> "In the Milky Way Galaxy, cosmic rays (high-energy protons and heavier nuclei) interact with galactic gas and dust to produce both gamma rays and neutrinos.", from the article
The detected neutrinos were not necessarily produced in the vicinity of the source of the cosmic rays as I understand it. Imagine it like a bank of fog lighting up in the night because a car with headlights is moving towards it.
Yes; however, because of the extreme kinetic energy and momentum of the cosmic ray inputs to these collision events, the output neutrinos will be emitted in a tightly focused cone parallel to the path of the original ray. You can be fairly confident that the original source is close to the line-of-sight.
As has been pointed out elsewhere, this is the first image of our galaxy in something other than light (radio, infrared, x rays, gamma rays are all photons).
I don't think I understand any of this. You have to bury a few zillion tons of water in a pitch-black salt mine a mile underneath the surface, and run detectors that see a dim flash of light when one of these hits a water molecule, right?
How is that in any way directional? Or is there a way to compile an image from this without directionality?
The detector is made of thousands of light sensors arranged in a a km-scale 3D grid. When a neutrino interaction causes a flash of light inside this volume, multiple sensors detect the photons with nanosecond-level time resolution. So a 3D map can be made of where the light started and how it moved over time.
For a macroscale analogue, imagine a large 3D grid of microphones all recording sound. If you fired a cannon from inside this grid and looked at the waveforms from all the microphones, you could work out where the cannon was, and also form a pretty good guess of what direction it was pointed.
The "dim flash of light" is called Cherenkov radiation and there's directional information encoded in it. Depending on which (or when a?) particular detector saw that light, the direction of the incoming particle can be calculated.
In addition to the other answers already posted, the neutrino may hit multiple water molecules along its path, or its decay products may hit other molecules themselves, so you get many flashes if you're lucky.
But another category of detector [1] adds additional signal by applying a strong, constant electric field vertically across the entire detection chamber (heavy noble gas, not ice, in this case). Then whatever charged particles are produced drift up to the top of the tank, are annihilated there, and you get a flash that gives you extra good localization in the z direction since you know how long it took them to get there.
The only neutrino detector placed in a salt mine was the IMB detector. That detector was ~10kt of water observed by ~2 thousand photo-detectors, it was located ~600meters underground. The only neutrino detector that's a mile underground was the SNO detector, which was ~1kt of water, observed by ~10 thousand photo-detectors. The SNO detector is still running today as the rechristened SNO+ experiment.
Both the IMB and SNO detectors used electron scattering to observe neutrinos, a neutrino comes in and bumps into an electron orbiting an atom, the electron & neutrino both then go flying off. The electron will usually go off in the same approximate direction that the neutrino was traveling, conservation of energy and momentum requires that. The electron, if energetic enough, emits Cherenkov radiation as it goes. Cherenkov radiation is just the light equivalent to a sonic-boom, it is emitted in a cone centered around the electrons direction of travel. The light from that cone is detected by the photo-detectors. Crucially, both the interaction process (electron-scattering) and the detection process (Cherenkov radiation) will preserve the directionality from the original neutrino (for the most part). The pattern of photo-detectors that gets hit by the Cerenkov light can be analyzed to reconstruct the Cerenkov cone and estimate the original neutrinos direction. Here's an example of an observed Cerenkov ring at the Super Kamiokande detector, although this example is very clear, the Cerenkov rings aren't always so obvious.
https://cerncourier.com/wp-content/uploads/2016/07/CCthe1_06...
The IceCube detector is somewhat different. Their photo-detectors are buried in the Antarctic ice at various depths from ~1-2km and spread out over a roughly 1-cubic km volume, which is ~1Gt of water. I'm not exactly sure how many PMTs in total they have, I reckon its probably around 5-10 thousand. Since their PMT array is so much less dense than the previously mentioned experiments, they can only observe very high energy, very bright, light flashes. So neutrino sources that are low energy, like the Sun, are invisible to them. But, they can see sources that are very high energy, and Ice Cube's extraordinary size lets them observe interactions that are rare/infrequent, such as those from very far away galaxies.
High energy neutrinos will almost always interact via "Deep Inelastic Scattering" (DIS), which is basically the neutrino hitting the protons & neutrons within an atomic nucleus. Since DIS is a scattering process, conservation of energy/momentum requires the scattered particles will preferentially travel in the same direction that the incoming neutrino was traveling in. After that Cerenkov radiation is produced from the scattered protons & neutrons, and that Cerenkov radiation still is emitted in a cone pointing in the direction of travel. So once again, the interaction (DIS) and detection (Cerenkov radiation) preserves directional information. So the pattern of which photo-detectors observe the light can be used to reconstruct that direction, and point back to the neutrinos source (approximately).
Water and ice versions exist, IceCube consists of detectors dug into ice and Super Kamiokande (example) is a big container of ultrapure water (not heavy). Hyper Kamiokande will be even bigger.
Not at the pole - the ice is essentially transparent, they made a 3d grid of detectors by melting deep holes in a grid and dropped strings of detectors that then froze in place
Unfortunately, the actual paper seems to be paywalled, but stories like this often seem to do a poor job of motivating why research like this is interesting.
For both this and all of the articles coming out about gravitational wave detection, these technologies allow us to sense things that can't be seen with light. Gravitational waves are produced by binary black hole systems and mergers, which don't give off any detectable radiation, and neutrinos can be produced by spin down of neutron stars we don't have any other easy way of detecting.
But, these also potentially give us a window into the deep past. The cosmic microwave background represents the furthest back in time we can ever see with light, and it happened during the first formation of neutral hydrogen atoms when the universe first cooled enough to allow that, and thus light could travel without being immediately scattered by free electrons, which was 378,000 years after the Big Bang. Seeing anything before that is impossible.
Neutrinos, however, first decoupled from matter 1 second after the Big Bang. The possibility of being able to detect a cosmic neutrino background from this event would allow us to detect the early universe much earlier than we can with light. And if gravity decoupling from the strong and electroweak forces is ever detectable in a cosmic gravitational wave background, that would have happened even earlier, and represents the earliest possible viewing of the universe by any means whatsoever.
I'm not a cosmologist and have no idea what usable data would ever come from being able to see these things, but keep in mind at least one reason we've had so much difficulty developing a grand unified theory and theory of everything even after conquering electroweak, is the inability to produce the enormous energies required to recouple the forces in a lab. A particle accelerator that could do it for strong force recoupling to electroweak would have to be the size of Pluto's orbit. But there is at least one event in nature where the necessary energy existed, which is the early universe. We've just never been able to see it.
Basically, with our current understanding, the only thing that makes the big bang and black holes seem connected is that they're both singularities.
But, there's no explanatory value associated with the idea of the universe being the inside of a black hole, and thus there's no way to attempt to test it.
>“What’s intriguing is that, unlike the case for light of any wavelength, in neutrinos, the universe outshines the nearby sources in our own galaxy,” says Francis Halzen, a professor of physics at the University of Wisconsin–Madison and principal investigator of IceCube.
As you'd expect, the resolution from a 86x60 neutrino detector array is not great.