Can I assume that the delay between Hanford and Livingston is < 12 ms? And between Hanford and Virgo < 39 ms?
\nFor 57c3183f4a, I get 13 ms and 55 ms, respectively, from a best fit under shifts…
\nEdit:
\nI would like to know whether the delays are bound, whether I could use BObyQA (https://www.damtp.cam.ac.uk/user/na/NA_papers/NA2009_06.pdf ), and they did not seem to be bound, which would be a bit odd, I'd reckon. I was using a sampler minimizer, but the Bobyqa is faster, and with 786 000 events to convert/transform… I think I will drop this idea anyways as it is unoriginal, and if it is not original it is boring, and if it is boring, it is not worth it.
Can I assume that the delay between Hanford and Livingston is < 12 ms? And between Hanford and Virgo < 39 ms?
\nFor 57c3183f4a, I get 13 ms and 55 ms, respectively, from a best fit under shifts…
\nEdit:
\nI would like to know whether the delays are bound, whether I could use BObyQA (https://www.damtp.cam.ac.uk/user/na/NA_papers/NA2009_06.pdf ), and they did not seem to be bound, which would be a bit odd, I'd reckon. I was using a sampler minimizer, but the Bobyqa is faster, and with 786 000 events to convert/transform… I think I will drop this idea anyways as it is unoriginal, and if it is not original it is boring, and if it is boring, it is not worth it.
I was wondering the same thing. It occurred to me that a shift may be due to the signal arriving at the detectors from different directions. I don't know how far apart the detectors are, but light takes about 42 ms to pass through the earth (12,700 km diameter of earth and 300,000 km/s speed of light). So in a worst case of the detectors being at opposite ends of the planet and the signal arriving along that line, there would be a 42 ms delay.) So that may explain at least some of the delay.
\nSo I guess the delay is going to be variable depending on the locations of the detectors and the direction of the signal.
", "rawMarkdown": "I was wondering the same thing. It occurred to me that a shift may be due to the signal arriving at the detectors from different directions. I don't know how far apart the detectors are, but light takes about 42 ms to pass through the earth (12,700 km diameter of earth and 300,000 km/s speed of light). So in a worst case of the detectors being at opposite ends of the planet and the signal arriving along that line, there would be a 42 ms delay.) So that may explain at least some of the delay.\n\nSo I guess the delay is going to be variable depending on the locations of the detectors and the direction of the signal.", "votes": null }, { "id": "1382632", "postDate": "07/10/2021 04:42:39", "content": "Hi @glimmung , may I ask how do you measure these delays, could you please share some code snippet? Thanks.
", "rawMarkdown": "Hi @glimmung , may I ask how do you measure these delays, could you please share some code snippet? Thanks.", "votes": null }, { "id": "1383024", "postDate": "07/10/2021 12:41:07", "content": "I hope it's alright if I refer you to https://arxiv.org/abs/1706.04191 and https://doi.org/10.1103/PhysRevLett.116.061102 ? My code snippet would in C#, if you still want it.
", "rawMarkdown": "I hope it's alright if I refer you to https://arxiv.org/abs/1706.04191 and https://doi.org/10.1103/PhysRevLett.116.061102 ? My code snippet would in C#, if you still want it.", "votes": null }, { "id": "1383027", "postDate": "07/10/2021 12:43:42", "content": "I used the coordinates (46.455, -119.408), (30.563, -90.774) and (43.631, 10.503) from maps.google
", "rawMarkdown": "I used the coordinates (46.455, -119.408), (30.563, -90.774) and (43.631, 10.503) from maps.google", "votes": null }, { "id": "1383053", "postDate": "07/10/2021 13:12:39", "content": "That answers half of the question; determining the distance between the detectors. The other part of the equation is to determine the direction in space from which each event is detected, and I'm not sure that information is available. For example, if the gravitational wave comes from a direction in space so that it arrives at Hanford first, then it will be the first to receive it and there will be a slight delay to the other detectors depending on the angle of its arrival. From other directions, the signal may arrive at two detectors almost simultaneously with a delay to the third. Make sense?
\nSo the big question is, does the test and training data provide us with the incoming angles of the signal relative to the detectors so we can calculate the expected delays for each event?
\nIf they provide us with the celestial coordinates (RA/Dec) and date/time of each event, then we would need to calculate the orientation of the earth to at that point in time to calculate the incoming angles of the wave and arrival time differences at each of the detectors. Not a trivial task, even if the data is available.
\nThe bottom line is that the delays are variable for each event.
\nI think the best solution is to ignore the mechanics of signal arrival times and delays and just try to determine the relative start of each event at the detectors by matching the phases of the signals to each other so they can all be aligned for processing. But it would have to be done separately for each event, because the delays are not just a simple constant for all the events.
", "rawMarkdown": "That answers half of the question; determining the distance between the detectors. The other part of the equation is to determine the direction in space from which each event is detected, and I'm not sure that information is available. For example, if the gravitational wave comes from a direction in space so that it arrives at Hanford first, then it will be the first to receive it and there will be a slight delay to the other detectors depending on the angle of its arrival. From other directions, the signal may arrive at two detectors almost simultaneously with a delay to the third. Make sense?\n\nSo the big question is, does the test and training data provide us with the incoming angles of the signal relative to the detectors so we can calculate the expected delays for each event?\n\nIf they provide us with the celestial coordinates (RA/Dec) and date/time of each event, then we would need to calculate the orientation of the earth to at that point in time to calculate the incoming angles of the wave and arrival time differences at each of the detectors. Not a trivial task, even if the data is available. \n\nThe bottom line is that the delays are variable for each event. \n\nI think the best solution is to ignore the mechanics of signal arrival times and delays and just try to determine the relative start of each event at the detectors by matching the phases of the signals to each other so they can all be aligned for processing. But it would have to be done separately for each event, because the delays are not just a simple constant for all the events.", "votes": null }, { "id": "1386547", "postDate": "07/13/2021 14:42:48", "content": "Hi Tord,
\nGravitational wave travel at the speed of light so the maximum allowed time delay between a signal arriving at 2 detectors will be when the source sky position is on the line that joins the 2 detectors. In that case the time delay would be the straight line distance between the detectors divided by the speed of light.
\nCheers
\nChris
", "rawMarkdown": "Hi Tord,\n\nGravitational wave travel at the speed of light so the maximum allowed time delay between a signal arriving at 2 detectors will be when the source sky position is on the line that joins the 2 detectors. In that case the time delay would be the straight line distance between the detectors divided by the speed of light. \n\nCheers\n\nChris", "votes": null }, { "id": "1386983", "postDate": "07/13/2021 20:35:10", "content": "12 ms = ![https://photos.app.goo.gl/FytjjVs46tarF6EHA]
\nSo, my method of fit must be wrong… But the fit looks pretty good ![https://photos.app.goo.gl/LmMH7ESWzR3Tui2Z8]
", "rawMarkdown": "12 ms = ![https://photos.app.goo.gl/FytjjVs46tarF6EHA]\n\nSo, my method of fit must be wrong... But the fit looks pretty good ![https://photos.app.goo.gl/LmMH7ESWzR3Tui2Z8]", "votes": null }, { "id": "1387136", "postDate": "07/14/2021 01:50:38", "content": "In this paper (page 33) they mention that maximum delay between detectors are the following:
\nThe arrival times of the gravitational waves at each detector must differ by no more than the the maximum time-of-flight between the detector.
", "rawMarkdown": "[In this paper](https://arxiv.org/pdf/1908.11170.pdf) (page 33) they mention that maximum delay between detectors are the following:\n- LIGO Hanford - LIGO Livingston pair is 10ms,with an extra 5 ms added in order to account for uncertainty \n- LIGO Hanford - Virgopair - 27 ms \n- LIGO Livingston - Virgo pair - 26 ms\n\nThe arrival times of the gravitational waves at each detector must differ by no more than the the maximum time-of-flight between the detector.", "votes": null }, { "id": "1563275", "postDate": "10/28/2021 06:53:06", "content": "Hey,
\nThank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.
\nThank you all,
\nRegards,
\nChris
I was wondering the same thing. It occurred to me that a shift may be due to the signal arriving at the detectors from different directions. I don't know how far apart the detectors are, but light takes about 42 ms to pass through the earth (12,700 km diameter of earth and 300,000 km/s speed of light). So in a worst case of the detectors being at opposite ends of the planet and the signal arriving along that line, there would be a 42 ms delay.) So that may explain at least some of the delay.
\nSo I guess the delay is going to be variable depending on the locations of the detectors and the direction of the signal.
", "votes": null, "replies": [ { "id": 1383027, "author_name": "glimmung", "author_url": "", "post_date": "07/10/2021 12:43:42", "content": "I used the coordinates (46.455, -119.408), (30.563, -90.774) and (43.631, 10.503) from maps.google
", "votes": null, "replies": [] }, { "id": 1383053, "author_name": "wilddave", "author_url": "", "post_date": "07/10/2021 13:12:39", "content": "That answers half of the question; determining the distance between the detectors. The other part of the equation is to determine the direction in space from which each event is detected, and I'm not sure that information is available. For example, if the gravitational wave comes from a direction in space so that it arrives at Hanford first, then it will be the first to receive it and there will be a slight delay to the other detectors depending on the angle of its arrival. From other directions, the signal may arrive at two detectors almost simultaneously with a delay to the third. Make sense?
\nSo the big question is, does the test and training data provide us with the incoming angles of the signal relative to the detectors so we can calculate the expected delays for each event?
\nIf they provide us with the celestial coordinates (RA/Dec) and date/time of each event, then we would need to calculate the orientation of the earth to at that point in time to calculate the incoming angles of the wave and arrival time differences at each of the detectors. Not a trivial task, even if the data is available.
\nThe bottom line is that the delays are variable for each event.
\nI think the best solution is to ignore the mechanics of signal arrival times and delays and just try to determine the relative start of each event at the detectors by matching the phases of the signals to each other so they can all be aligned for processing. But it would have to be done separately for each event, because the delays are not just a simple constant for all the events.
", "votes": null, "replies": [] } ] }, { "id": 1382632, "author_name": "superchenhao", "author_url": "", "post_date": "07/10/2021 04:42:39", "content": "Hi @glimmung , may I ask how do you measure these delays, could you please share some code snippet? Thanks.
", "votes": null, "replies": [ { "id": 1383024, "author_name": "glimmung", "author_url": "", "post_date": "07/10/2021 12:41:07", "content": "I hope it's alright if I refer you to https://arxiv.org/abs/1706.04191 and https://doi.org/10.1103/PhysRevLett.116.061102 ? My code snippet would in C#, if you still want it.
", "votes": null, "replies": [] } ] }, { "id": 1386547, "author_name": "bayeswolf", "author_url": "", "post_date": "07/13/2021 14:42:48", "content": "Hi Tord,
\nGravitational wave travel at the speed of light so the maximum allowed time delay between a signal arriving at 2 detectors will be when the source sky position is on the line that joins the 2 detectors. In that case the time delay would be the straight line distance between the detectors divided by the speed of light.
\nCheers
\nChris
", "votes": null, "replies": [ { "id": 1386983, "author_name": "glimmung", "author_url": "", "post_date": "07/13/2021 20:35:10", "content": "12 ms = ![https://photos.app.goo.gl/FytjjVs46tarF6EHA]
\nSo, my method of fit must be wrong… But the fit looks pretty good ![https://photos.app.goo.gl/LmMH7ESWzR3Tui2Z8]
", "votes": null, "replies": [] } ] }, { "id": 1387136, "author_name": "meaninglesslives", "author_url": "", "post_date": "07/14/2021 01:50:38", "content": "In this paper (page 33) they mention that maximum delay between detectors are the following:
\nThe arrival times of the gravitational waves at each detector must differ by no more than the the maximum time-of-flight between the detector.
", "votes": null, "replies": [] }, { "id": 1563275, "author_name": "zerafachris", "author_url": "", "post_date": "10/28/2021 06:53:06", "content": "Hey,
\nThank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.
\nThank you all,
\nRegards,
\nChris