaging platform into a manipulation platform.
Dentro del marco actual del software offline de ATLAS (Athena ), dos tipos diferentes de algoritmos de clusterización son utilizados. La reconstrucción de la deposición electromagnética que proporciona una medida óptima de electrones y fotones se realiza mediante un algoritmo que busca el centro de dicha deposición. Este algoritmo es un cono alrededor del centro de gravedad de una celda y forma en una ventana de 3x5 celdas (para fotones no convertidos) o 3x7 celdas (para electrones y fotones convertidos). En el caso de la reconstrucción de las cascadas hadrónicas, las celdas cercanas al centro con deposiciones de energía por encima de un cierto umbral son añadidas al cluster. Este algoritmo es conocido como algoritmo topológico.
Atomic level manipulation of matter using Scanning Transmission Electron Micr...Ondrej Dyck
Discussion of developments surrounding the transformation of the scanning transmission electron microscope from an imaging platform into a manipulation platform.
A whole new world in Biology is exposed. Cells can now be "dissected" into nanometre thin "slices" while at the same time the composition and 3-D ultrastructure of each "slice", using Nano Scanning Auger Microscopy, are determined
by
Prof. J.L.F. Kock (Ph.D.)
Department of Microbial, Biochemical and Food Biotechnology
University of the Free State
P. O. Box 339
Bloemfontein 9300
South Africa
This presentation has been delivered to "Online Medical Conference" at http://conferences.medicalia.org
Trabajo Final de Grado Física(UV): Angular distribution and energy spectrum o...Christiaan Roca Catala
Thesis of my bachellor in Physics.
We analise the angular distribution and the energy spectrum of neutrinos coming from decaying pions in a boosted frame. From this we observe the benefits of placing a detector at an off-axis angle respect to the trajectory of the pion.
In concrete we derive the effects of adding first order corrections to the mass of the initially set massless neutrino in the kinematical scheme. We compare the results with the well-known biography and determine that those corrections lead no contribution.
Finally we discuss the importance of this scheme on the neutrino experiments nowadays. A higher detection rate leads better results on the actual detections. In a near future this could shed some light on some of the most elusive problems nowadays in neutrino physics. For example, the neutrino mass hierarchy or the CP violation in the leptonic sector.
We pay special attention to the recent results of T2K (Tokai to Kamioka) and NOvA.
I gave 1 hour seminar at ANSTO (Australian Nuclear Science and Technology Organization) to introduce my approach to magnetism. I see myself as an experimental physicist who is studying magnetism by using neutron scattering techniques. Throughout my career, I had learned local structure analysis (PDF), magnetic structural analysis, and inelastic neutron scattering technique to investigate superconductor, multiferroics, antiferromagnets, helimagnets, and frustrated magnets. I was trying to explain my approach to magnetism as an experiment physicist to both professional scientists and novices.
Measurement-induced long-distance entanglement of superconducting qubits usin...Ondrej Cernotik
Although superconducting systems provide a promising platform for quantum computing, their networking poses a challenge as they cannot be interfaced to light—the medium used to send quantum signals through channels at room temperature. We show that mechanical oscillators can mediate such coupling and light can be used to measure the joint state of two distant qubits. The measurement provides information on the total spin of the two qubits such that entangled qubit states can be postselected.
Our scheme works in analogy to experimental technique already established in the microwave domain but employs an optical channel at room temperature. The use of light greatly enhances the distance over which the qubits can become entangled. The generalization to the optical domain—although relatively straightforward from the experimental point of view—is highly nontrivial and requires a systematic investigation of new sources of decoherence; thermal mechanical noise and optical transmission loss have to be analysed. Such an analysis requires adiabatic elimination of the complex transducer dynamics since the Hilbert space dimension is too large to allow numerical simulations.
Compared to earlier proposals of optomechanical transducers, our strategy requires no time-dependent control. This simplicity leads to modest requirements on the system parameters; optomechanical cooperativity moderately larger than unity is sufficient and large transmission losses can be tolerated. The approach is scalable to generation of multipartite entanglement and represents a crucial step towards quantum networks with superconducting circuits.
The presentation file on workshop on Neutron and X-ray Characterisation on Caloric Materials, introduction to neutron scattering experiment with triple axis spectrometer for material scientist
The Neutrino: Elusive Misfit and Evolutionary DiscoveriesSon Cao
We live in a matrix of neutrinos, the most abundant and perhaps the most elusive of all the known massive particles. The neutrino’s interactions dictate how the Sun shines, how the Sun will evolve, and the dynamics of dying stars. The neutrino, a tangible misfit, also tells us that our theory of the fundamental building blocks of Nature called the “Standard Model” is incomplete. There have been four neutrino-related Nobel prizes in physics awarded since 1995, but to date, the neutrino is still among the most mysterious of all known particles. A recent publication of the T2K experiment, one of the ten most remarkable discoveries of science in 2020, suggests that neutrinos do not respect the charge-conjugation parity-reversal (CP) symmetry, which in turn could explain how our matter-dominated Universe has emerged. The talk will highlight what we have known and what we expect to know in the following decades about this elusive particle. Also, we will discuss how to weigh the extraordinarily tiny mass of the neutrino and detect the CP violation via a quantum mechanical phenomenon called neutrino oscillation.
VSoN Lab Training: A Concept for Neutrino DetectorSon Cao
A concept for Neutrino Detector with plastic scintillator, wavelength shifting fiber and MPPC. NIM is used for signal processing. This is for the hardware training conducted at Vietnam School on Neutrino http://www-he.scphys.kyoto-u.ac.jp/member/nuICISE/vson
My little stories, Vietnam & Experimental High Energy PhysicsSon Cao
This is a science dialog for high school students in Japan. I did tell some little stories of my academic path: from a small village of Vietnam, to Hanoi as undergraduate student, then to United State for Ph.D degree and went to Japan after graduation. The main message from these stories is that I enjoyed my journey which I could not imagine when I was in high school and I hope that they (students in class) will enjoy their journey as well. That is the most important thing in life. Then I expressed my inspiration toward symmetry, one of the most beautiful and important concepts in physics. I gave examples from classical physics to model physics. Then I brought up a very fundamental question how the universe begin if the symmetry is held everywhere. I went through the concept of symmetry breaking, and introduced neutrino particle from which we could find the answer for this fundamental question. I also mentioned about the Nobel and Breakthrough prize last year for the contributions in the neutrino physics which I am working on. In the final part of this section, I mentioned about the benefit obtained if they choose to be particle physicists and what they need to prepare for this career. In the final section, I introduced a bit about Vietnam culture in the corresponding to Japanese culture. This section is required from the school since their student will come to Vietnam for excursion this December. While I made some comparison which I think students can find useful when they explore Vietnam culture, the main message I delivered is that culture, which is not better or worse, just different and we need to respect the difference.
Các nhà khoa học tin rằng, mỗi thiên hà luôn có một dải vật chất tối bao quanh. Vì vậy mà quan sát tương tác giữa các thiên hà cho phép các nhà khoa học tìm hiểu về tính chất của vật chất tối thông qua tương tác giữa chúng. Ở đây, các nhà khoa học dùng kính thiên văn Hubble và kính thiên văn cực lớn của đài quan sát Nam châu Âu để quan sát chùm thiên hà Abell 3827 (1.3 tỷ năm ánh sáng từ Trái Đất, tức là khoảng 1/10 độ tuổi của vũ trụ tính từ thời điểm “Vụ nổ lớn”[3] ). Chùm thiên hà này có 4 thiên hà lớn ở trung tâm. Các nhà khoa học quan sát thấy một thiên hà (trong hình ) có khoảng cách vào cỡ 5000 năm ánh sáng so với dải vật chất tối tương ứng của nó. Bạn có thể hình dung như có bốn người bạn gặp nhau, mỗi người đều mang theo một con chó. Những con chó này đứng sau người chủ của mình, nhưng có một con vì một lý do nào đó mà nó đứng cách xa hơn bình thường so với những con khác (ví dụ: nhỏ con hơn..). Đó có thể là bằng chứng cho sự tương tác giữa vật chất tối không chỉ thông qua lực hấp dẫn.
Francesca Gottschalk - How can education support child empowerment.pptxEduSkills OECD
Francesca Gottschalk from the OECD’s Centre for Educational Research and Innovation presents at the Ask an Expert Webinar: How can education support child empowerment?
Model Attribute Check Company Auto PropertyCeline George
In Odoo, the multi-company feature allows you to manage multiple companies within a single Odoo database instance. Each company can have its own configurations while still sharing common resources such as products, customers, and suppliers.
Introduction to AI for Nonprofits with Tapp NetworkTechSoup
Dive into the world of AI! Experts Jon Hill and Tareq Monaur will guide you through AI's role in enhancing nonprofit websites and basic marketing strategies, making it easy to understand and apply.
2024.06.01 Introducing a competency framework for languag learning materials ...Sandy Millin
http://sandymillin.wordpress.com/iateflwebinar2024
Published classroom materials form the basis of syllabuses, drive teacher professional development, and have a potentially huge influence on learners, teachers and education systems. All teachers also create their own materials, whether a few sentences on a blackboard, a highly-structured fully-realised online course, or anything in between. Despite this, the knowledge and skills needed to create effective language learning materials are rarely part of teacher training, and are mostly learnt by trial and error.
Knowledge and skills frameworks, generally called competency frameworks, for ELT teachers, trainers and managers have existed for a few years now. However, until I created one for my MA dissertation, there wasn’t one drawing together what we need to know and do to be able to effectively produce language learning materials.
This webinar will introduce you to my framework, highlighting the key competencies I identified from my research. It will also show how anybody involved in language teaching (any language, not just English!), teacher training, managing schools or developing language learning materials can benefit from using the framework.
A Strategic Approach: GenAI in EducationPeter Windle
Artificial Intelligence (AI) technologies such as Generative AI, Image Generators and Large Language Models have had a dramatic impact on teaching, learning and assessment over the past 18 months. The most immediate threat AI posed was to Academic Integrity with Higher Education Institutes (HEIs) focusing their efforts on combating the use of GenAI in assessment. Guidelines were developed for staff and students, policies put in place too. Innovative educators have forged paths in the use of Generative AI for teaching, learning and assessments leading to pockets of transformation springing up across HEIs, often with little or no top-down guidance, support or direction.
This Gasta posits a strategic approach to integrating AI into HEIs to prepare staff, students and the curriculum for an evolving world and workplace. We will highlight the advantages of working with these technologies beyond the realm of teaching, learning and assessment by considering prompt engineering skills, industry impact, curriculum changes, and the need for staff upskilling. In contrast, not engaging strategically with Generative AI poses risks, including falling behind peers, missed opportunities and failing to ensure our graduates remain employable. The rapid evolution of AI technologies necessitates a proactive and strategic approach if we are to remain relevant.
The French Revolution, which began in 1789, was a period of radical social and political upheaval in France. It marked the decline of absolute monarchies, the rise of secular and democratic republics, and the eventual rise of Napoleon Bonaparte. This revolutionary period is crucial in understanding the transition from feudalism to modernity in Europe.
For more information, visit-www.vavaclasses.com
Honest Reviews of Tim Han LMA Course Program.pptxtimhan337
Personal development courses are widely available today, with each one promising life-changing outcomes. Tim Han’s Life Mastery Achievers (LMA) Course has drawn a lot of interest. In addition to offering my frank assessment of Success Insider’s LMA Course, this piece examines the course’s effects via a variety of Tim Han LMA course reviews and Success Insider comments.
Read| The latest issue of The Challenger is here! We are thrilled to announce that our school paper has qualified for the NATIONAL SCHOOLS PRESS CONFERENCE (NSPC) 2024. Thank you for your unwavering support and trust. Dive into the stories that made us stand out!
Palestine last event orientationfvgnh .pptxRaedMohamed3
An EFL lesson about the current events in Palestine. It is intended to be for intermediate students who wish to increase their listening skills through a short lesson in power point.
Embracing GenAI - A Strategic ImperativePeter Windle
Artificial Intelligence (AI) technologies such as Generative AI, Image Generators and Large Language Models have had a dramatic impact on teaching, learning and assessment over the past 18 months. The most immediate threat AI posed was to Academic Integrity with Higher Education Institutes (HEIs) focusing their efforts on combating the use of GenAI in assessment. Guidelines were developed for staff and students, policies put in place too. Innovative educators have forged paths in the use of Generative AI for teaching, learning and assessments leading to pockets of transformation springing up across HEIs, often with little or no top-down guidance, support or direction.
This Gasta posits a strategic approach to integrating AI into HEIs to prepare staff, students and the curriculum for an evolving world and workplace. We will highlight the advantages of working with these technologies beyond the realm of teaching, learning and assessment by considering prompt engineering skills, industry impact, curriculum changes, and the need for staff upskilling. In contrast, not engaging strategically with Generative AI poses risks, including falling behind peers, missed opportunities and failing to ensure our graduates remain employable. The rapid evolution of AI technologies necessitates a proactive and strategic approach if we are to remain relevant.
The Roman Empire A Historical Colossus.pdfkaushalkr1407
The Roman Empire, a vast and enduring power, stands as one of history's most remarkable civilizations, leaving an indelible imprint on the world. It emerged from the Roman Republic, transitioning into an imperial powerhouse under the leadership of Augustus Caesar in 27 BCE. This transformation marked the beginning of an era defined by unprecedented territorial expansion, architectural marvels, and profound cultural influence.
The empire's roots lie in the city of Rome, founded, according to legend, by Romulus in 753 BCE. Over centuries, Rome evolved from a small settlement to a formidable republic, characterized by a complex political system with elected officials and checks on power. However, internal strife, class conflicts, and military ambitions paved the way for the end of the Republic. Julius Caesar’s dictatorship and subsequent assassination in 44 BCE created a power vacuum, leading to a civil war. Octavian, later Augustus, emerged victorious, heralding the Roman Empire’s birth.
Under Augustus, the empire experienced the Pax Romana, a 200-year period of relative peace and stability. Augustus reformed the military, established efficient administrative systems, and initiated grand construction projects. The empire's borders expanded, encompassing territories from Britain to Egypt and from Spain to the Euphrates. Roman legions, renowned for their discipline and engineering prowess, secured and maintained these vast territories, building roads, fortifications, and cities that facilitated control and integration.
The Roman Empire’s society was hierarchical, with a rigid class system. At the top were the patricians, wealthy elites who held significant political power. Below them were the plebeians, free citizens with limited political influence, and the vast numbers of slaves who formed the backbone of the economy. The family unit was central, governed by the paterfamilias, the male head who held absolute authority.
Culturally, the Romans were eclectic, absorbing and adapting elements from the civilizations they encountered, particularly the Greeks. Roman art, literature, and philosophy reflected this synthesis, creating a rich cultural tapestry. Latin, the Roman language, became the lingua franca of the Western world, influencing numerous modern languages.
Roman architecture and engineering achievements were monumental. They perfected the arch, vault, and dome, constructing enduring structures like the Colosseum, Pantheon, and aqueducts. These engineering marvels not only showcased Roman ingenuity but also served practical purposes, from public entertainment to water supply.
Experimental Neutrino Physics Concepts in Nutshell
1. Experimental Neutrino
Physics Concepts in Nutshell
Son Cao
(KEK/J-PARC)
i.e A practical guide for
How to THINK/ADDRESS the things as a neutrino experimentalist
2. !2
To simplify something such as…
(GeV)νE
0.5 1 1.5 2 2.5 3
Osc.Prob
0
0.02
0.04
0.06
0.08
0.1
flux
µ
νOff-axis°2.5
ν, NH,°=0cpδ
ν, NH,°=270cpδ
ν, NH,°=0cpδ
ν, NH,°=270cpδ
eν→µν,eν→µν
Hypothesis
(parameters)
observed data
T2K has made the first observation of electron neutrino
appearance in a muon neutrino beam…with
a significance of 7.3𝜎 over the the hypothesis of sin2
2θ13 = 0
data-based statement
Experiment setup
4. !4
Some confessions
• Call me “teaching assistant” or senpai ( 先輩 in Japanese) rather than lecturer
• I shared almost same background as most of Vietnamese theoretical students,
and found lot of difficulties for first years of Ph.D in experimental HEP
• I requested this lecture since I thought it could make bridge between young
theorists and young experimentalist. But it may be very bias…
• My background is with accelerator-based neutrino experiments, might not
enough to generalize for other fields.
• Many materials are borrowed from other talks, but sometime missing citation
• Your feedback is valuable and I’m alway be a student for listening your
feedback
5. !5
Contents
• Basic steps as scientists
• Ask question, Design, Build experiment, Collect data and Make
statement based on data
• Examples with neutrino experiments
• Neutrino detection principle
• A complicated, interdisciplinary field of Particle & Nuclear physics,
Material science, Mechanics, Electronics, Data mining
• Some selected topics (personal choices)
1) Signal and background
2) Hypothesis test
3) Sensitivity &Parameter estimation
4) Neutrino energy reconstruction
5) Systematics
6) Monte Carlo usage
Number of illustrations will be shown
Source code: https://github.com/cvson/nushortcourse
Feel free to download and play!
6. !6
Basic steps as scientists
Ask a question about Nature
Formulate the hypothesis to test
Experiment(s)
Design
Exp.
Build
Exp.
Collect
data
Make
data-based
statement
7. !7
Basic steps as scientists
All connected on one foot!
Ask a question about Nature
Formulate the hypothesis to test
Design
Exp.
Build
Exp.
Collect
data
Make
data-based
statement
Experiment(s)
8. !8
An example: searchνμ → νe
Do muon neutrinos transform into
electron neutrinos at given distance
of travel?
“neutrino oscillations” from Oyama-san & Boris
Why is addressing this question important?
• Confirm non-zero mixing angle, 𝜃13 >0 or
set lower limit for mixing angle 𝜃13 (𝜃13 >𝛼)
• If non-zero, open the door to measure 𝛿CP,
might a source of matter-antimatter
asymmetry in the Universe
What have you already know at the
time question posed?
• Neutrino oscillations confirmed
• Some upper limit on 𝜃13 from reactor
• etc…
https://arxiv.org/abs/hep-ex/0106019
Supported knowledge
at point A at point B
L (distance)
9. !9
Do muon neutrinos transform into
electron neutrinos at given distance
of travel?
Driven by
theoretical models
An example: searchνμ → νe
T2K, NOvA, etc
Goal of your experiments
Level 2: Estimate parameters which
used to describe your model
• Model give good description of our
data?
• Allowed region for sin2 𝜃13 at 68%
C.L. (1σ) or 90% C.L., etc…
Level 1: Test theoretical hypothesis; basically
yes/no question
• At some C.L., we observe the appearance of
electron neutrino
• At some C.L., we reject the hypothesis that
electron neutrino appeared
Hypothesis test and parameter
estimation will be discussed later
may not know yet?
10. !10
An example: searchνμ → νe
Do muon neutrinos transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
In principle, how can we conduct the search?
1. Need source of
2. Put detector at some distance from source
3. Look for appeared from source in detector
Is it simple for you?
νμ
νe νμ
νμ
at point A at point B
L (distance)
11. !11
An example: searchνμ → νe
In principle, how we can perform the search?
1. Need source of
2. Put detector at some distance from source
3. Look for appeared from source in detector
νμ
νe νμ
νμ
Do muon neutrino transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
(1)
(2) (3)
(4) (5)
(1) How can source be created? How well you understand the
source? (composition, density, energy, timing, etc)
(2) What kind of detector you need? how big it is? Where do you put?
(3) How can you choose distance? Typically your detector can’t move
from place to place.
(4) How can you identify ?
(5) How do you know it coming from source but not others?
νe
νμ
12. !12
Design experiments
Do muon neutrino transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
Design
Exp.
(1) How can source be created? How well you understand the
source? (density, energy, timing, etc)
(2) What kind of detector you need? how big it is? Where do you put?
(3) How can you choose distance? Typically your detector can’t move
from place to place.
(4) How can you identify ?
(5) How do you know it coming from source but not others?
νe
νμ
You have to address at least below questions when designing experiment. In reality,
there are many more questions to address.
• Think big, make cheap. HEP is really expensive game
• What available facilities to use, ref. T2K
• How do you know you have the best among many possible experiment setups
• Most important: guarantee of success (doesn’t mean you will get signal, but your
experiment should achieve some measurement w/ unprecedented level of precision)
can be conservative
13. !13
Short example answer: T2K
Do muon neutrino transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
Design
Exp.
(1) How can source be created? How well you understand the
source? (density, energy, timing, etc)
(2) What kind of detector you need? how big it is? Where do you put?
(3) How can you choose distance? Typically your detector can’t move
from place to place.
(4) How can you identify ?
(5) How do you know it coming from source but not others?
νe
νμ
(1)
(2)
(3)
(4)
(5)
GPS, timing synchronize
details from Oyama-san
14. !14
Short example answer: NOvA
Do muon neutrino transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
Design
Exp.
(1) How can source be created? How well you understand the
source? (density, energy, timing, etc)
(2) What kind of detector you need? how big it is? Where do you put?
(3) How can you choose distance? Typically your detector can’t move
from place to place.
(4) How can you identify ?
(5) How do you know it coming from source but not others?
νe
νμ
(1)
(2)
(3)
(4)
(5)
GPS,timingsynchronize
15. !15
Design experiments for νμ → νe
Do muon neutrino transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
Design
Exp.
Neutrino Energy (GeV)
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
)eν→µνoreν→µνProb.(
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
2
eV-3
10×|=2.5032
2
m∆|
2
eV
-5
10×|=7.6021
2
m∆|
=1.0023θ22
sin
=0.8712
θ22
sin
=0.0813
θ22
sin
L=295 km
=0
CP
δ, NH,ν
°
=270CP
δ,NH,ν
=0CP
δ, NH,ν
°
=270CP
δ, NH,ν
T2K 𝜈 energy range
Neutrino Energy (GeV)
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
)eν→µνoreν→µνProb.(
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
2
eV-3
10×|=2.5032
2
m∆|
2
eV
-5
10×|=7.6021
2
m∆|
=1.0023θ22
sin
=0.8712
θ22
sin
=0.0813
θ22
sin
L=810 km
=0
CP
δ, NH,ν
°
=270CP
δ,NH,ν
=0CP
δ, NH,ν
°
=270CP
δ, NH,ν
NOvA 𝜈 energy range
At sin22𝜃13 = 0.08,
Prob. is around 5%
(depend on 𝛿CP value;
smaller for anti-neutrino)
Since oscillation probability depends on neutrino
energy, it’s important to know energy of incoming
neutrinos.
https://github.com/cvson/nushortcourse/tree/master/OscCalculatorPMNS
Let’s look at basic quality: probability as function of
energy. Basically this is counting experiment
νμ → νe
T2K baseline NOvA baseline
Nνe
∼ Prob.(νμ → νe)
16. !16
Design experiments for νμ → νe
Do muon neutrino transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
Design
Exp.
Neutrino Energy (GeV)
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
)eν→µνoreν→µνProb.(
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
2
eV-3
10×|=2.5032
2
m∆|
2
eV
-5
10×|=7.6021
2
m∆|
=1.0023θ22
sin
=0.8712
θ22
sin
=0.0413
θ22
sin
L=295 km
=0
CP
δ, NH,ν
°
=270CP
δ,NH,ν
=0CP
δ, NH,ν
°
=270CP
δ, NH,ν
T2K 𝜈 energy range
Neutrino Energy (GeV)
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
)eν→µνoreν→µνProb.(
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
2
eV-3
10×|=2.5032
2
m∆|
2
eV
-5
10×|=7.6021
2
m∆|
=1.0023θ22
sin
=0.8712
θ22
sin
=0.0413
θ22
sin
L=810 km
=0
CP
δ, NH,ν
°
=270CP
δ,NH,ν
=0CP
δ, NH,ν
°
=270CP
δ, NH,ν
NOvA 𝜈 energy range
At sin22𝜃13 = 0.04,
Prob. is around 2.5%
(depend on 𝛿CP value;
smaller for anti-neutrino)
The probability depends on values of sin22𝜃13
(also 𝛿CP). Smaller probability means with the
same amount of neutrino, number of event you
will observe is smaller in average.
https://github.com/cvson/nushortcourse/tree/master/OscCalculatorPMNS
Let’s look at basic quality: probability as function of
energy. Basically this is a counting experiment
νμ → νe
Nνe
∼ Prob.(νμ → νe)
T2K baseline NOvA baseline
17. !17
Design experiments for νμ → νe
Do muon neutrino transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
Design
Exp.
sin2
2θμe=sin2
2θ13.sin2
θ23
Normally, capability (how good/“sensitivity”) of designed
detector much be evaluated for various scenario of
underlying parameters, e.g.:
- Range of parameter in which detector can explore
- At what values of parameter, detector can make
observation, measurement with some precision
hep-ex/0106019
18. !18
Build experiment
Do muon neutrino transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
Design
Exp.
Build
Exp.
• Typically, neutrino detector is big, really big (with
few exception) since cross section is small
• Normally, detector is located in deep
underground to cancel the noise from cosmic ray
• big MONEY for this, NOvA is on surface
although it is design to be underground
• India controversy on INO building due to
natural conservation
• etc…
Example/ Details in Oyama-san
Many additional considerations,
• How to access it?
• How to monitor it?
• How to maintain it?
19. !19
Collect data
• Data taking needs time: from year to decade
(ex, Super-K 20 years, T2K 7 years)
• Your detector is NOT at same condition during
data-collecting period
• Detector position can be unintentionally
moved due to, eg, Earthquake
• Some photon detectors can be out-of-
function or mis-behaved
• Light yield (No. of photon per fixed amount
of deposited energy) can be changed due
to water quality, aging of scintillator, etc…
• etc…
Do muon neutrino transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
Design
Exp.
Build
Exp.
Collect
data
You need to take good data. Keep
condition of your detector in control
as much as you can
Atmospheric ! FCFV events
20. !20
Make statement based on data
• The statement is never simple like “yes” or “no”
• It always goes with level of uncertainty/
confidence
• If an observation is claimed, some parameters
are estimated, if not, a limit is presented
• E.g:
Do muon neutrino transform into
electron neutrinos at given distance
of travel?
T2K, NOvA, etc
Design
Exp.
Build
Exp.
Collect
data
make
statement
10.1103/PhysRevLett.112.061802
21. !21
Make statement based on data
Hypothesis
(parameters)
Data
Probability
How interpolate data and parameters to hypothesis, and interpolation of
probability for uncertain/random events to happen.
Two schools of thought: Frequentist and Bayesian
22. Frequentist vs. Bayesian
!22
Frequentist
• Data are a repeatable
random sample, frequency
• Underlying parameters
remains constant
• State of knowledge worlds is
sharp
Bayesian
• Data are observed from the
realized sample, prior
knowledge
• Parameters are unknown and
described probabilistically
• State of knowledge worlds is
always updated
Not going to detail. We can discuss after class
It’s about philosophy of probability for
and uncertain/random event happen,
e.g World cup’s history record play any
role in the result of current/future
soccer match? (in very rude sense)
23. Frequentist vs. Bayesian
!23
Frequentist uses impeccable logic to deal with an issue of no
interest to anyone
Bayesian addresses the question everyone is interested in, by
using assumptions no-one believes
L. Lyon
Experimentalists tend to make data-based
statements with both approaches
Frequentist
Bayesian
Reach quite similar conclusion:
CP conserving values lies outside of 2σ confidence/credible interval
T2K, Neutrino 2018
24. !24
Make your own path?
• Neutrinos are Majorana or Dirac particles?
• CP symmetry violated in neutrino oscillation?
• Neutrino mass is inverted or normal?
• Is there 4th generation of neutrino?
• etc…
More questions from other lecturers
25. !25
Make your own path?
• Neutrinos are Majorana or Dirac particles?
• CP symmetry violated in neutrino oscillation?
• Neutrino mass is inverted or normal?
• Is there 4th generation of neutrino?
• etc…
Some experiments are going to build (DUNE, HYPER-K, JUNO,etc)
Some are waiting for your!
More questions from other lecturers
27. !27
Neutrino detection principle
Neutrinos basically can’t see/directly but we know it via their trace when interacting with nucleon/nuclei
with help of photon detectors/sensors
This is just illustration. Many detection technique out there.
Photon detectors
Event reconstruction
Neutrino interactions
Youwanttoknowthisguys
what you observed
nucleon/nuclei
Charged particles
(𝜇, e,𝝅, p,…)
Neutral particles
(n,𝜋0,𝛾,…)
“eyes”
“lot of eyes”
“Pattern of light induced by
neutrino interaction”
28. !28
Neutrino detection is complicate
PN physics
Material science
Mechanics
Electronic
Data mining
Neutrino detection is a complicate, interdisciplinary field
29. !29
Involved Particle and Nuclear physics
PN physics
Material science
Mechanics
Electronic
Data mining
F. Sanchez, neutrino 2018
• Neutrino-nucleon/nuclei interaction is complicated
• For oscillation analysis, you need, essentially
(1) Particle identity
(2) Neutrino energy
Van Nguyen’s lecture
Charged particles
neutral particles
Based on induced charged particle
in final state interaction
30. !30
Material science in neutrino experiments
PN physics
Material science
Mechanics
Electronic
Data mining
T2K far detector use water; NOvA use liquid scintillator; MINOS
used magnetized steel, OPERA used Emulsion, etc…?
• T2K, NOvA needs to identify both 𝜈 𝜇 and 𝜈e
• MINOS focus on 𝜈 𝜇 and its antineutrino
• OPERA need to see 𝜈 𝝉
Material selection depends on particle you want to detect and
its properties. Also detector size & our understanding of
neutrino interaction on selected material are important factors.
Water
Liquid scintillator
magnetized iron
Emulsion
31. !31
Mechanics in neutrino detection
PN physics
Material science
Mechanics
Electronic
Data mining
• In Nov 2001, Super-K suffered a serious blow, ~700 PMT tubes
exploded (cost $3000 per each) (5000 PMT remain undamaged)
• Cause: one tubes (contain a vacuum) exploded, released
energy, caused shock wave —> chain reaction of explosion
• To mitigate this possibility: Acrylic shield is developed and used
Bare PMT
PMT w/ acrylic shield
Similar structure in aquarium
One example
32. !32
Electronics in neutrino experiments
PN physics
Material science
Mechanics
Electronic
Data mining
• Number of photon sensor/ “eyes” per each detector is often very
large: 13,000 channels in Super-K, 334,000 channels in NOvA far
detector
• With many “eyes”, you need a “nervous” system to manipulate
and collect data efficiently
• “Eyes” don’t not always open, no need and not good for
lifetime of electronics
• “Eyes” actually operate when receiving “trigger” signal, and
often within a predefined time window
Depend on how often your detector
get data; how many events interact in
your detector in a time window, etc…
Ex: NOvA electronics at Near Detector
Ex: INGRID data acquisition
33. !33
Data mining in neutrino experiments
PN physics
Material science
Mechanics
Electronic
Data mining
• How do you know this is likely due to 𝜈e interaction?
• Basically, you need guidance from simulation
• The method is something like this:
1. Build detector simulation to simulate what you can observe when particles
enter your detector (so called Geant 3,4 and your detector geometry)
2. Simulate neutrinos (you know true info. such as neutrino type, energy,
direction, interaction point in detector)
3. Obtain pattern for simulated neutrino events and store as a library
4. Compare your data pattern with library and see how likely data match with
what simulated events
34. !34
Neutrino detection is complicate
PN physics
Material science
Mechanics
Electronic
Data mining
Neutrino detection is a complicate, interdisciplinary field.
You don’t need to know all of these. Expert in one field is probably enough.
35. !35
Before going to some selected topics, let’s have a
quick digest on Histogram, a conventional way to
visualize your data in HEP
• Taking about experiment is to talk about data.
• To make number less boring, a sexy way to
visualize it is invented, so-call Histogram
Some lecturers showed but not
sure if you understand it.
36. !36
A histogram is an accurate representation of the
distribution of numerical data. It is an estimate of the
probability distribution of a continuous variable (quantitative
variable) and was first introduced by Karl Pearson.
https://en.wikipedia.org/wiki/Histogram
Histogram
Go google image and type: histogram neutrino
39. !39
100 entries in your sample
Histogram
Values of variable X
10 15 20 25 30 35 40
Numberofentries
0
2
4
6
8
10
12
14
Histogram
Entries 100
Mean 24.82
RMS 4.83
Can you guess data following
which distribution?
https://github.com/cvson/nushortcourse/tree/master/basic01
40. !40
1000 entries in your sample
(as data sample increased)
Histogram
Values of variable X
10 15 20 25 30 35 40
Numberofentries
0
10
20
30
40
50
60
70
80
90
Histogram
Entries 1000
Mean 25.14
RMS 5.131
Can you guess data following
which distribution?
https://github.com/cvson/nushortcourse/tree/master/basic01
41. !41
Histogram
Values of variable X
10 15 20 25 30 35 40
Numberofentries
0
100
200
300
400
500
600
700
800
Histogram
Entries 10000
Mean 25.03
RMS 4.906
Can you guess data following
which distribution?
https://github.com/cvson/nushortcourse/tree/master/basic0110000 entries in your sample
(as data sample increased)
42. !42
Histogram
Values of variable X
10 15 20 25 30 35 40
Numberofentries
0
1000
2000
3000
4000
5000
6000
7000
8000
Histogram
Entries 100000
Mean 25
RMS 4.92
Can you guess data following
which distribution?
100,000 entries in your sample
(as data sample increased)
https://github.com/cvson/nushortcourse/tree/master/basic01
43. !43
Histogram
Histogram
Entries 100000
Mean 25
RMS 4.92
/ ndf2
χ 42.97 / 27
Prob 0.02633
Constant 31.6±7985
Mean 0.0±25
Sigma 0.012±4.993
Values of variable X
10 15 20 25 30 35 40
Numberofentries
0
1000
2000
3000
4000
5000
6000
7000
8000
Histogram
Entries 100000
Mean 25
RMS 4.92
/ ndf2
χ 42.97 / 27
Prob 0.02633
Constant 31.6±7985
Mean 0.0±25
Sigma 0.012±4.993
Indeed, it generated with Gaussian
distribution with Mean = 25 and RMS =5
100,000 entries in your sample
(as data sample increased)
Your data might be underlying a particular
distribution/pattern but it might not be easy
to reveal if your data sample is not enough.
https://github.com/cvson/nushortcourse/tree/master/basic01
44. !44
1) Underlying distribution
Two-dimensional histogram
Values of variable X
10 20 30 40
ValuesofvariableY
10
20
30
Histogram
Entries 10000
Mean x 24.99
Mean y 19.94
RMS x 5.005
RMS y 4.01
Histogram
Entries 10000
Mean x 24.99
Mean y 19.94
RMS x 5.005
RMS y 4.01
Histogram_py
Entries 10000
Mean 19.94
RMS 4.01
0 200 400 600 800 1000
ValuesofvariableY
10
20
30
Histogram_px
Entries 10000
Mean 24.99
RMS 5.005
Values of variable X
10 20 30 40
0
200
400
600
800
Histogram_px
Entries 10000
Mean 24.99
RMS 5.005
The histogram can be shown in more than
one-dimension
https://github.com/cvson/nushortcourse/tree/master/basic01
45. !45
1) Underlying distribution
Two-dimensional histogram
0
10
20
30
40
50
60
70
80
90
Values of variable X
10 20 30 40
ValuesofvariableY
10
20
30
Histogram
Entries 10000
Mean x 24.99
Mean y 19.94
RMS x 5.005
RMS y 4.01
0
10
20
30
40
50
60
70
80
90
Histogram
Entries 10000
Mean x 24.99
Mean y 19.94
RMS x 5.005
RMS y 4.01
Histogram_py
Entries 10000
Mean 19.94
RMS 4.01
0 200 400 600 800 1000
ValuesofvariableY
10
20
30
Histogram_px
Entries 10000
Mean 24.99
RMS 5.005
Values of variable X
10 20 30 40
0
200
400
600
800
Histogram_px
Entries 10000
Mean 24.99
RMS 5.005
https://github.com/cvson/nushortcourse/tree/master/basic01
Zoom
-in
bin, ~90 entries
bin width in Y-view
bin width in X-view
Bin width might be varies, not constant
46. !46
Some selected topics
1) Signal and background
2) Hypothesis test
3) Sensitivity & Parameter estimation
4) Neutrino energy reconstruction
5) Systematics
6) Monte Carlo usage 101
47. !47
Signal and Background
Signal: For what you are placing as object to study, e.g. 𝜈e from 𝜈 𝜇 beam
Background: Anything else
Measurement is performed on a selected sample which contains both signal and
background. Background is always there since your sample selection is not perfect.
• It’s important to define clearly what’s signal. Sometime it’s not straightforward, e.g.
• For oscillation analysis, 𝜈e from 𝜈 𝜇 beam observed at Far Detector is signal but
intrinsic 𝜈e is background
• For understanding neutrino source composition, cross-section at Near
Detector, intrinsic 𝜈e is signal
• In selected sample, ratio of signal to background is matter, not just number of signal
alone. E.g, you want to find a Japanese students, which class you should go?
• A class with 10 Japanese students and 1000 non-Japanese
• A class with 5 Japanese students and 5 non-Japanese
48. !48
Signal vs. Background
Electron with EM activity, look more fuzzy than muon.
NOvA, scintillator technique
Karol, Miura-san’s lecture & training
You need machinery/tool to separate signal from background. The “fuzzy” thing is quantized into
one or multiple variables which is used to build likelihood of data to be signal or to be
background. (Some machine learning can skip the middle steps.)
Super-Kamiokande,
Water-cherenkov technique
This guides your eyes but science need
quantitative thing
52. !52
Signal vs. Background: ID parameter
Partice ID parameter
-10 -8 -6 -4 -2 0 2 4 6 8 10
0
50
100
150
200
250
300
350
Super Kamiokande IV 2166.5 days : Monitoring
e-like muon-like
Numberofevents
• To make selection (or decision rule/boundary),
typically a likelihood of data to be signal/
background is built.
• Background is unavoidable
• Enhance signal and suppress background is
important in HEP analysis, especially in
neutrino experiment where statistics is limited.
• The improvement can be from hardware side
or software side
• It’s (big) money can be saved when you can
improve your selection since it is effectively
equivalent to collecting more data or enlarge
your detector
Example with atmospheric neutrinos observed in Super-Kamiokande
Red curve is what machine learned
Black dots are your data
53. !53
Fun exercise
We have 19 students, 10 are from Vietnam.
Can you teach computer how to identify them?
54. !54
Fun exercise
Partice ID parameter
-10 -8 -6 -4 -2 0 2 4 6 8 10
0
50
100
150
200
250
300
350
Super Kamiokande IV 2166.5 days : Monitoring
e-like muon-like
Numberofevents
Nationality ID parameter
Numberofstudents
Vietnamese Non-Vietnamese
55. !55
Signal and Background
Reconstructed Energy (GeV)ν
0 0.2 0.4 0.6 0.8 1 1.2
NumberofEvents
0
2
4
6
8
10
12
eν→µν
eν→µν
NC
intrinsiceν/eν
intrinsicµν
intrinsicµν
T2K Run1-8 PreliminaryFinal systematics pending
NOvA, Neutrino 2018
T2K, Moriond 2018
Examples on selected sample in real
analyses from neutrino experiments
56. !56
Some selected topics
1) Signal and background
2) Hypothesis test
3) Sensitivity & Parameter estimation
4) Neutrino energy reconstruction
5) Systematics
6) Monte Carlo usage 101
57. !57
Hypothesis test
Hypothesis H0: 𝜃13 = 0
Errors are alway there. What matter is the probability to
make error and how much you can tolerate the error
In testing a hypothesis H0, there are two kinds of errors:
• Type I: reject H0 although it is true
• i.e. “observe” 𝜃13 ≠ 0 although the true is 𝜃13 = 0
• Type II: accept H0 although it is false
• i.e. “fail to observe” 𝜃13 ≠ 0 although the true is 𝜃13 ≠ 0
58. !58
Hypothesis test: Error toleration
Hypothesis H0: Train is arrived at train station on-time.
In Japan, toleration (unconventionally?)
is within 1minute
E.g. Train company apologize for 20s-early
departure
In Vietnam, toleration (unconventionally?)
is order of 30 minutes
(Rarely you hear apologies for delay in arrival)
59. !59
Hypothesis test
When you make statement, it should include two errors ( 𝛼, β).
• (1- 𝛼) (%) is normally mentioned as Confidence Level (C.L.), set at
beginning as toleration level, e.g 0.05 or 95% C.L.
• (1-β) (%) is probability that you make “observation” at (1- 𝛼) (%) C.L.
We care this error especially when e.g. due to statistic fluctuation, you are
very lucky to make observation
Hypothesis H0: 𝜃13 = 0
In testing a hypothesis H0, there are two kinds of errors:
• Type I: reject H0 although it is true
• i.e. “observe” 𝜃13 ≠ 0 although the true is 𝜃13 = 0
• Type II: accept H0 although it is false
• i.e. “fail to observe” 𝜃13 ≠ 0 although the true is 𝜃13 ≠ 0
denoted by 𝛼
denoted by β
60. !60
Hypothesis test: example from T2K
T2K measurement of “appearance” of
• Background: 6.5 events
• Data: 9 events
Value of toy experiments
0 10 20 30 40 50Numberoftoyexperiments
1−
10
1
10
2
10
3
10
4
10
5
10
6
10
7
10
if you are familiar with ROOT, can try
p-value = 1-ROOT::Math::poisson_cdf(8,6.5);
https://github.com/cvson/nushortcourse/tree/master/basic01
νe
p-value =
No. of toy exp., value > 9
All of toy exp.
= 0.208
p-value> 0.05: you failed to reject hypothesis H0 and
the result is statistically nonsignificant.
Hypothesis H0: data agree with background
61. !61
Hypothesis test: Example from NOvA
NOvA appearance of
• Background: 5.3 events
• Data: 18 events
P value is 5.27e-06
pvalue = 1-ROOT::Math::poisson_cdf(17,5.3);
sigma = TMath::NormQuantile(1-pvalue);
Value of toy experiments
0 10 20 30 40 50
Numberoftoyexperiments
1−
10
1
10
2
10
3
10
4
10
5
10
6
10
7
10
8
10
P value is 5.27e-06
5.41634e-06
4.39985
https://github.com/cvson/nushortcourse/tree/master/basic01
νe
p-value =
No. of toy exp., value > 9
All of toy exp.
= 5.27e − 6
p-value< 0.05: you reject hypothesis H0 and
the result is statistically significant, but not observation yet!
< 5σ (level of discovery)
62. !62
Some selected topics
1) Signal and background
2) Hypothesis test
3) Sensitivity & Parameter estimation
4) Neutrino energy reconstruction
5) Systematics
6) Monte Carlo usage 101
63. !63
Sensitivity
You might hear, e.g.
• T2K has good “sensitivity” on CP violation
• NOvA has good “sensitivity” on mass hierarchy
“Good” sensitivity means you can reject some
hypothesis, i.e make observation of something, with high
confidence level (1-𝛼)(%) with high probability (1-β)(%)
For sensitivity, normally only quote C.L while keep (1-β)
(%) = 50%
without mentioning
to data
64. !64
Prediction of event rate in detector
To give some sense on sensitivity and parameter
estimation, we will follow one specific example
Ni = Φflux × σ × ϵ
[GeV]νE
0 5 10 15 20 25
p.o.t.]21
/50MeV/1x102
Flux[/cm
1−
10
1
10
2
10
3
10
4
10
5
10
6
10
flux at Super-K with 250 kA operationµνT2K
fluxµν
fluxµν
fluxeν
fluxeν
flux at Super-K with 250 kA operationµνT2K
[GeV]µν
True E
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
/GeV/atom]2
[cmν/Eσ
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0.22
36−
10×
O
16
8Cross section on
Total
CCQE
πCC1
CC coherent
CC other
NC
O
16
8Cross section on
Energy [GeV]ν
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
p.o.t]21
Events[/50MeV/1x10
0
1
2
3
4
5
3
10×
Event rate at Near Detector (1ton H20)
Total
CCQE
πCC1
CC coherent
CC other
NC
Event rate at Near Detector (1ton H20)
https://github.com/cvson/nushortcourse/tree/master/eventpred
= x
Detection efficiency
assume a constant (<1)
This assumes there is no oscillation
68. !68
Prediction with oscillation
Neutrino Energy [GeV]
0 0.5 1 1.5 2 2.5 3
NumberofEvents
0
10
20
30
40
50
60
70
80
Data
Unoscillated Prediction
=0.52)θ22
=0.00239,sin2
m∆(
=0.95)θ22
=0.00239,sin2
m∆(
=0.95)θ22
=0.00299,sin2
m∆(
Neutrino Energy [GeV]
0 0.5 1 1.5 2 2.5 3
Data/UnoscillatedPred. 0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
⃗o = (Δm2
21, Δm2
31, θ12, θ13, θ23, δCP)
Ni( ⃗o ) = Φflux × σ × ϵ × P( ⃗o )
Data is well described by the blue line. But is it
the “best” parameter yet?
⃗otrue
69. !69
Parameter estimation
log χ2
=
∑
i
2(Nexp. − Nobs.) − 2Nobs. . log(Nexp./Nobs.)
Neutrino Energy [GeV]
0 0.5 1 1.5 2 2.5 3
NumberofEvents
0
10
20
30
40
50
60
70
80
Data
Unoscillated Prediction
Neutrino Energy [GeV]
0 0.5 1 1.5 2 2.5 3
Data/Prediction
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
• When we talk about parameter, we need
predefine a mode
• Given data and model, how do we estimate
parameter?
• you need quantify the difference
between data and prediction at various
parameter
• Method to quantify is not unique, e.g
Maximum likelihood
χ2
( ⃗ok, ⃗otrue) =
∑
i
χ2
(Ni( ⃗ok), Ni( ⃗otrue))
P(νμ → νμ) ∼ 1 − sin2
2θ23 sin2
Δm2
31L
4Eν
Nexp
Nobs
bin i
70. !70
What’s the best parameters to describe your data?
https://github.com/cvson/nushortcourse/tree/master/sensitivity
Illustration for comparing data with various prediction on the
left with 𝜒2 calculated corresponding on the right. From this,
confidence intervals are extracted (not going to detail now)
χ2
72. • Essentially, your experiment is measuring oscillation probability with
some uncertainty
• If measured probability is definitely larger than 0, you will have
allowed region respectively to some C.L.
!72
Excluded and Allowed region
xbest − a < P(νμ → νx) ∼ sin2
2θ23 sin2
Δm2
31L
4Eν
< xbest + b
θ22
sin
3−
10 2−
10 1−
10 1
)4
/c2
(eV2
m∆
5−
10
4−
10
3−
10
2−
10
1−
10
1
10
2
10
3
10
P(νμ→νx)<0.01
P(νμ→νx)>0.4
Excluded region
• If measured probability is not
definitely larger than 0 but upper
limit is known with some C.L., you
will have excluding region (ref.
sterile neutrino search)
P(νμ → νx) ∼ sin2
2θ23 sin2
Δm2
31L
4Eν
< xlimit
do you understand this?
73. !73
Excluded and Allowed region
• Essentially, your experiment is measuring oscillation probability with
some uncertainty
• If measured probability is definitely larger than 0, you will have
allowed region respectively to some C.L.
xbest − a < P(νμ → νx) ∼ sin2
2θ23 sin2
Δm2
31L
4Eν
< xbest + b
• If measured probability is not
definitely larger than 0 but upper
limit is known with some C.L., you
will have excluding region (ref.
sterile neutrino search)
P(νμ → νx) ∼ sin2
2θ23 sin2
Δm2
31L
4Eν
< xlimit
Shape of allowed region is quite different. Why?
θ22
sin
3−
10 2−
10 1−
10 1
)4
/c2
(eV2
m∆
5−
10
4−
10
3−
10
2−
10
1−
10
1
10
2
10
3
10
Excluded region
P(νμ→νx)<0.01
P(νμ→νx)>0.4
Allowedregion
smooth the rapid variation
74. !74
Allowed region
• Essentially, your experiment is measuring oscillation probability with
some uncertainty
• If measured probability is definitely larger than 0, you will have
allowed region respectively to some C.L.
xbest − a < P(νμ → νx) ∼ sin2
2θ23 sin2
Δm2
31L
4Eν
< xbest + b
• If measured probability is not
definitely larger than 0 but upper
limit is known with some C.L., you
will have excluding region (ref.
sterile neutrino search)
P(νμ → νx) ∼ sin2
2θ23 sin2
Δm2
31L
4Eν
< xlimit
Shape of allowed region is quite different. Why?
75. !75
Some selected topics
1) Signal and background
2) Hypothesis test
3) Sensitivity & Parameter estimation
4) Neutrino energy reconstruction
5) Systematics
6) Monte Carlo usage 101
76. !76
Why is it important?
• Oscillation probability depends on energy
• Basically we don’t know precisely neutrino energy in
neutrino experiment —> most challenging for precision
measurement
• Neutrino energy is estimated based on the produced
particles observed in detector, but you might be fooled
and get wrong estimation on energy —> wrong
estimation on underlying parameters
P(νμ → νμ) ∼ 1 − sin2
2θ23 sin2
Δm2
31L
4Eν
Neutrino Energy [GeV]
0 0.5 1 1.5 2 2.5 3
NumberofEvents
0
10
20
30
40
50
60
70
80
Data
Data w/ energy bias (50+/-5 MeV)
Unoscillated Prediction
Neutrino Energy [GeV]
0 0.5 1 1.5 2 2.5 3
Data/Prediction
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
[GeV]νE
0 5 10 15 20 25
p.o.t.]21
/50MeV/1x102
Flux[/cm
1−
10
1
10
2
10
3
10
4
10
5
10
6
10
flux at Super-K with 250 kA operationµνT2K
fluxµν
fluxµν
fluxeν
fluxeν
flux at Super-K with 250 kA operationµνT2K
Here is one example where energy is
added with a Gaussian (50+/- 5MeV)
for the red points. Effect on oscillation
parameter estimation can be followed
(you can do this.)
Oscillation dip changed
77. !77
How do we reconstruct energy?
• Method 1: Kinematics of out-going particle when neutrino interact with nucleon/
nuclei. selecting the interaction you know how to reconstruct energy, example
CCQE. Sometime they call “CCQE-enhanced” (ref. T2K)
• Method 2: Using hadronic information: sum of all charge in the detector (ref.
NOvA)
Comments on methods:
• On method 1: rely on kinematics, and capability of detector to look at final state
particle —> depend on model. For example
• On method 2: neutron basically doesn’t show up in detector; charge sum is
underestimated and you have to correct it with a model.
Dream of neutrino source with well-know energy: yes, we can but not yet
However in the mean time, you have to live with this by learning how your energy
estimation effects on your measured parameters
78. !78
Some selected topics
1) Signal and background
2) Hypothesis test
3) Sensitivity & Parameter estimation
4) Neutrino energy reconstruction
5) Systematics
6) Monte Carlo usage 101
79. !79
Systematic sources
For neutrino experiments, there are basically three sources of errors
• Neutrino source
• Proton beam condition (use monitors but still error)
• pion/Kaon production when proton hits on target: this is the most
dominant error, external data from other experiments are used
• Current uncertainty level of 10%, but can improved
• Neutrino interaction model
• Statistic is challenging
• nuclear effect
• Final state interaction
• Detector systematics
• Secondary interaction
• Detector response
without quoting error, your result is meaningless
80. !80
Systematic sources
0
10
20
30
40
50
Values of variable X
0.6 0.8 1 1.2 1.4
ValuesofvariableY
0.6
0.8
1
1.2
1.4
Histogram
Entries 10000
Mean x 0.9999
Mean y 0.9986
RMS x 0.1001
RMS y 0.1003
0
10
20
30
40
50
Histogram
Entries 10000
Mean x 0.9999
Mean y 0.9986
RMS x 0.1001
RMS y 0.1003
Histogram_py
Entries 10000
Mean 0.9986
RMS 0.1003
0 200 400 600
ValuesofvariableY
0.6
0.8
1
1.2
1.4
Histogram_px
Entries 10000
Mean 0.9999
RMS 0.1001
Values of variable X
0.6 0.8 1 1.2 1.4
0
200
400
600
Histogram_px
Entries 10000
Mean 0.9999
RMS 0.1001
0
10
20
30
40
50
60
70
Values of variable X
0.6 0.8 1 1.2 1.4
ValuesofvariableY
0.6
0.8
1
1.2
1.4
Histogram
Entries 10000
Mean x 0.9999
Mean y 0.9988
RMS x 0.1001
RMS y 0.09973
0
10
20
30
40
50
60
70
Histogram
Entries 10000
Mean x 0.9999
Mean y 0.9988
RMS x 0.1001
RMS y 0.09973
Histogram_py
Entries 10000
Mean 0.9988
RMS 0.09973
0 200 400 600
ValuesofvariableY
0.6
0.8
1
1.2
1.4
Histogram_px
Entries 10000
Mean 0.9999
RMS 0.1001
Values of variable X
0.6 0.8 1 1.2 1.4
0
200
400
600
Histogram_px
Entries 10000
Mean 0.9999
RMS 0.1001
0
10
20
30
40
50
60
Values of variable X
0.6 0.8 1 1.2 1.4
ValuesofvariableY
0.6
0.8
1
1.2
1.4
Histogram
Entries 10000
Mean x 0.9999
Mean y 0.999
RMS x 0.1001
RMS y 0.1006
0
10
20
30
40
50
60
Histogram
Entries 10000
Mean x 0.9999
Mean y 0.999
RMS x 0.1001
RMS y 0.1006
Histogram_py
Entries 10000
Mean 0.999
RMS 0.1006
0 200 400 600
ValuesofvariableY
0.6
0.8
1
1.2
1.4
Histogram_px
Entries 10000
Mean 0.9999
RMS 0.1001
Values of variable X
0.6 0.8 1 1.2 1.4
0
200
400
600
Histogram_px
Entries 10000
Mean 0.9999
RMS 0.1001
No covariance btw 2 parameters
Positive covariance btw 2 parameters
Negative covariance btw 2 parameters
Not just systematic value matter
but covariance btw systematics is also important
81. !81
Systematic sources
No covariance btw 2 parameters
Positive covariance btw 2 parameters
Negative covariance btw 2 parameters
Total systematics effects
0.6 0.8 1 1.2 1.4
0
100
200
300
Histogram_effect
Entries 10000
Mean 0.998
RMS 0.1407
Total systematics effects
0.6 0.8 1 1.2 1.4
Numberofentries
0
50
100
150
200
Histogram_effect
Entries 10000
Mean 1.002
RMS 0.1751
Total systematics effects
0.6 0.8 1 1.2 1.4
Numberofentries
0
100
200
300
400
500
Histogram_effect
Entries 10000
Mean 0.9924
RMS 0.08575
Not just systematic value matter
but covariance btw systematics is also important
Because of this, global combination is non-trivial
83. !83
Some selected topics
1) Signal and background
2) Hypothesis test
3) Sensitivity & Parameter estimation
4) Neutrino energy reconstruction
5) Systematics
6) Monte Carlo usage
84. !84
Monte Carlo usage
Pi calculation method
• Random throw points (dots)
• If inside the circle, you count
• Ratio of counts inside the circle to
all throw points is proportional to
area ratio of 1/4 circle and
corresponding square
https://github.com/cvson/nushortcourse/tree/master/mctoy
85. !85
Monte Carlo usage
• MC integral: to find the mass and
central of mass position
mass : 1.05322 theoretical 1.05221
xcm : -0.00182585
ycm : 0.00821742
zcm : -0.266806 theoretical -0.269071 https://github.com/cvson/nushortcourse/tree/master/mctoy
86. !86
Monte Carlo usage
Neutrino Event generator is one example
• Neutrino follows a distribution —> NO worry, we
just generate a lot of neutrinos with energy follow
that distribution
• There are many possible interactions for a
definitive energy with different cross section —>
NO worry, there a lot of neutrinos generated, each
of them can go different interactions
• Etc….
Keep in mind: randomly go through all possibilities under predefined rule
[GeV]νE
0 5 10 15 20 25
p.o.t.]21
/50MeV/1x102
Flux[/cm
1−
10
1
10
2
10
3
10
4
10
5
10
6
10
flux at Super-K with 250 kA operationµνT2K
fluxµν
fluxµν
fluxeν
fluxeν
flux at Super-K with 250 kA operationµνT2K
[GeV]µν
True E
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
/GeV/atom]2
[cmν/Eσ 0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0.22
36−
10×
O
16
8Cross section on
Total
CCQE
πCC1
CC coherent
CC other
NC
O
16
8Cross section on
details from H. Van
88. !88
See-saw or life balances?
Keep your dream big. But keep in mind, you have to live with small things
Does it somehow link to See-saw phenomenon? Ref. Prof. Kayser’s lecture
Does he look familiar to you?