WEBVTT
00:00:00.000 --> 00:00:18.024
35c3 preroll music
00:00:18.024 --> 00:00:27.449
Herald: Our speaker is Jost Migenda and he
is PhD student in astroparticle physics
00:00:27.449 --> 00:00:34.980
from the University of Sheffield in the UK
and Jost is going to talk about going deep
00:00:34.980 --> 00:00:41.500
underground to watch the stars. Please
give a huge round of applause for Jost
00:00:41.500 --> 00:00:42.530
Migenda.
00:00:42.530 --> 00:00:50.860
applause
00:00:50.860 --> 00:00:56.220
J: Good morning everybody. I'm glad you
managed the first day of congress. Now
00:00:56.220 --> 00:01:02.300
physics rarely makes highlight news. And
if and when it does it is often treated a
00:01:02.300 --> 00:01:09.500
black box, where you pourd in money and
scientist on one end, you wait a while and
00:01:09.500 --> 00:01:16.240
knowledge drops out. So today in this talk
I want to do this a bit differently. I
00:01:16.240 --> 00:01:20.249
want to give you a glimpse behind the
scenes of an experiment, I have been
00:01:20.249 --> 00:01:26.290
working on for over 4 years now. First as
part of my master's thesis and then as a
00:01:26.290 --> 00:01:33.920
PhD student. Now earlier this year we
published a design report which is over
00:01:33.920 --> 00:01:38.030
300 pages long and contains much more
detail about the experiment than you
00:01:38.030 --> 00:01:45.240
probably want to know. So I'll focus on
just some of the highlights in this talk.
00:01:45.240 --> 00:01:48.719
But before we actually talk about the
detector I'll have to introduce you to the
00:01:48.719 --> 00:01:55.389
particles we're looking for. And that
story begins over 100 years ago with
00:01:55.389 --> 00:02:02.279
radioactive beta decay. Now in radioactive
beta decay, you have a nucleus of one
00:02:02.279 --> 00:02:07.850
chemical element that turns into a nucleus
of a different element and emits an
00:02:07.850 --> 00:02:16.060
electron or in modern language we would
say a neutron decays into a proton and an
00:02:16.060 --> 00:02:21.920
electron. Now after that was discovered
there were lots of experiments done to
00:02:21.920 --> 00:02:26.570
measure the energy of the outgoing
electron and experiment after experiment
00:02:26.570 --> 00:02:34.910
found that there was some variance in
energy but was always lower than expected.
00:02:34.910 --> 00:02:39.810
And physicists at the time came up with
all sorts of possible explanations for
00:02:39.810 --> 00:02:45.650
what might be going wrong with these
experiments but they excluded those
00:02:45.650 --> 00:02:51.870
explanations very quickly as well. So
after a while physicists became desperate
00:02:51.870 --> 00:02:56.250
and some pretty well-known physicists
actually thought: "Well, maybe we'll just
00:02:56.250 --> 00:03:04.070
have to give up on conservation of
energy". So in this desperate situation a
00:03:04.070 --> 00:03:09.260
guy called Wolfgang Pauli came up with
what he himself call "a desperate way
00:03:09.260 --> 00:03:14.540
out". So in this letter to a group of his
colleagues which he addressed as "Dear
00:03:14.540 --> 00:03:21.420
radioactive ladies and gentlemen", Pauli
suggested that maybe there's another
00:03:21.420 --> 00:03:28.570
particle created in this beta decay. And
Pauli originally called this particle
00:03:28.570 --> 00:03:33.560
neutron but of course two years later the
particle we nowadays know as Neutron was
00:03:33.560 --> 00:03:40.340
discovered. So Pauli's particle was re-
named neutrino. Now you might be wondering
00:03:40.340 --> 00:03:46.720
well why didn't they observe this particle
already. And the answer is very simple.
00:03:46.720 --> 00:03:52.160
Neutrinos are like ghosts. So what I mean
by there is they can quite literally, you
00:03:52.160 --> 00:03:58.620
know, go through walls or through your
body. And in effect we can do a little
00:03:58.620 --> 00:04:03.740
experiment right now to try and detect
neutrinos. So to help me with this
00:04:03.740 --> 00:04:12.810
experiment, please give me thumb's up.
Everyone? Okay so there's two things
00:04:12.810 --> 00:04:16.750
happening right now. First thing of all
you're giving me a massive confidence
00:04:16.750 --> 00:04:23.560
boost. But, you know, more importantly
somewhere out there the sun is shining and
00:04:23.560 --> 00:04:28.040
it's producing a lot of neutrinos and
nuclear fusion. Now these neutrinos are
00:04:28.040 --> 00:04:35.480
flying to Earth through the roof of this
building and then through your thumbnail.
00:04:35.480 --> 00:04:43.060
And right now as you're listening to me
around 60 billion neutrinos are flying
00:04:43.060 --> 00:04:49.180
through your thumbnail. 60 billion
neutrinos flying through your thumbnail
00:04:49.180 --> 00:04:57.970
every second. How does that feel? Hmm? You
don't feel any of them? Right, so that's
00:04:57.970 --> 00:05:02.740
how ghost-like neutrinos are. And of
course physicists are clever and shortly
00:05:02.740 --> 00:05:08.210
after Pauli had this idea some of them
estimated that how often neutrinos
00:05:08.210 --> 00:05:13.570
interact with normal matter and they found
that there is no practically possible way
00:05:13.570 --> 00:05:18.410
of observing the neutrino. And that
remained true for over 20 years
00:05:18.410 --> 00:05:28.440
afterwards. So now that I have introduced
you to neutrinos. Let's talk about
00:05:28.440 --> 00:05:36.370
building a detector to actually detect
them. And the original motivation for
00:05:36.370 --> 00:05:44.150
building this detector was something a bit
different. I talked about beta decay and
00:05:44.150 --> 00:05:48.580
over the next decades physicists slowly
discovered more particles. They discovered
00:05:48.580 --> 00:05:55.620
that protons and neutrons are made up out
of quarks and in the 1970s theoretical
00:05:55.620 --> 00:06:00.960
physicists came up was are some Grand
Unified Theories basically precursors to
00:06:00.960 --> 00:06:06.800
string theory. And these theories
predicted that the proton should decay as
00:06:06.800 --> 00:06:14.820
well. So of course you know we build
detectors to look for that and a group in
00:06:14.820 --> 00:06:20.610
Japan built a detector near the town of
Kamioka which they called the Kamioka
00:06:20.610 --> 00:06:27.290
Nucleon Decay Experiment or Kamiokande for
short. Now they didn't observe any proton
00:06:27.290 --> 00:06:33.030
decay but shortly after they built it
somebody had a suggestion that if we
00:06:33.030 --> 00:06:37.130
changed just a little bit above their
detector, if we modified just a little, we
00:06:37.130 --> 00:06:44.650
would also be able to detect neutrinos
with that. So they modified the detector
00:06:44.650 --> 00:06:49.290
switched it back on and just a couple of
weeks later they actually observe
00:06:49.290 --> 00:06:57.770
neutrinos from an exploding star just
outside our Milky Way. And that was the
00:06:57.770 --> 00:07:03.540
birth of neutrino astronomy. And for that
the then-leader of the experiment received
00:07:03.540 --> 00:07:11.860
the Nobel Prize in 2002. Now after over a
decade of running physicists were
00:07:11.860 --> 00:07:15.860
basically hitting the limits of what we
could do with a detector of their size. So
00:07:15.860 --> 00:07:22.470
we needed to build a bigger detector and
that one was very creatively named Super-
00:07:22.470 --> 00:07:29.360
Kamiokande. And it's about 20 times
bigger, started running in 1996 and still
00:07:29.360 --> 00:07:39.680
running to this day. Now Super-K did not
discover proton decay but did detect a lot
00:07:39.680 --> 00:07:43.889
of neutrinos and made very fascinating
discoveries. For example, they discovered
00:07:43.889 --> 00:07:49.070
that different types of neutrinos can
change into each other back and forth as
00:07:49.070 --> 00:07:55.259
they travel. That's like you buying a cone
of vanilla ice cream and then as you walk
00:07:55.259 --> 00:08:01.570
out it suddenly turns into chocolate ice
cream. That's really weird. And for their
00:08:01.570 --> 00:08:08.860
discovery just a few years ago they
received the Nobel Prize again. But today
00:08:08.860 --> 00:08:12.580
we are again hitting the limit of you know
what we can learn from a detector that
00:08:12.580 --> 00:08:18.520
size. So of course the next step is to
build an even bigger detector and we're
00:08:18.520 --> 00:08:26.280
calling it Hyper-Kamiokande. By the way
"super" and "hyper" mean exactly the same
00:08:26.280 --> 00:08:34.099
thing. Just one is Latin and one is Greek.
So we're currently getting ready the plans
00:08:34.099 --> 00:08:37.950
to build Hyper-Kamiokande and we will
start construction probably in the spring
00:08:37.950 --> 00:08:47.340
of 2020. Details of the Noble Prize are
still to be determined. Now I said that 60
00:08:47.340 --> 00:08:52.630
billion neutrinos go through your
thumbnail every second. Of course Super-
00:08:52.630 --> 00:08:58.170
Kamiokande which is running right now is
much larger than your thumbnail. So
00:08:58.170 --> 00:09:03.630
there's not just 60 billion but 10000
billion billion neutrinos passing through
00:09:03.630 --> 00:09:13.680
every day and only 10 or 15 of those get
detected. So let's look at what this
00:09:13.680 --> 00:09:20.060
detection process looks like. Now this is
the water inside Super-K. And there's a
00:09:20.060 --> 00:09:24.140
bunch of electrons in there but they'll
show just one. And there's neutrinos
00:09:24.140 --> 00:09:31.940
flying through not just one, not just a
few, but loads of them. And most of them
00:09:31.940 --> 00:09:36.520
go straight through without leaving a
trace. But every once in a while we're
00:09:36.520 --> 00:09:41.450
lucky and one of those neutrinos will
actually hit the electron and give it a
00:09:41.450 --> 00:09:47.190
little kick. And that little kick you know
like billiard balls basically and that
00:09:47.190 --> 00:09:55.170
little kick accelerates the electron to
faster than the speed of light in water.
00:09:55.170 --> 00:09:59.870
Still slower than the speed of light in
vacuum which is the absolute cosmic tempo
00:09:59.870 --> 00:10:07.200
limit. But faster than the speed of light
in water. And then you get basically a
00:10:07.200 --> 00:10:12.920
sonic boom, but with light, which is this
cone of light. And let's just show the
00:10:12.920 --> 00:10:19.680
animation again. So you've got this cone
of light that hits the wall of the
00:10:19.680 --> 00:10:25.660
detector you see a little ring, this ring
of light. Well we've got very sensitive
00:10:25.660 --> 00:10:33.260
photo sensors all over the inside walls to
detect this flash of light and from how
00:10:33.260 --> 00:10:39.760
bright it is we can tell the energy of the
neutrino. And we can also tell, you know
00:10:39.760 --> 00:10:43.550
just like was billiard balls, we can
approximately tell what directions a
00:10:43.550 --> 00:10:51.230
neutrino came from just based on in which
direction it pushed the electron. So
00:10:51.230 --> 00:10:57.190
that's the basic idea how we detect
neutrinos from the sun. Now let's talk
00:10:57.190 --> 00:11:03.420
about what it's actually like to build one
of these detectors. So this is a drawing
00:11:03.420 --> 00:11:11.120
of Hyper-Kamiokande and you can see it's
78 meters high, 74 meters in diameter and
00:11:11.120 --> 00:11:16.459
on the top left there is a truck for
comparison. But maybe a better size
00:11:16.459 --> 00:11:22.200
comparison is to compare this to buildings
which you're familiar with. Like the
00:11:22.200 --> 00:11:28.880
entrance hall which you just came in
through this morning. Or the Statue of
00:11:28.880 --> 00:11:36.899
Liberty and it doesn't quite fit in there.
The arm still looks out. But you could
00:11:36.899 --> 00:11:41.940
drown the Statue of Liberty in this
detector which nowadays is probably some
00:11:41.940 --> 00:11:51.060
sort of political metaphor. So this is the
giant detector. And what's more we're
00:11:51.060 --> 00:11:57.310
building it inside the mountain about 650
meters underground. So that all the rock
00:11:57.310 --> 00:12:02.279
on top will act as a kind of a natural
shield against all sorts of other
00:12:02.279 --> 00:12:06.560
particles, that are raining down on the
atmosphere from outer space so that all
00:12:06.560 --> 00:12:12.640
other particles get stuck and only
Neutrinos can make it through. Now of
00:12:12.640 --> 00:12:18.420
course to build such a huge cavern inside
the mountain - that's something that we
00:12:18.420 --> 00:12:25.540
physicists can't do on our own. So we need
to talk to geologists who look at the rock
00:12:25.540 --> 00:12:31.640
quality and tell us, you know, what's a
good place to build this cavern - where is
00:12:31.640 --> 00:12:37.580
the rock stable enough to do that. And to
figure out the rock quality, they drill
00:12:37.580 --> 00:12:46.860
bore holes in what's actually called a
boring survey. laughter Now, during my
00:12:46.860 --> 00:12:51.510
years working on this experiment, I had to
listen to several hours of talks on these
00:12:51.510 --> 00:12:58.300
geological surveys and I can tell you that
name is quite appropriate. laughter
00:12:58.300 --> 00:13:02.000
though of course, there's a reason I'm not
a geologist, so, you know, take this with
00:13:02.000 --> 00:13:09.420
a grain of salt. But okay, let's say, you
know, we talked to geologists, they told
00:13:09.420 --> 00:13:15.160
us where we can build our detector. The
next step is: We need to actually excavate
00:13:15.160 --> 00:13:21.140
the cavern. And something to keep in mind
is that we are building this somewhere in
00:13:21.140 --> 00:13:26.400
the mountains of Japan, you know, pretty
far away from any major city. So we have
00:13:26.400 --> 00:13:32.690
to think about stuff like lack of local
infrastructure like what's the electricity
00:13:32.690 --> 00:13:40.610
supply like. Do we need to add the power
line. Or what are the local roads like.
00:13:40.610 --> 00:13:44.480
And do they have enough capacity for, you
know, dozens of trucks every day to
00:13:44.480 --> 00:13:50.620
transport away the excavated rock. And, by
the way, where do you store all that
00:13:50.620 --> 00:13:56.710
excavated rock? Because we will be moving
something like half a million cubic meters
00:13:56.710 --> 00:14:01.390
of rock. You can't just store that in your
backyard. You need to find a place where
00:14:01.390 --> 00:14:08.050
all that fits. And, of course, if you've
listened to or watch the Lord of the
00:14:08.050 --> 00:14:15.119
Rings, you'll know it's dangerous to dig
too deeply, to greedily. So we need a
00:14:15.119 --> 00:14:23.211
Balrog early warning system as well. But
okay let's say we've got all those and we
00:14:23.211 --> 00:14:31.300
managed to build a cavern, and now we need
to fill it. And as detector material we
00:14:31.300 --> 00:14:36.971
use water. Both because it's actually
pretty good for detecting neutrinos, but
00:14:36.971 --> 00:14:42.130
also because it's cheap and there's lots
of it. So you can afford to build a
00:14:42.130 --> 00:14:49.690
detector of this size. A detector so big
that that little dot there is a scuba
00:14:49.690 --> 00:14:59.459
diver. But even with water, you hit limits
of, you know, how much you can get. So to
00:14:59.459 --> 00:15:04.680
fill Hyper-Kamiokande you need about as
much water as 5000 people use in a year.
00:15:04.680 --> 00:15:13.860
And that's for drinking, for showering,
for washing their car and so on. Now
00:15:13.860 --> 00:15:17.830
that's easy if you're near a big city. But
we are not, we're somewhere in the
00:15:17.830 --> 00:15:23.560
mountains in Japan where the next biggest
town has far fewer than 5000 people. So
00:15:23.560 --> 00:15:30.980
how do we get enough water to actually
fill our detector? And... we could use
00:15:30.980 --> 00:15:38.060
rivers nearby, we could use springs. We
could wait for for the end of winter and
00:15:38.060 --> 00:15:43.460
for the snow in the mountains to melt and
use that to fill our detector. But if you
00:15:43.460 --> 00:15:48.660
use melting snow to fill the detector, you
can only fill it once a year. So, you
00:15:48.660 --> 00:15:54.950
know, even "where do you get the water" is
is a pretty... pretty important question
00:15:54.950 --> 00:16:03.779
that you need to solve. And then, we're
not just using any water but we actually
00:16:03.779 --> 00:16:09.160
have, we will build our own water
purification system. So that we don't have
00:16:09.160 --> 00:16:15.959
any, you know, traces of radioactivity in
there, any trace of dust and stuff in
00:16:15.959 --> 00:16:23.420
there. And let me let me just tell you
just how pure this water will be. So, this
00:16:23.420 --> 00:16:29.450
is my supervisor, who, when he was a PhD
student, worked in the detector on some
00:16:29.450 --> 00:16:35.610
maintenance work, so he was working on a
boat doing the work, and then at the end
00:16:35.610 --> 00:16:41.149
of his shift he leaned back on the boat,
and just the tip of his long hair fell
00:16:41.149 --> 00:16:48.019
into the water, which he didn't know just
didn't think about too much until at the
00:16:48.019 --> 00:16:54.790
end of his shift. He went home, you know,
went to bed, fell asleep, and then woke up
00:16:54.790 --> 00:17:00.980
in the middle of the night, with his whole
head itching like mad. Now, what had
00:17:00.980 --> 00:17:06.569
happened there? The ultra pure water had
sucked all of the nutrients out the tip of
00:17:06.569 --> 00:17:12.179
his hair and then through osmosis over
time those had sucked the nutrients out of
00:17:12.179 --> 00:17:30.660
the rest of his hair, and then his skin.
So that's how pure that water is. Now, I
00:17:30.660 --> 00:17:35.410
said "all over the inside walls". And here
is, kind of, a photo of the inside of the
00:17:35.410 --> 00:17:41.910
detector. And all these kind of golden
hemispheres, those are what we call
00:17:41.910 --> 00:17:48.850
Photomultiplier tubes, of PMTs for short.
And those are, basically, giant very
00:17:48.850 --> 00:17:54.990
sensitive pixels. And we will have 40,000
of those lining the inside wall of the
00:17:54.990 --> 00:18:02.220
detector. Now, in smaller detectors, you
could just have from each PMT one cable
00:18:02.220 --> 00:18:05.990
leading to the top of the detector and
then have your computers there to analyze
00:18:05.990 --> 00:18:12.799
the signal. With a detector of this size,
you just can't do that because you would
00:18:12.799 --> 00:18:21.559
need 40,000 cables some of which are over
100 meters long. That wouldn't work. So we
00:18:21.559 --> 00:18:26.169
have to put some electronics in the water
to digitize the signal and combine signal
00:18:26.169 --> 00:18:33.940
from multiple PMTs into one, and then use
just a single cable, to bring it up
00:18:33.940 --> 00:18:41.070
to the top where we analyze the signal.
But that means we have to put electronics
00:18:41.070 --> 00:18:47.240
into the water, so that creates a whole
bunch of new problems. For example, we
00:18:47.240 --> 00:18:51.169
need these electronics to be watertight.
And I'm not talking about the level of
00:18:51.169 --> 00:18:54.890
watertight as you'd expect from your
smartwatch where it survives, you know,
00:18:54.890 --> 00:18:59.719
you standing under the shower for five
minutes. I'm talking below 60 metres of
00:18:59.719 --> 00:19:07.900
water for 20+ years. This electronics also
need to be very low power, because we
00:19:07.900 --> 00:19:13.390
can't heet up the water too much.
Otherwise these pixels, the
00:19:13.390 --> 00:19:21.700
Photomultiplier Tubes, would become noisy
and this would kill our signal. And then
00:19:21.700 --> 00:19:26.160
also, because we don't want one defective
cable to kill a whole section of the
00:19:26.160 --> 00:19:30.510
detector. We need to implement some sort
of mesh networking to introduce some
00:19:30.510 --> 00:19:37.660
redundancy. Now each of that by itself is
not... not a heart problem. Each of these
00:19:37.660 --> 00:19:42.870
problems can be solved. It's just a lot of
additional work you suddenly have to do,
00:19:42.870 --> 00:19:51.510
because your detectors is that huge. And
it gets even worse: This is what one of
00:19:51.510 --> 00:19:58.169
these PMTs looks like. It's about 50
centimeters in diameter, and inside that
00:19:58.169 --> 00:20:06.760
glass bulb is a vacuum. So it's under a
lot of pressure. Plus we add 60 metres of
00:20:06.760 --> 00:20:12.370
water on top of it, which adds additional
pressure. So you need to make absolutely
00:20:12.370 --> 00:20:16.830
sure when you're manufacturing those that
you don't have any weak points in the
00:20:16.830 --> 00:20:25.640
glass. And they don't just have to
withstand that water pressure, but there
00:20:25.640 --> 00:20:32.990
will probably be some PMTs that have some
weak points, some air bubbles or something
00:20:32.990 --> 00:20:40.490
in the glass. Some structural weakness.
And the neighboring PMT don't only have to
00:20:40.490 --> 00:20:44.950
survive the normal water pressure. They
also need to survive whether their
00:20:44.950 --> 00:20:51.100
neighboring PMT is imploding and sending
out a pressure wave. And that's not just a
00:20:51.100 --> 00:20:57.950
hypothetical - that actually happened, you
know, 18 years ago. It's 17 years ago in
00:20:57.950 --> 00:21:05.120
Superkamiokande. And that, you know,
within seconds killed more than half of
00:21:05.120 --> 00:21:13.670
the PMTs we had in there and it took years
to restore the detector to full capacity.
00:21:13.670 --> 00:21:19.830
So, lots of problems to solve, and, you
know, one group, one university can't
00:21:19.830 --> 00:21:24.539
solve all these on their own. So we've got
this multinational collaboration with over
00:21:24.539 --> 00:21:29.919
300 people from 17 different countries
marked in green here and across many
00:21:29.919 --> 00:21:36.850
different time zones. So right here right
now, here it's about you know just before
00:21:36.850 --> 00:21:42.430
noon. In Japan people have already had
dinner and they're going to bed soon. In
00:21:42.430 --> 00:21:47.410
the U.S., people haven't even got up yet
in the morning. So good luck finding a
00:21:47.410 --> 00:21:57.540
time for phone meetings which works for
all of these people. So, that's kind of a
00:21:57.540 --> 00:22:01.720
glimpse behind the scenes of what it's
like to work on this detector and to
00:22:01.720 --> 00:22:07.780
actually build it. But now I want to talk
about what we use the detector for. And
00:22:07.780 --> 00:22:13.330
I've got two examples. But of course
there's a whole bunch more that we do, I
00:22:13.330 --> 00:22:23.940
just don't have time to talk about it
today. So, first example: Why does the sun
00:22:23.940 --> 00:22:30.060
shine? That seems like such a simple
question, right? And yet it turns out it's
00:22:30.060 --> 00:22:36.330
really difficult to answer. So, in the
days of the Industrial Revolution, when,
00:22:36.330 --> 00:22:41.480
you know, burning coal and steam power was
all the rage, people thought that, you
00:22:41.480 --> 00:22:48.120
know, maybe it's, you know, a giant ball
of burning coal. But when you when you do
00:22:48.120 --> 00:22:54.780
the math, it turns out that would burn for
a few thousand years maybe. So that
00:22:54.780 --> 00:23:01.000
definitely doesn't work. A bit later,
physicists suggested that maybe the sun is
00:23:01.000 --> 00:23:06.530
just slowly shrinking and shrinking and
it's that gravitational energy which is
00:23:06.530 --> 00:23:13.050
released as light. And that would give you
a life time of a few million years. But
00:23:13.050 --> 00:23:17.070
then you've got, you know pesky geologists
coming along and saying "no no, we've got
00:23:17.070 --> 00:23:21.830
these rock formations or, I don't know,
fossils maybe, for more than a few million
00:23:21.830 --> 00:23:28.310
years old on Earth. So the sun has to live
longer than a few million years. And the
00:23:28.310 --> 00:23:33.750
arguments from the 19th century between
Lord Kelvin and the geologists back then
00:23:33.750 --> 00:23:41.400
are just amazing to read. If you find
those somewhere. But, of course, nowadays
00:23:41.400 --> 00:23:49.250
we know that the answer is nuclear fusion.
And here's, you know, a bunch of reactions
00:23:49.250 --> 00:23:54.790
which lead you the energy generation of
the sun. But now the question is "how can
00:23:54.790 --> 00:24:02.720
we check that? How can we check that this
is actually what's going on?" And the
00:24:02.720 --> 00:24:08.570
answer is neutrinos, because many of these
reactions produce neutrinos, and we can
00:24:08.570 --> 00:24:16.309
detect them. So the one we typically
detect in Super- and later Hyperkamiokande
00:24:16.309 --> 00:24:21.220
is the one on the bottom right here,
called "Bore and eight neutrinos" because
00:24:21.220 --> 00:24:28.929
those have the highest energy. So they are
easiest for us to detect. And the rate of
00:24:28.929 --> 00:24:35.559
various of these processes depends very
much on the temperature. So by measuring
00:24:35.559 --> 00:24:40.320
how many of these neutrinos we see, we can
measure the temperature inside the core of
00:24:40.320 --> 00:24:48.650
the sun. And we have done that, and we
know that it's about 15.5 million Kelvin, plus or
00:24:48.650 --> 00:24:54.740
minus 1 percent. So we know the
temperature in the core of the sun to less
00:24:54.740 --> 00:25:05.640
than 1 percent uncertainty. That's pretty
amazing if you ask me. And, you know, I
00:25:05.640 --> 00:25:10.560
said that we could detect the directions
the neutrinos were coming from. So we can
00:25:10.560 --> 00:25:16.700
actually take a picture of the sun with
neutrinos. Now this is a bit blurry and
00:25:16.700 --> 00:25:21.400
pixelated, not as nice as what you'd get
from a, you know, from the optical
00:25:21.400 --> 00:25:27.810
telescope. But this is still a completely
different way of looking at the sun. And
00:25:27.810 --> 00:25:33.600
what this tells us is that this, you know,
giant glowing orb in the sky. That's not
00:25:33.600 --> 00:25:45.900
some optical illusion but that actually
exists. Okay. So onto our next topic.
00:25:45.900 --> 00:25:54.279
Exploding stars or supernovae which is
what my own research is mostly about. So
00:25:54.279 --> 00:25:59.669
supernovae are these giant explosions
where one single star, like in this
00:25:59.669 --> 00:26:05.760
example here, can shine about as bright as
a whole galaxy consisting of billions of
00:26:05.760 --> 00:26:15.060
stars. And the rule of thumb is this:
However big you think supernovae are
00:26:15.060 --> 00:26:23.830
you're wrong. They're bigger than that.
Randall Munroe as an XKCD fan had this
00:26:23.830 --> 00:26:30.290
excellent example of just how big
supernovae are so he asks which of the
00:26:30.290 --> 00:26:33.980
following would be brighter in terms of
the amount of energy delivered to your
00:26:33.980 --> 00:26:42.441
retina. Option 1: A supernova, seen from
as far away as the sun is from earth or
00:26:42.441 --> 00:26:49.920
Option 2: A hydrogen bomb pressed against
your eyeball. So which of these would be
00:26:49.920 --> 00:27:00.280
brighter. What do you think? You remember
the the rule of thumb we had earlier this
00:27:00.280 --> 00:27:08.769
supernova is bigger than that. In fact
it's a billion times bigger than that. So
00:27:08.769 --> 00:27:15.340
supernovae are some of the biggest bangs
since the original big bang. They also
00:27:15.340 --> 00:27:19.990
leave a neutron star or a black hole which
are really interesting objects to study in
00:27:19.990 --> 00:27:26.240
their own right and the outgoing shockwave
also leads to the creation of lots of new
00:27:26.240 --> 00:27:33.770
stars. But maybe most importantly
supernovae are where many of the chemical
00:27:33.770 --> 00:27:40.779
elements around you come from. So whether
it's things like the oxygen in your lungs
00:27:40.779 --> 00:27:46.860
or the calcium in your bones or the
silicon in your favorite computer chip.
00:27:46.860 --> 00:27:51.539
Life as we know it, and congress as we
know it could not exist without
00:27:51.539 --> 00:28:01.569
supernovae. And yet we don't actually
understand how these explosions happen...
00:28:01.569 --> 00:28:18.890
well I have no idea what's happening....
life as we know it could not exist without
00:28:18.890 --> 00:28:25.560
supernovae and yet we don't actually
understand how these explosions happen.
00:28:25.560 --> 00:28:31.380
And even observing them with telescopes
doesn't really help us because telescopes
00:28:31.380 --> 00:28:37.059
can only ever see the surface of a star.
They can't look inside the core of the
00:28:37.059 --> 00:28:44.529
star where the explosion actually takes
place. So that's why we need neutrinos.
00:28:44.529 --> 00:28:51.020
And we have tens of thousands of
supernovae with optical telescopes, with
00:28:51.020 --> 00:28:58.710
neutrinos we've observed just one. This
one in February of 1987. And we've seen
00:28:58.710 --> 00:29:06.889
two dozen neutrinos which you see here on
the right. That's what we know. So we know
00:29:06.889 --> 00:29:11.770
basically that there were many neutrinos
emitted during the first one second or so
00:29:11.770 --> 00:29:18.980
and then fewer and fewer for the next 10
seconds. We know that neutrinos make up
00:29:18.980 --> 00:29:25.990
most of the energy of the explosion of the
supernova, with the actual energy of the
00:29:25.990 --> 00:29:32.390
explosion and the visible light making
just up a tiny fraction. We know that the
00:29:32.390 --> 00:29:42.450
neutrinos arrive a few hours before the
light. And that's all, that's all we know.
00:29:42.450 --> 00:29:46.830
And still about these two dozen events,
these two dozen neutrinos more than
00:29:46.830 --> 00:29:52.710
sixteen hundred papers are written. That's
more than one paper a week for over 30
00:29:52.710 --> 00:29:59.730
years. So this gives you an idea of just
how important this event was. And how
00:29:59.730 --> 00:30:03.770
creative physicists are. Or I guess, you
could call it desperate.
00:30:03.770 --> 00:30:09.079
Laughter
But you know, I prefer creative. In fact
00:30:09.079 --> 00:30:14.740
this, you know, this one Supernova, we
observed was such big deal, that 30 years
00:30:14.740 --> 00:30:22.690
later February of last year we had a
conference in Tokyo on supernova neutrinos
00:30:22.690 --> 00:30:30.700
and we had a 30th anniversary celebration.
So there we were about 40 or 50 physicists
00:30:30.700 --> 00:30:38.270
looking over the skyline of Tokyo having
dinner, there's the now leader of the
00:30:38.270 --> 00:30:43.680
Super-Kamiokande experiment who was a PhD
student back then when it happened, and in
00:30:43.680 --> 00:30:49.650
his hand he's holding the original data
tape with the events that he himself
00:30:49.650 --> 00:30:56.830
analyzed back then. So there we were, and
at one point that evening we actually
00:30:56.830 --> 00:31:04.419
started to sing Happy Birthday. So you
know how it goes - happy birthday to you,
00:31:04.419 --> 00:31:11.779
happy birthday to you, happy birthday dear
supernova 1987-A. You know it absolutely
00:31:11.779 --> 00:31:21.029
doesn't work, but it was still amazing. So
that's all we know, and then there's what
00:31:21.029 --> 00:31:26.689
we think we know, and most of that comes
from computer simulations of supernovae.
00:31:26.689 --> 00:31:33.440
But the problem is, those are really
really hard. You know it's one of these
00:31:33.440 --> 00:31:38.880
extremely rare situations where all four
fundamental forces: gravity,
00:31:38.880 --> 00:31:44.910
electromagnetism, and the weak, and the
strong nuclear force, all play a role. You
00:31:44.910 --> 00:31:49.399
know normally, in particle physics you
don't have to worry about gravity, and in
00:31:49.399 --> 00:31:53.899
pretty much all other areas of physics you
only have to worry about gravity and
00:31:53.899 --> 00:31:59.620
electromagnetism. Here all four play a
role. You've also got nonlinear
00:31:59.620 --> 00:32:05.170
hydrodynamics of the gas and plasma inside
the star. You've got the matter moving
00:32:05.170 --> 00:32:09.620
relativistically at 10 or 20 percent of
the speed of light and you've got extreme
00:32:09.620 --> 00:32:14.320
pressures and extreme temperatures that
are sometimes beyond what we can produce
00:32:14.320 --> 00:32:23.470
in the laboratory on earth. So that's why
these simulations even in 2018 are still
00:32:23.470 --> 00:32:28.260
limited by the available computing power.
So we need to do a lot of approximations
00:32:28.260 --> 00:32:34.929
to actually get our code to run in a
reasonable time, but that gives you, some,
00:32:34.929 --> 00:32:41.429
that produces some problems, and in fact
the week I started my PhD one of those
00:32:41.429 --> 00:32:46.210
groups doing the supernova simulations
published a paper saying that there is a
00:32:46.210 --> 00:32:52.080
long list of numerical challenges and code
verification issues. Basically we're using
00:32:52.080 --> 00:32:57.890
this approximations and we don't know
exactly how much error they introduce, and
00:32:57.890 --> 00:33:03.390
the results of different groups are still
too far apart, and that's not because
00:33:03.390 --> 00:33:06.670
those people are dummies. Quite the
opposite they're some of the smartest
00:33:06.670 --> 00:33:14.799
people in the world. It's just that the
problem is so damn hard. In fact, in many
00:33:14.799 --> 00:33:20.090
of these simulations the stars don't even
explode on their own, and we don't know
00:33:20.090 --> 00:33:25.070
whether that means that some of these
approximations just introduce numerical
00:33:25.070 --> 00:33:31.390
errors which change the result, or whether
it means that there are some completely
00:33:31.390 --> 00:33:36.340
new physics happening in there which we
don't know about, or whether that is
00:33:36.340 --> 00:33:44.149
actually realistic and some stars you know
in the universe don't explode but just
00:33:44.149 --> 00:33:52.730
implode silently into a blackhole. We
don't know. We just don't know. So take
00:33:52.730 --> 00:34:02.280
any, you know, any results of the
simulations with a grain of salt. That said
00:34:02.280 --> 00:34:09.029
here's our best guess for what happens. So
we start out was a massive star that's at
00:34:09.029 --> 00:34:15.179
least eight times the mass of our sun, and
it starts fusing hydrogen to helium, and
00:34:15.179 --> 00:34:22.659
then on and on into heavier elements until
finally it reaches iron, and at that point
00:34:22.659 --> 00:34:31.440
fusion stops because you can't gain energy
from fusing two iron nuclei. So there the
00:34:31.440 --> 00:34:37.378
iron just accretes in the core of the star
while in the outer layers - here in orange
00:34:37.378 --> 00:34:43.668
- nuclear fusion is still going on, but as
more and more iron accretes and that core
00:34:43.668 --> 00:34:50.839
reaches about one-and-a-half solar masses
it can't hold its own weight anymore and
00:34:50.839 --> 00:34:55.819
it starts to collapse, and inside the core
in nuclear reactions you're starting to
00:34:55.819 --> 00:35:04.770
form neutrinos which I'm showing us ghost
emoji here. Now let's zoom in a bit. The
00:35:04.770 --> 00:35:13.279
core continues to collapse until at the
center it surpasses nuclear density, and
00:35:13.279 --> 00:35:21.669
at that point it's so dense that the
neutrinos are actually trapped in there.
00:35:21.669 --> 00:35:27.469
So even neutrinos which literally can go
through walls can not escape from there
00:35:27.469 --> 00:35:35.779
because matter is so dense, and the
incoming matter basically hits a wall
00:35:35.779 --> 00:35:39.690
because the matter in the center can't be
compressed any further, so just hits the
00:35:39.690 --> 00:35:46.209
wall and bounces back. So from that
collapse you suddenly get an outgoing
00:35:46.209 --> 00:35:51.299
shock wave, and in the wake of that
shockwave suddenly you get a whole burst
00:35:51.299 --> 00:35:58.650
of neutrinos which escape the star
quickly. Now that shockwave moves on and
00:35:58.650 --> 00:36:05.920
slows down and as it slows down the matter
from outer layer still falls in, and in
00:36:05.920 --> 00:36:13.390
this kind of collision region where
neutrinos in the center are still trapped
00:36:13.390 --> 00:36:18.660
in that collision region, neutrinos are,
you know, being produced at a relatively
00:36:18.660 --> 00:36:25.140
steady rate and the shock wave has pretty
much stopped and just wavers back and
00:36:25.140 --> 00:36:32.529
forth, and we see some neutrino emission.
Now after about half a second, maybe a
00:36:32.529 --> 00:36:37.589
second, neutrinos from the center are
slowly starting to escape. You know most
00:36:37.589 --> 00:36:44.079
of them are still trapped but some are
making their way outside and some of those
00:36:44.079 --> 00:36:48.809
actually manage to leave the star while
others interact with matter in this
00:36:48.809 --> 00:36:54.410
shockwave layer and give that matter a
little energy transfer a little push and
00:36:54.410 --> 00:37:02.670
heat it back up. So the shock wave gets
revived and the star actually explodes,
00:37:02.670 --> 00:37:09.520
and all of that took just one second, and
then over the next 10 seconds or so the
00:37:09.520 --> 00:37:16.410
neutrinos remaining at the core slowly
make their way outwards and then travel
00:37:16.410 --> 00:37:22.719
away at the speed of light hopefully to
Earth to our detector. While that shock wave
00:37:22.719 --> 00:37:27.160
moves much slower than the speed of light,
you know, slowly makes its way outwards
00:37:27.160 --> 00:37:32.999
and only a few hours later when that shock
wave reaches the surface of the star do we
00:37:32.999 --> 00:37:41.049
actually see something with telescopes.
So, remember earlier the neutrinos signal
00:37:41.049 --> 00:37:45.759
we saw was something like a bunch of
neutrinos in the first second and then
00:37:45.759 --> 00:37:53.760
fewer and fewer neutrinos for 10 seconds.
Not a lot of detail but what we might see
00:37:53.760 --> 00:37:58.730
is something like this: a brief and
intense burst when the matter hits the
00:37:58.730 --> 00:38:03.680
wall and is thrown back in this first
shock wave, then as the shock wave
00:38:03.680 --> 00:38:08.690
stagnates we might see some wiggles
corresponding to the shock wave, you know,
00:38:08.690 --> 00:38:17.269
sloshing around aimlessly until the shock
wave is revived. The explosion starts and
00:38:17.269 --> 00:38:22.079
then over the next 10 or so seconds we
would see fewer and fewer neutrinos as a
00:38:22.079 --> 00:38:29.010
star cools down and as neutrinos escape.
So if we have good neutrino detectors we
00:38:29.010 --> 00:38:34.099
should be able to watch, you know,
millisecond by millisecond what exactly
00:38:34.099 --> 00:38:43.430
happens inside the star. Now luckily we've
got many more neutrino detectors by now.
00:38:43.430 --> 00:38:46.690
Probably the biggest one is the Super-
Kamiokande in the Mozumi Mine which would
00:38:46.690 --> 00:38:52.359
see about 4000 events from an average
supernova in our Milky Way, and then we've
00:38:52.359 --> 00:38:58.069
got a bunch of other detectors which was
typically you know hundreds of events, and
00:38:58.069 --> 00:39:01.579
some of these detectors are part of
something called the supernova early
00:39:01.579 --> 00:39:07.789
warning system or SNEWS, and snooze is
meant to act as a wake up call to
00:39:07.789 --> 00:39:14.130
astronomers. So when when these detectors
observe neutrinos which are probably from
00:39:14.130 --> 00:39:20.309
a supernova they will send out an alert to
astronomers to get their telescopes ready
00:39:20.309 --> 00:39:27.150
to be able to see that supernova from the
very beginning and then of course just in
00:39:27.150 --> 00:39:31.769
the past few years we've also had
gravitation wave detectors like LIGO in
00:39:31.769 --> 00:39:40.009
the U.S, Virgo in Italy, and in just a few
years we will get another one called KAGRA
00:39:40.009 --> 00:39:45.709
which is located in Japan actually inside
the same mountain as Super-Kamiokande. So
00:39:45.709 --> 00:39:50.260
they're literally next door neighbors, and
then we might get another detector in
00:39:50.260 --> 00:39:56.300
India, maybe one China in the future. So
that's three completely different ways of
00:39:56.300 --> 00:40:05.200
looking at supernovae. So when we observe
a supernova it will be headline news, and
00:40:05.200 --> 00:40:10.209
now you know what's behind those
headlines. So I've introduced you to
00:40:10.209 --> 00:40:16.631
neutrinos, I've told you a bit about what
it's like to work on on such a detector
00:40:16.631 --> 00:40:22.479
and, you know the challenges of building a
detector of this scale and I've showed you
00:40:22.479 --> 00:40:27.519
how with neutrinos we can observe things
but we can't directly observe otherwise
00:40:27.519 --> 00:40:33.619
like the interior of exploding stars, and
with that I want to thank you for your
00:40:33.619 --> 00:40:36.949
attention and please let me know if you
have any questions.
00:40:36.949 --> 00:40:51.044
applause
00:40:51.044 --> 00:40:53.019
Herald: Thank you Jost, it was an amazing
00:40:53.019 --> 00:40:59.530
talk. We have plenty of time for questions
and there are two microphones. Microphone
00:40:59.530 --> 00:41:03.630
one is on the left side of the stage,
microphone two is in the middle so queue
00:41:03.630 --> 00:41:10.839
up and we're going to take some questions.
First question from microphone two.
00:41:10.839 --> 00:41:17.410
Q: Yeah, thank you, I do have a question -
I come from a mining area and I just
00:41:17.410 --> 00:41:24.450
looked up how deep other mines go and I'm
wondering why do you dig into useless rock
00:41:24.450 --> 00:41:30.729
if you can just go to some area where
there are mines that are no longer used,
00:41:30.729 --> 00:41:37.059
my area they go as deep as 1,200 metres I
think. I just looked up and was surprised
00:41:37.059 --> 00:41:41.289
that the deepest mine on Earth is almost
four kilometres in South Africa - an
00:41:41.289 --> 00:41:46.839
active goldmine - so why don't you use
those?
00:41:46.839 --> 00:41:52.059
A: So, part of why we're using that
particular location is because we used it
00:41:52.059 --> 00:42:00.390
for Super-K and Kamiokande before, and the
mountain that Kamiokande was in actually
00:42:00.390 --> 00:42:07.759
is a mine. So we had some previous
infrastructure there, and then there's I
00:42:07.759 --> 00:42:12.470
guess some tradeoff between the benefits
you get from going deeper and deeper and
00:42:12.470 --> 00:42:19.670
the additional cost I think.
Herald: Thank you, we have a question from
00:42:19.670 --> 00:42:24.650
the Internet, that's going to be narrated
by our wonderful Signal Angel.
00:42:24.650 --> 00:42:37.839
Jost: Hello Internet. So the question, I
didn't understand then whole question, but
00:42:37.839 --> 00:42:42.520
something about earthquake. Okay.
Q: Does the earthquake affect the
00:42:42.520 --> 00:42:49.969
detector.
A: Well there's two parts of the answer:
00:42:49.969 --> 00:42:58.660
Part one I'm not a geologist. Part two I
think the earthquakes are mostly centered
00:42:58.660 --> 00:43:06.440
in the cavern on the east coast of Japan
and we're about 200-300 kilometers away
00:43:06.440 --> 00:43:15.009
from there, so the region we're in is
relatively stable, and in fact we've been
00:43:15.009 --> 00:43:22.949
running, since 1983 and we haven't had
problems with earthquakes, and during the
00:43:22.949 --> 00:43:31.380
Fukushima earthquake our detector was
mostly fine but we've actually had, so in
00:43:31.380 --> 00:43:37.599
addition to what I was talking about we're
also producing a beam of neutrinos at an
00:43:37.599 --> 00:43:43.180
accelerator which we shoot at the
detector, and that accelerator is right at
00:43:43.180 --> 00:43:48.510
the east coast. So, the only damage from
the Fukushima earthquake was to that
00:43:48.510 --> 00:43:57.379
accelerator, not to the detector itself. I
hope that answers your question.
00:43:57.379 --> 00:44:03.329
Herald: Next is from microphone two.
Q: Hello, thanks for an interesting talk.
00:44:03.329 --> 00:44:10.090
Do you, or does science, have any theory
if the neutrinos who hit the electron are
00:44:10.090 --> 00:44:16.680
affected themselves from this hit, are
they like directed in another direction or
00:44:16.680 --> 00:44:22.979
lose some sort of energy themselves or
just hit the electron and pass through.
00:44:22.979 --> 00:44:27.999
A: So, conservation of energy and of
momentum still holds, so they would lose
00:44:27.999 --> 00:44:31.289
some energy as they give the electron a
little kick.
00:44:31.289 --> 00:44:34.849
OK. Thank you.
Herald: Thank you, one more question from
00:44:34.849 --> 00:44:38.279
mic two.
Q: Hello, thanks for your talk. My
00:44:38.279 --> 00:44:44.359
question is you said that the only
supernova where we detected some neutrinos
00:44:44.359 --> 00:44:50.509
from is from the 80's. So what is so
special about that supernova that with all
00:44:50.509 --> 00:44:54.670
the new detectors built there was never
another detection?
00:44:54.670 --> 00:45:00.149
A: So, the special thing about that one is
that it was relatively close. So it was in
00:45:00.149 --> 00:45:05.940
the Large Magellanic Cloud about 150,000
light years away which is, you know, on
00:45:05.940 --> 00:45:13.279
cosmic scales our next door neighbor,
whereas other supernovae which we observe
00:45:13.279 --> 00:45:17.059
can be you know millions of light years
away. We can easily see them at that
00:45:17.059 --> 00:45:23.249
distance but we can't detect any
neutrinos, and we expect about between 1
00:45:23.249 --> 00:45:28.999
and 3 supernovae in our Milky Way per
century, so we're in this for the long
00:45:28.999 --> 00:45:32.249
term. Okay.
00:45:32.249 --> 00:45:36.039
Herald: Thank you, microphone two again
please.
00:45:36.039 --> 00:45:43.630
Q: Hi, thanks for your talk. My question
is you said that changing the water once a
00:45:43.630 --> 00:45:50.169
year is not often enough, how often do you
change the water?
00:45:50.169 --> 00:45:56.709
A: How often do we change the water in the
detector? - Yes - So we completely drain
00:45:56.709 --> 00:46:05.789
and refill the detector only for repair
work which, you know, happens every
00:46:05.789 --> 00:46:10.680
depending on what we want to do but
typically every couple of years to, you
00:46:10.680 --> 00:46:17.519
know, 10 plus years and apart from that we
recirculate the water all the time to
00:46:17.519 --> 00:46:22.479
purify it because there will always be
some traces of radioactivity from the
00:46:22.479 --> 00:46:27.599
surrounding rock which make the way in the
water over time.
00:46:27.599 --> 00:46:33.640
Mic 2: Thank you.
Herald: Thank you, and that would be all,
00:46:33.640 --> 00:46:38.635
that was a wonderful start to the
Congress, thank you Jost.
00:46:38.635 --> 00:46:41.405
applause
00:46:41.405 --> 00:46:46.598
35c3 postroll music
00:46:46.598 --> 00:47:03.000
subtitles created by c3subtitles.de
in the year 2019. Join, and help us!