Why UTokyo?
Home > Why UTokyo? > In Depth > Transcript of Professor Kajita's Nobel Lecture
In Depth

Transcript of Professor Kajita's Nobel Lecture

Discovery of Atmospheric Neutrino Oscillations

Professor Kajita delivered his Nobel Lecture in Stockholm on December 8th, 2015. Here is the transcript of his nearly 30-minute lecture in English, edited for clarity and interspersed with a selection of the actual slides used in his presentation. On the occasion of this prestigious lecture, which only Nobel Laureates are entitled to give, what did Professor Kajita discuss? As you read through his lecture, let your mind wander in wonder over neutrinos, the tiny particles that connect the infinitesimally small with the extremely large vastness of space.


The Aula Magna at Stockholm University, where Professor Kajita delievered his Nobel Lecture
Good morning. First of all, I'd like to thank you for the very kind introduction1. And also I want to mention that it is really the greatest honor for me to give this lecture today. Today I am going to talk about atmospheric neutrinos.
The outline of this talk is shown here. First I will give an introduction of the Kamiokande experiment that was the starting point of my research. Then, I want to discuss the atmospheric neutrino deficit. Then, the discovery of neutrino oscillations and recent results and the future. Then, I’ll summarize and finally I want to mention the acknowledgements.

Introduction: Kamiokande - the starting point

Now for the introduction.

In the late 1970s, new theories2 that unify strong, weak and electromagnetic forces were proposed. These theories predicted that protons and neutrons—namely, nucleons—should decay with a lifetime of approximately 1028 to 1032 years. This is long enough to think about the real effect. However, it is short enough to observe proton decay. Therefore, several proton decay experiments began in the early 1980s, and one of them was the Kamiokande experiment3.

This is a photo of the Kamiokande experiment. It is a three-kiloton water Cherenkov detector and the fiducial mass for the masses of neutrino events, or proton decays, was about one kiloton. And I'd like to tell you the basics of this experiment. If relativistic charged particles propagate in the water, they will emit Cherenkov light4. And these photons' Cherenkov light is detected by photon detectors that are placed on the walls of the detector. You can see these dots are actually photo detectors.
I want to mention that initially Kamiokande was a rather small experiment. This was the, I would say, construction team in the spring of 1983. Here, you can see Professor Koshiba, who was the Nobel Laureate in Physics in 2002. And you can also see Professor Totsuka, Professor Kifune and some students maybe two to three meters behind Professor Koshiba. (laughs) One of them was me.

Atmospheric neutrino deficit

We had an introduction of neutrinos earlier, so I can keep my explanation short here. However, I would still like to repeat some of the main points.
Neutrinos are fundamental particles like electrons and quarks. They have no electric charge, and they have three types, or three flavors: electron-neutrinos (νe), muon-neutrinos (νμ) and tau-neutrinos (ντ). They are produced in various places, such as the Earth’s atmosphere, or the center of the Sun, and they can easily penetrate through the Earth, or maybe even the Sun. So, if neutrinos are produced here, they can easily propagate through the Earth, unfortunately also through the detector, and go into space. However, they can interact with matter very rarely. If interactions occur, a muon-neutrino produces a muon, and an electron-neutrino produces an electron. And also I want to mention that in the very successful Standard Model of particle physics, neutrinos are assumed to have no mass. However, physicists have been asking if neutrinos really have no mass.
Now I want to move on to the atmospheric neutrino deficit. I want to explain a little bit about how cosmic ray particles entering the atmosphere interact with the air nucleus. These interactions typically produce pions and these pions decay into muons, and then into electrons. In this decay chain, two muon-neutrinos and one electron-neutrino are produced. These are observed in underground detectors.
Now, I want to mention a little bit about my work in the early days. I got my PhD in March 1986 based on the search for proton decay. Of course, I did not find any proton decay. But, anyway, I felt that the analysis software was not good enough to select the signal that is proton decay from the background that is atmospheric neutrino interactions most efficiently. Therefore, as soon as I submitted my thesis, I began to work to improve the software. One of the programs was an analysis software for identifying the particle type for multi Cherenkov ring events5. Namely, I wanted to know, if, say, each Cherenkov ring in a multi-ring event is produced by an electron or a muon. For example, this is an event observed in Kamiokande. You can see three Cherenkov rings here. I wanted to know if they were produced by a muon or an electron.

Anyway, the new software was ready and was applied to single Cherenkov ring events, which were the easiest events to analyze. It's natural to start from the easiest thing.
Then, the neutrino flavors were studied for the atmospheric neutrino events. These are the typical events observed in Kamiokande. This is a typical event pattern for electron-neutrino interaction, and this is a typical event pattern for muon-neutrino interaction. We immediately found that the result was strange. The number of muon-neutrino events was much fewer than expected. At first, I thought that I had made some serious mistake. In order to know where I made a mistake, I decided to scan the events. Immediately, I realized that the analysis software was right. However, I was not optimistic yet. I thought that it was very likely that there were some mistakes somewhere in the simulation, data reduction, and/or event reconstruction. So, we, mostly Professor Masato Takita and myself, started various studies to find the mistakes in late 1986.
After more than one year of studies, we concluded that the muon-neutrino deficit could not be due to any major problem in the data analysis nor the simulation. Therefore, Kamiokande decided to publish this result. Basically, in this paper, we reported these numbers. The number of electron-neutrino events was 93, while the predicted number was 88.5. So, basically the numbers agreed. However, for the number of muon-neutrino events, the data was 85, while the prediction was 144. It's clear that there was a deficit in muon-neutrino events. In the paper we concluded: “We are unable to explain the data as the result of systematic detector effects or uncertainties in the atmospheric neutrino fluxes. Some as-yet-unaccounted-for physics such as neutrino oscillations might explain the data.” At this stage we still said ''might.''
However, I was very much excited. I want to mention my personal memory. I was mostly excited with the possibility of neutrino oscillations with a large mixing angle. Namely, muon-neutrinos seemed to oscillate maximally to the other neutrino type, which was not expected. This gave me the strong motivation to continue the study.

Discovery of neutrino oscillations

Now, since I mentioned neutrino oscillations, I want to explain a little about them. If neutrinos have mass, they will change their flavor (type) from one flavor (type) to another. For example, oscillations could occur between muon-neutrinos and tau-neutrinos. If muon-neutrinos are produced here, and then, if they travel long distances, they disappear and reappear. Looking at this diagram, as a muon-neutrino disappears, a tau-neutrino appears. This is neutrino oscillation. One important thing is if the mass of a neutrino is smaller, the oscillation length (L/E) gets longer. Now, I also want to mention the people who theoretically predicted neutrino oscillations: Professor Shoichi Sakata, Professor Jiro Maki and Professor Masami Nakagawa of Nagoya University, and Italian physicist Bruno Pontecorvo. They contributed greatly to this discovery.
So far, I have discussed results from the Kamiokande experiments, but there was another detector experiment called IMB6. The IMB experiment, which also used a large water Cherenkov detector, similarly reported a deficit of muon-neutrino events. That was encouraging.
However, we thought that a simple deficit of muon-neutrino events would not really be enough to conclude that the cause was neutrino oscillations. We wanted more positive evidence for oscillations.
Actually, the physics is simple. If neutrinos are produced in the atmosphere over the detector, then travel distances are so short that there is no time for the neutrinos to oscillate. However, if neutrinos are produced on the other side of the Earth, they need to travel long distances before reaching the detector. So, they may oscillate.

Therefore, we concluded that we should observe a deficit of upward going muon-neutrinos. In fact, we tried to observe such an event with Kamiokande, but it was clear that Kamiokande was not big enough to give us conclusive data. We needed a much larger detector. We needed Super-Kamiokande7.
Now, I want to move on to the neutrino oscillation discovery. As you know, the discovery of neutrino oscillations was made by Super-Kamiokande. Super-Kamiokande is a 50,000-ton water Cherenkov detector with a fiducial volume of 22,500 tons. It is more than 20 times larger than the fiducial mass of Kamiokande. This is a big experiment and a big international collaboration. In fact, over 120 people have collaborated with us on this experiment.
This photo shows the beginning of the Super-Kamiokande collaboration between Japan and the United States. Here, you can see Professor Tostuka who was the spokesperson of Super-Kamiokande. And you can see other key people. This photo was taken at the Institute for Cosmic Ray Research probably in 1991 or 1992.
Now, this is a photo of the Super-Kamiokande detector. It was taken in January 1996. You can see water is filled to this level and you can realize how big the detector is. You can also see three people on the rubber raft.
Anyway, the Super-Kamiokande detector was constructed and started taking data in 1996. Since then, Super-Kamiokande has been observing a lot of events. The figure on the left shows a typical single Cherenkov ring event observed in Super-Kamiokande. The right figure shows a multiple Cherenkov ring event, and here is a partially-contained event and an upward-going muon event. All of these events were used in the analysis. We needed the cooperation of many people, particularly young people. That was essential to the discovery.
Anyway, after two years from the beginning of the Super-Kamiokande experiment, we concluded that the data gave the evidence for neutrino oscillations. This is a slide used in a 1998 international conference on neutrinos. Neutrino 1998. It was the most important part of that presentation. The "e" graph on top shows electron-neutrino data, while the "μ" graph below shows muon-neutrino data. The total number of muon-neutrinos is significantly under the predicted amount. We can clearly see the deficit. From this result, together with the other data, Super-Kamiokande concluded that the observed zenith angle dependent deficit and the other supporting data gave evidence for neutrino oscillations.
At that time, there were two other atmospheric neutrino experiments: MACRO8 in Italy and Soudan-29 in the United States. These experiments observed atmospheric neutrinos and confirmed neutrino oscillations. So, neutrino oscillations were quickly accepted by the neutrino community.

Recent results and the future

I want to talk about the recent results and the future. It was very nice to see that approximately half of the long-traveling muon-neutrinos disappeared. However, we wanted to confirm that this phenomenon really meant that the neutrinos were “oscillating.”
Let's compare the 1998 data with the data we collected in 2015. During 1998, 531 events were observed, while in 2015, the number of events observed was 5485.
This data tells us two things. First, the heaviest neutrino mass is approximately 10,000,000 times smaller than the mass of an electron, which is the lightest particle next to a neutrino. So you can realize that the mass of a neutrino is extremely, extremely tiny. In addition, we found that muon-neutrinos oscillate maximally to tau-neutrinos, which was really surprising. We would like to understand why this is so.
Now I want to mention a little bit about the overall history, or development, of neutrino oscillation experiments.
As I said, the muon-neutrino deficit began to be observed around 1990, and neutrino oscillations were discovered in the 1990s. There have been long-baseline neutrino oscillation experiments10 until the 2000s, and neutrino oscillations among the three neutrino flavors were confirmed in the 2010s. However, the story of neutrinos has not been completed yet. The neutrino community is trying to pursue further experiments, such as the Hyper-Kamiokande11, which we hope will be operational in the 2020s. Our understanding on neutrino oscillations has been improving tremendously, but we still have much to learn.


An unexpected muon-neutrino deficit in the atmospheric neutrino flux was observed in Kamiokande (1988).
Subsequently, in 1998, Super-Kamiokande discovered neutrino oscillations, which show that neutrinos have mass.
I feel that I have been extremely lucky because I have been involved in the excitement of this discovery from the beginning.
The discovery of non-zero neutrino masses opened a window to study physics beyond the Standard Model of elementary particle physics, and will probably lead us to the Grand Unified Theory of elementary particle interactions.
There are still many things to be observed about neutrinos. Further studies of neutrinos might give us fundamental information for understanding nature, such as the origin of the matter in the Universe.


I would like to thank those who collaborated with us for the Kamiokande and Super-Kamiokande experiments. In particular, I would like to thank Masatoshi Koshiba and Yoji Totsuka for their continuous support and encouragement of my research throughout my career. Ed Kearns worked with me on the analyses of atmospheric neutrinos in Super-Kamiokande for many years. Masato Takita and Kenji Kaneyuki worked with me on the Kamiokande analysis. Yoji Totsuka, Yoichiro Suzuki and Masayuki Nakahata have been leading the Super-Kamiokande experiment. Hank Sobel and Jim Stone have been leading the Super-Kamiokande efforts in the US. Kenzo Nakamura and Atsuto Suzuki played very important roles in the early stages of Super-Kamiokande. And, most importantly, the hard work from young collaborators with Super-Kamiokande was essential for the discovery.
Also, I would like to thank Morihiro Honda for the neutrino flux calculation. Finally, Super-Kamiokande acknowledges the Japanese Ministry of Education, Culture, Sports, Science and Technology, the United States Department of Energy, and Kamioka Mining and Smelting Company.
Thank you very much for your attention.


After giving his Nobel Lecture, Professor Kajita shares some friendly words with Professor Arthur McDonald, the joint winner of the 2015 Nobel Prize in Physics
1: Preceding Professor Kajita's lecture, Professor Anne L'Huillier (of Lund University in Sweden), the chair of the Nobel Committee for Physics, gave a brief explanation of Professor Kajita's background and the reason for him receiving the Nobel Prize. Her speech focused on how Professor Kajita's fantastic achievements have captivated the minds of many people.
2: The natural world is thought to be comprised of four fundamental forces (electromagnetic, weak, strong, and gravitational). The Grand Unified Theory seeks to explain the three forces other than gravity in an integrated way, but this theory is as of yet incomplete.
3: The name "Kamiokande" comes from the phrase "KAMIOKA Nucleon Decay Experiment." "Kamioka" is from the area called "KAMIOKA," and "-nde" is the abbreviation of "Nucleon Decay Experiment." As the name suggests, the original purpose of this experiment was to detect proton decay (protons and neutrons are both types of nucleons).
4: The Cherenkov light phenomena is named after its discoverer, Soviet physicist Pavel Cherenkov. Cherenkov received the Nobel Prize in Physics in 1958 for his discovery.
5: A Cherenkov ring forms when Cherenkov light radiates in a conical shape along the direction in which charged particles move, projecting an annular pattern on a wall.
6: A joint project by the University of California, Irvine, the University of Michigan, and Brookhaven National Laboratory. The detector was located underground in a salt mine in Fairport Harbor, Ohio, USA. IMB stands for Irvine-Michigan-Brookhaven.
7: The name "Super-Kamiokande" is derived from "Super-Kamioka Neutrino Detection Experiment." The "-nde" in Super-Kamiokande thus has a different meaning from the "-nde" in Kamiokande.
8: An experiment that took place in Gran Sasso National Laboratory in Italy. Its name is an acronym for "Monopole, Astrophysics and Cosmic Ray Observatory."
9: A detector located within the Soudan Mine in Minnesota, USA.
10: Professor Kajita is referring to an experiment called T2K (Tokai to Kamioka). In this experiment, neutrinos created with an accelerator were sent off from Tokai-mura in Ibaraki, then observed by Super-Kamiokande 295 kilometers away in Kamioka, Gifu.
11: Plans for a one-million ton Cherenkov light detector—20 times the size of Super-Kamiokande—that will bring neutrino research to the next level. The goals for the experiments conducted at this facility will be to verify a unified particle theory and unravel the evolutionary history of the universe.

You can view a video of Professor Kajita's Nobel Lecture and the slides he used in his presentation here:

* This article is a translation of an article that was originally printed in Tansei 32 (Japanese language only).
** This is Part 2 of a three-part series on the University of Tokyo and the Nobel Prize (Read Part 1 and Part 3 here.).

Inquiries about the content of this page: Public Relations GroupSend inquiry

Access Map
Kashiwa Campus
Hongo Campus
Komaba Campus