5 ms·

To do that you would need the sensor to provide either the distance or direction of the “signal,” right? Does this actually do that? Given how unreactive neutri

To do that you would need the sensor to provide either the distance or direction of the “signal,” right? Does this actually do that? Given how unreactive neutrinos are I can’t imagine there being much decay or defraction or distortion over distance. Can this experimental sensor produce a direction that the neutrinos are coming from?

dragonwriter 3y ago
> To do that you would need the sensor to provide either the distance or direction of the “signal,” right? It was specifically called out that one of the things this relied on is that their detector method provides an estimated direction. OTOH, when you have to integrate a half year of data to detect any signal, triangulation of non-stationary sources, especially if there is more than one potential source, is going to be…beyond difficult.
- jjk166 3y ago
  You can integrate the same amount of data in less time by increasing the size of the detector. With a detector 182.5 times greater in volume (which is only 6 times the radius), you only need a day's worth of data - still probably not useful for targeting but now you know the search area to look within. You can also build a network of multiple detectors, which in addition to increasing the amount of data you collect also makes it easier to differentiate signals. For a nation state this would be chump change.
  - dragonwriter 3y ago
    > With a detector 182.5 times greater in volume (which is only 6 times the radius), you only need a day’s worth of data Oh, okay, so you just want to scale this up ~200 times, and build multiple of them (deep underground, because the water is shielding for the ambient radiation from the rock, which it has a couple km of shielding it from external radiation): “The heart of the SNO+ detector is a 12m diameter acrylic sphere filled with 780 tonnes of liquid scintillator that is surrounded by 7000 tonnes of ultrapure water which serves as shielding from external ambient radioactivity. This volume is monitored by nearly 10,000 photomultiplier tubes (PMTs) that are very sensitive light detectors.” [0] > You can also build a network of multiple detectors Well, yes, you can’t triangulate with only one detector, you have to do that at a minimum. That’s not an “also you can” option, its a minimum requirement because you need multiple simultaneous (or close enough) signals from the same source to triangulate. [0] https://snoplus.phy.queensu.ca/about.html https://snoplus.phy.queensu.ca/about.html
    - snophysics 3y ago
      I worked on the SNO+ experiment for grad school and actually got to go in a boat in the ultra pure water surrounding the acrylic vessel to search for leaks. It is an impressive piece of engineering to construct such a detector requiring a lot of money and effort. They actually had to bring down the acrylic vessel in pieces because the whole thing wouldn't fit down the mine shaft elevator. Then they glued them all together underground (this is no small feat since the acrylic vessel has to be optically clear for the photons to travel through to the photomultiplier tubes).
    - kadoban 3y ago
      You say that like it's ridiculous, but is it? Seems pretty reasonable if a large nation wants to know where nukes are, or if a smaller one _really_ wants to know. Spies and satellite monitoring might be cheaper, but less reliable.
    - Dylan16807 3y ago
      > Oh, okay, so you just want to scale this up ~200 times, and build multiple of them The equipment should only need to scale with the square of the radius, right? But sure, why not? The budget is orders of magnitude bigger. > Well, yes, you can’t triangulate with only one detector, you have to do that at a minimum. That’s not an “also you can” option, its a minimum requirement because you need multiple simultaneous (or close enough) signals from the same source to triangulate. You don't need to triangulate. It's going to be in a thin shell near the earth's surface. So using more is optional.
    - jjk166 3y ago
      Yeah, building a 70m diameter device is obviously going to be more expensive than building a 12m diameter device, but assuming the budget is available there's no technical obstacle to doing so. SNO+ cost about $100 million to build which is a decent bit for a science project but nothing for a military procurement project. Assuming cost scales with volume (which I am skeptical of), 200 times more expensive would be $20 Billion - about 1.5x the cost of a ford class carrier. Is knowing the location of nuclear reactors to that level of precision worth that price tag? I don't know but it's within the range that someone could make a plausible argument for it. If we assume construction takes 10 years, as it did for SNO+, then the project would represent 0.2% of the US's annual budget. And these are directional detectors, they don't need to triangulate. You can most definitely identify the location of a reactor with one detector, which is exactly what this study has done. If you were to go with multiple detectors, they can all get proportionally smaller and thus cheaper, to say nothing of potential economies of scale.
jjk166 3y ago
Yes, when neutrinos do react (rare though that is) it produces electrons that move in the same direction as the neutrino, and which move at extremely high speed, faster than the speed of light in the medium in fact, which produces cherenkov radiation. This radiation is what the sensors actually detect. Essentially they are giant cameras looking for brief streaks of light. By seeing how long and in what direction the streak is, you can determine both the direction and energy of the incoming neutrino.
- snophysics 3y ago
  In this case they are detecting inverse beta decay events. The antineutrino interacts with a proton producing a positron and a neutron. The positron annihilates with an electron producing two gamma rays and the neutron then captures on a proton and releases a 2.22 MeV gamma ray. I think the only realistic way to determine the direction is by looking at the direction of the neutron by comparing the position of the positron annihilation with the neutron capture, but typically the neutron scatters so much in these detectors that it is not possible to do on an event by event basis (although see [0] for a possible detector that could). What you're talking about is elastic scattering and I think this may be possible (see [1]), but I don't know of anyone who is able to do this at these low energies. As a side note, the WATCHMAN collaboration has been studying the ability to monitor nuclear reactors using these sorts of detectors for a while. [0] https://arxiv.org/pdf/1410.8530.pdf https://arxiv.org/pdf/1410.8530.pdf [1] https://www.osti.gov/servlets/purl/1332127 https://www.osti.gov/servlets/purl/1332127
  - jjk166 3y ago
    No, I'm not referring to elastic scattering. It's the positron that they are tracking, and it moves in approximately the same direction as the neutrino to conserve linear momentum. It's exactly the same process, the only difference is anti-neutrinos produce positrons, neutrinos produce electrons.
    - snophysics 3y ago
      I don't think that's true. The energies of the neutrinos they are looking at are relative low (less than 15 MeV), and according to https://journals.aps.org/prd/pdf/10.1103/PhysRevD.60.053003 https://journals.aps.org/prd/pdf/10.1103/PhysRevD.60.053003 the positron is isotropic around 15 MeV and is only forward pointing at much higher energies.

dragonwriter 3y ago

> To do that you would need the sensor to provide either the distance or direction of the “signal,” right? It was specifically called out that one of the things this relied on is that their detector method provides an estimated direction. OTOH, when you have to integrate a half year of data to detect any signal, triangulation of non-stationary sources, especially if there is more than one potential source, is going to be…beyond difficult.

jjk166 3y ago
You can integrate the same amount of data in less time by increasing the size of the detector. With a detector 182.5 times greater in volume (which is only 6 times the radius), you only need a day's worth of data - still probably not useful for targeting but now you know the search area to look within. You can also build a network of multiple detectors, which in addition to increasing the amount of data you collect also makes it easier to differentiate signals. For a nation state this would be chump change.
- dragonwriter 3y ago
  > With a detector 182.5 times greater in volume (which is only 6 times the radius), you only need a day’s worth of data Oh, okay, so you just want to scale this up ~200 times, and build multiple of them (deep underground, because the water is shielding for the ambient radiation from the rock, which it has a couple km of shielding it from external radiation): “The heart of the SNO+ detector is a 12m diameter acrylic sphere filled with 780 tonnes of liquid scintillator that is surrounded by 7000 tonnes of ultrapure water which serves as shielding from external ambient radioactivity. This volume is monitored by nearly 10,000 photomultiplier tubes (PMTs) that are very sensitive light detectors.” [0] > You can also build a network of multiple detectors Well, yes, you can’t triangulate with only one detector, you have to do that at a minimum. That’s not an “also you can” option, its a minimum requirement because you need multiple simultaneous (or close enough) signals from the same source to triangulate. [0] https://snoplus.phy.queensu.ca/about.html https://snoplus.phy.queensu.ca/about.html
  - snophysics 3y ago
    I worked on the SNO+ experiment for grad school and actually got to go in a boat in the ultra pure water surrounding the acrylic vessel to search for leaks. It is an impressive piece of engineering to construct such a detector requiring a lot of money and effort. They actually had to bring down the acrylic vessel in pieces because the whole thing wouldn't fit down the mine shaft elevator. Then they glued them all together underground (this is no small feat since the acrylic vessel has to be optically clear for the photons to travel through to the photomultiplier tubes).
  - kadoban 3y ago
    You say that like it's ridiculous, but is it? Seems pretty reasonable if a large nation wants to know where nukes are, or if a smaller one _really_ wants to know. Spies and satellite monitoring might be cheaper, but less reliable.
  - Dylan16807 3y ago
    > Oh, okay, so you just want to scale this up ~200 times, and build multiple of them The equipment should only need to scale with the square of the radius, right? But sure, why not? The budget is orders of magnitude bigger. > Well, yes, you can’t triangulate with only one detector, you have to do that at a minimum. That’s not an “also you can” option, its a minimum requirement because you need multiple simultaneous (or close enough) signals from the same source to triangulate. You don't need to triangulate. It's going to be in a thin shell near the earth's surface. So using more is optional.
  - jjk166 3y ago
    Yeah, building a 70m diameter device is obviously going to be more expensive than building a 12m diameter device, but assuming the budget is available there's no technical obstacle to doing so. SNO+ cost about $100 million to build which is a decent bit for a science project but nothing for a military procurement project. Assuming cost scales with volume (which I am skeptical of), 200 times more expensive would be $20 Billion - about 1.5x the cost of a ford class carrier. Is knowing the location of nuclear reactors to that level of precision worth that price tag? I don't know but it's within the range that someone could make a plausible argument for it. If we assume construction takes 10 years, as it did for SNO+, then the project would represent 0.2% of the US's annual budget. And these are directional detectors, they don't need to triangulate. You can most definitely identify the location of a reactor with one detector, which is exactly what this study has done. If you were to go with multiple detectors, they can all get proportionally smaller and thus cheaper, to say nothing of potential economies of scale.

jjk166 3y ago

Yes, when neutrinos do react (rare though that is) it produces electrons that move in the same direction as the neutrino, and which move at extremely high speed, faster than the speed of light in the medium in fact, which produces cherenkov radiation. This radiation is what the sensors actually detect. Essentially they are giant cameras looking for brief streaks of light. By seeing how long and in what direction the streak is, you can determine both the direction and energy of the incoming neutrino.

snophysics 3y ago
In this case they are detecting inverse beta decay events. The antineutrino interacts with a proton producing a positron and a neutron. The positron annihilates with an electron producing two gamma rays and the neutron then captures on a proton and releases a 2.22 MeV gamma ray. I think the only realistic way to determine the direction is by looking at the direction of the neutron by comparing the position of the positron annihilation with the neutron capture, but typically the neutron scatters so much in these detectors that it is not possible to do on an event by event basis (although see [0] for a possible detector that could). What you're talking about is elastic scattering and I think this may be possible (see [1]), but I don't know of anyone who is able to do this at these low energies. As a side note, the WATCHMAN collaboration has been studying the ability to monitor nuclear reactors using these sorts of detectors for a while. [0] https://arxiv.org/pdf/1410.8530.pdf https://arxiv.org/pdf/1410.8530.pdf [1] https://www.osti.gov/servlets/purl/1332127 https://www.osti.gov/servlets/purl/1332127
- jjk166 3y ago
  No, I'm not referring to elastic scattering. It's the positron that they are tracking, and it moves in approximately the same direction as the neutrino to conserve linear momentum. It's exactly the same process, the only difference is anti-neutrinos produce positrons, neutrinos produce electrons.
  - snophysics 3y ago
    I don't think that's true. The energies of the neutrinos they are looking at are relative low (less than 15 MeV), and according to https://journals.aps.org/prd/pdf/10.1103/PhysRevD.60.053003 https://journals.aps.org/prd/pdf/10.1103/PhysRevD.60.053003 the positron is isotropic around 15 MeV and is only forward pointing at much higher energies.