The webinar documentation states that we can expect 10^10 events caused by atmospheric muons. That should be big enough a number to assume that many of these appear as pollution in the actual neutrino events. The question arises whether these have been considered in the simulations, and how.
\nOf course we could estimate the number of polluted events by this formula:
\n[sum(maxtime) - sum(mintime)] / yeartime * 10^10
\n(for all events)
That would be roughly 500000 out of the 134 million events. But if the simulation does not consider it this way, it might be misleading.
\nAny insights on this?
", "messages": [ { "id": "2153737", "postDate": "02/21/2023 15:54:07", "content": "The webinar documentation states that we can expect 10^10 events caused by atmospheric muons. That should be big enough a number to assume that many of these appear as pollution in the actual neutrino events. The question arises whether these have been considered in the simulations, and how.
\nOf course we could estimate the number of polluted events by this formula:
\n[sum(maxtime) - sum(mintime)] / yeartime * 10^10
\n(for all events)
That would be roughly 500000 out of the 134 million events. But if the simulation does not consider it this way, it might be misleading.
\nAny insights on this?
", "rawMarkdown": "The webinar documentation states that we can expect 10^10 events caused by atmospheric muons. That should be big enough a number to assume that many of these appear as pollution in the actual neutrino events. The question arises whether these have been considered in the simulations, and how.\n\nOf course we could estimate the number of polluted events by this formula:\n\n[sum(maxtime) - sum(mintime)] / yeartime * 10^10 \n(for all events)\n\nThat would be roughly 500000 out of the 134 million events. But if the simulation does not consider it this way, it might be misleading.\n\nAny insights on this?", "votes": null }, { "id": "2153763", "postDate": "02/21/2023 16:13:02", "content": "IMO one of the major noise issues faced by Icecube has to be atmospheric muons.
\nIf they already have a great filter of this noise than not much reason to give too much of it to us :)
\nIf they don't have a great filter than you would expect our data to have its fair share of noise from this source. In my notes from the video presentation I have \"expect 25% real junk\". Would have to re-watch the video to see why I wrote that down.
", "rawMarkdown": "IMO one of the major noise issues faced by Icecube has to be atmospheric muons. \n\nIf they already have a great filter of this noise than not much reason to give too much of it to us :)\n\nIf they don't have a great filter than you would expect our data to have its fair share of noise from this source. In my notes from the video presentation I have \"expect 25% real junk\". Would have to re-watch the video to see why I wrote that down.", "votes": null }, { "id": "2154702", "postDate": "02/22/2023 07:46:27", "content": "From the hosts' presentation and discussions here, I see the following picture.
\nThe data is simulated, and every event was injected with a neutrino, coming from a random direction.
\nAs the energy and flavor of injected neutrinos vary, we have three possible event forms:
1) muon neutrunos produce track-like events
\n2) tau and e-neutrinos produce clump-like events
\n3) some neutrinos (due to low energy, maybe) fail to register many pulses, and events are hardly reconstructible
I understand that we have roughly the same proportions of those three types, like 25-35% each.
\nOn top of that, some events also contain a muon (mostly from above, due to Earth shading). IIRC from the presentation, less than 10% of events contain such muons.
", "rawMarkdown": "From the hosts' presentation and discussions here, I see the following picture.\n\nThe data is simulated, and every event was injected with a neutrino, coming from a random direction.\nAs the energy and flavor of injected neutrinos vary, we have three possible event forms:\n\n1) muon neutrunos produce track-like events\n2) tau and e-neutrinos produce clump-like events\n3) some neutrinos (due to low energy, maybe) fail to register many pulses, and events are hardly reconstructible\n\nI understand that we have roughly the same proportions of those three types, like 25-35% each.\n\nOn top of that, some events also contain a muon (mostly from above, due to Earth shading). IIRC from the presentation, less than 10% of events contain such muons.", "votes": null }, { "id": "2155234", "postDate": "02/22/2023 14:11:26", "content": "I had just dug a bit deeper into that, and found in a paper on page 10 the following statement:
\n\n\nfor roughly 9% of the events collected by the data collection algorithm, more than one muon will be passing though the detector simultaneously
\n
That would be roughly 12 million of the total 134 million. Seems worthwhile to identify and remove these.
", "rawMarkdown": "I had just dug a bit deeper into that, and found in [a paper](https://arxiv.org/abs/1308.5501) on page 10 the following statement:\n>for roughly 9% of the events collected by the data collection algorithm, more than one muon will be passing though the detector simultaneously\n\nThat would be roughly 12 million of the total 134 million. Seems worthwhile to identify and remove these.", "votes": null }, { "id": "2156383", "postDate": "02/23/2023 09:57:42", "content": "You are right, it is definitely worth trying to identify the events with two tracks, and process them separately.
\nI'd add two notes to this: 1) our dataset may contain a different percentage of atmospreric muons, and 2) we probably should not remove these events, as they contain not only muon, but a neutrino as well.
", "rawMarkdown": "You are right, it is definitely worth trying to identify the events with two tracks, and process them separately.\n\nI'd add two notes to this: 1) our dataset may contain a different percentage of atmospreric muons, and 2) we probably should not remove these events, as they contain not only muon, but a neutrino as well.", "votes": null } ], "comments": [ { "id": 2153763, "author_name": "pcjimmmy", "author_url": "", "post_date": "02/21/2023 16:13:02", "content": "IMO one of the major noise issues faced by Icecube has to be atmospheric muons.
\nIf they already have a great filter of this noise than not much reason to give too much of it to us :)
\nIf they don't have a great filter than you would expect our data to have its fair share of noise from this source. In my notes from the video presentation I have \"expect 25% real junk\". Would have to re-watch the video to see why I wrote that down.
", "votes": null, "replies": [] }, { "id": 2154702, "author_name": "alexz0", "author_url": "", "post_date": "02/22/2023 07:46:27", "content": "From the hosts' presentation and discussions here, I see the following picture.
\nThe data is simulated, and every event was injected with a neutrino, coming from a random direction.
\nAs the energy and flavor of injected neutrinos vary, we have three possible event forms:
1) muon neutrunos produce track-like events
\n2) tau and e-neutrinos produce clump-like events
\n3) some neutrinos (due to low energy, maybe) fail to register many pulses, and events are hardly reconstructible
I understand that we have roughly the same proportions of those three types, like 25-35% each.
\nOn top of that, some events also contain a muon (mostly from above, due to Earth shading). IIRC from the presentation, less than 10% of events contain such muons.
", "votes": null, "replies": [] }, { "id": 2155234, "author_name": "taqseorangpun", "author_url": "", "post_date": "02/22/2023 14:11:26", "content": "I had just dug a bit deeper into that, and found in a paper on page 10 the following statement:
\n\n\nfor roughly 9% of the events collected by the data collection algorithm, more than one muon will be passing though the detector simultaneously
\n
That would be roughly 12 million of the total 134 million. Seems worthwhile to identify and remove these.
", "votes": null, "replies": [ { "id": 2156383, "author_name": "alexz0", "author_url": "", "post_date": "02/23/2023 09:57:42", "content": "You are right, it is definitely worth trying to identify the events with two tracks, and process them separately.
\nI'd add two notes to this: 1) our dataset may contain a different percentage of atmospreric muons, and 2) we probably should not remove these events, as they contain not only muon, but a neutrino as well.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "2153737": "The webinar documentation states that we can expect 10^10 events caused by atmospheric muons. That should be big enough a number to assume that many of these appear as pollution in the actual neutrino events. The question arises whether these have been considered in the simulations, and how.\n\nOf course we could estimate the number of polluted events by this formula:\n\n[sum(maxtime) - sum(mintime)] / yeartime * 10^10 \n(for all events)\n\nThat would be roughly 500000 out of the 134 million events. But if the simulation does not consider it this way, it might be misleading.\n\nAny insights on this?", "2153763": "IMO one of the major noise issues faced by Icecube has to be atmospheric muons. \n\nIf they already have a great filter of this noise than not much reason to give too much of it to us :)\n\nIf they don't have a great filter than you would expect our data to have its fair share of noise from this source. In my notes from the video presentation I have \"expect 25% real junk\". Would have to re-watch the video to see why I wrote that down.", "2154702": "From the hosts' presentation and discussions here, I see the following picture.\n\nThe data is simulated, and every event was injected with a neutrino, coming from a random direction.\nAs the energy and flavor of injected neutrinos vary, we have three possible event forms:\n\n1) muon neutrunos produce track-like events\n2) tau and e-neutrinos produce clump-like events\n3) some neutrinos (due to low energy, maybe) fail to register many pulses, and events are hardly reconstructible\n\nI understand that we have roughly the same proportions of those three types, like 25-35% each.\n\nOn top of that, some events also contain a muon (mostly from above, due to Earth shading). IIRC from the presentation, less than 10% of events contain such muons.", "2155234": "I had just dug a bit deeper into that, and found in [a paper](https://arxiv.org/abs/1308.5501) on page 10 the following statement:\n>for roughly 9% of the events collected by the data collection algorithm, more than one muon will be passing though the detector simultaneously\n\nThat would be roughly 12 million of the total 134 million. Seems worthwhile to identify and remove these.", "2156383": "You are right, it is definitely worth trying to identify the events with two tracks, and process them separately.\n\nI'd add two notes to this: 1) our dataset may contain a different percentage of atmospreric muons, and 2) we probably should not remove these events, as they contain not only muon, but a neutrino as well." }, "source": "meta" }