4 ms·

I absolutely love LIGO. YC actually did a great interview with one of the lead physicists on the project where he described some of the technical hardware and s

I absolutely love LIGO. YC actually did a great interview with one of the lead physicists on the project where he described some of the technical hardware and software challenges - https://www.youtube.com/watch?v=1D2j8nTjOZ4 https://www.youtube.com/watch?v=1D2j8nTjOZ4

jackmott42 3y ago
That's the guy who when someone told him they got their first detection was incredulous "I don't have time for this" or something to that affect, because he assumed it was a false positive to have gotten something so quickly, but it was real!
- throwawaymaths 3y ago
  How do we know it was real and not overinterpreting noise again?
  - kmote00 3y ago
    It's my understanding that it was correlated by data from the twin facility on the other side of the country.
    - throwawaymaths 3y ago
      The problem is that we are seeing a lot of correlations. I'd be worried.
      - eaq 3y ago
        One technique that is straightforward to understand is that folks have done "time-slide" analyses for many years - i.e. offsetting one detector's data in time - to understand what the "baseline" rate of correlations is in data streams where there is no possible physical origin. Naturally, much more statistical analysis has been done to ground the claims of "detection"; beyond detailed academic publications, LIGO and others have been producing layperson-accessible science summaries for years/decades that address these and other questions.
        throwawaymaths 3y ago
        > have been producing layperson-accessible science summaries for years/decades that address these and other questions. Citation please. Every layperson accessible summary has said "we use advanced statistics and machine learning" and I haven't found a simple high school statistics accessible explanation yet. Unlike say the higgs boson, I think for this experiment a simple statistical treatment is not an unreasonable request. Please show me and correct me. I would love to be able to believe we have detected gravitational waves.
        ssl232 3y ago
        https://iopscience.iop.org/article/10.1088/1361-6382/ab685e https://iopscience.iop.org/article/10.1088/1361-6382/ab685e Probably not high school statistics accessible but Bayesian statistics and Gaussian noise and GR isn't.
        throwawaymaths 3y ago
        thank you, that is helpful.
  - Gare 3y ago
    https://en.wikipedia.org/wiki/GW170817 https://en.wikipedia.org/wiki/GW170817
    - bsder 3y ago
      That event is amazing. I had no idea how much cross correlation they produced (see the "Scientific Importance" sections). I love the fact that measurements got like 10 orders of magnitude or more better--that's just absolutely absurd.
    - throwawaymaths 3y ago
      You can't take two detectors and just say "they correlate" without presenting the blinded correlation data. If you apply the crazy filters that they do, how many of those signals happen in both detectors by chance? In fact, you can't do a proper statistical analysis without three identical detectors, or at least two pairs of detectors, which we will have once ligo1,2 and Virgo are all online at the same time, and watch how the filtered signals drop in count based on how many detectors you look at. > It is the first GW observation that has been confirmed by non-gravitational means. This is a pretty weak claim. Worryingly, iirc it is the only multi-messenger observation we've made to date, so, again. How many supernovae do we see at any time in any given segment of the sky, and what is the likelihood that it would have happened to be in whatever ~1/20 spatial angle the Gw detectors happen to have resolutions over, and out of the other n candidate neutron-neutron mergers we've "detected" what are the odds that we wouldn't have seen a supernova in the EM by chance in all of the others? If almost every single event came with a mm detection, I'd be convinced, but honestly, IMO it's looking kinda grim right now. We will find out if after we jack up the sensitivity we continue to fail to make multi messenger observations at the same rate as we have so far
      - pharrington 3y ago
        > How many supernovae do we see at any time in any given segment of the sky, and what is the likelihood that it would have happened to be in whatever ~1/20 spatial angle the Gw detectors happen to have resolutions over, and out of the other n candidate neutron-neutron mergers we've "detected" what are the odds that we wouldn't have seen a supernova in the EM by chance in all of the others? The data's all publicly available. You go ahead, do the general relativity, and tell us.

jackmott42 3y ago

That's the guy who when someone told him they got their first detection was incredulous "I don't have time for this" or something to that affect, because he assumed it was a false positive to have gotten something so quickly, but it was real!

throwawaymaths 3y ago
How do we know it was real and not overinterpreting noise again?
- kmote00 3y ago
  It's my understanding that it was correlated by data from the twin facility on the other side of the country.
  - throwawaymaths 3y ago
    The problem is that we are seeing a lot of correlations. I'd be worried.
    - eaq 3y ago
      One technique that is straightforward to understand is that folks have done "time-slide" analyses for many years - i.e. offsetting one detector's data in time - to understand what the "baseline" rate of correlations is in data streams where there is no possible physical origin. Naturally, much more statistical analysis has been done to ground the claims of "detection"; beyond detailed academic publications, LIGO and others have been producing layperson-accessible science summaries for years/decades that address these and other questions.
      - throwawaymaths 3y ago
        > have been producing layperson-accessible science summaries for years/decades that address these and other questions. Citation please. Every layperson accessible summary has said "we use advanced statistics and machine learning" and I haven't found a simple high school statistics accessible explanation yet. Unlike say the higgs boson, I think for this experiment a simple statistical treatment is not an unreasonable request. Please show me and correct me. I would love to be able to believe we have detected gravitational waves.
        ssl232 3y ago
        https://iopscience.iop.org/article/10.1088/1361-6382/ab685e https://iopscience.iop.org/article/10.1088/1361-6382/ab685e Probably not high school statistics accessible but Bayesian statistics and Gaussian noise and GR isn't.
        throwawaymaths 3y ago
        thank you, that is helpful.
- Gare 3y ago
  https://en.wikipedia.org/wiki/GW170817 https://en.wikipedia.org/wiki/GW170817
  - bsder 3y ago
    That event is amazing. I had no idea how much cross correlation they produced (see the "Scientific Importance" sections). I love the fact that measurements got like 10 orders of magnitude or more better--that's just absolutely absurd.
  - throwawaymaths 3y ago
    You can't take two detectors and just say "they correlate" without presenting the blinded correlation data. If you apply the crazy filters that they do, how many of those signals happen in both detectors by chance? In fact, you can't do a proper statistical analysis without three identical detectors, or at least two pairs of detectors, which we will have once ligo1,2 and Virgo are all online at the same time, and watch how the filtered signals drop in count based on how many detectors you look at. > It is the first GW observation that has been confirmed by non-gravitational means. This is a pretty weak claim. Worryingly, iirc it is the only multi-messenger observation we've made to date, so, again. How many supernovae do we see at any time in any given segment of the sky, and what is the likelihood that it would have happened to be in whatever ~1/20 spatial angle the Gw detectors happen to have resolutions over, and out of the other n candidate neutron-neutron mergers we've "detected" what are the odds that we wouldn't have seen a supernova in the EM by chance in all of the others? If almost every single event came with a mm detection, I'd be convinced, but honestly, IMO it's looking kinda grim right now. We will find out if after we jack up the sensitivity we continue to fail to make multi messenger observations at the same rate as we have so far
    - pharrington 3y ago
      > How many supernovae do we see at any time in any given segment of the sky, and what is the likelihood that it would have happened to be in whatever ~1/20 spatial angle the Gw detectors happen to have resolutions over, and out of the other n candidate neutron-neutron mergers we've "detected" what are the odds that we wouldn't have seen a supernova in the EM by chance in all of the others? The data's all publicly available. You go ahead, do the general relativity, and tell us.