{
  "id": 379847,
  "title": "One string is not perfectly vertical?",
  "url": "/competitions/icecube-neutrinos-in-deep-ice/discussion/379847",
  "author_name": "",
  "post_date": "2023-01-21T09:43:38.431028100Z",
  "votes": 7,
  "comment_count": 2,
  "views": 0,
  "content": "<p>For a vertical string we would expect a single x-y value.</p>\n<p>Sensor ids 1200-1259 seem to have slight x-y variation, suggesting a slightly deviated well (we're talking about a step out of less than 1m, so nothing significant for this task)</p>\n<p>Question to the organisers - Is this because this was the only well to receive a gyro survey and the other 85 wells are assumed vertical (constant x-y)? If that is the case it gives us some sort of idea of x-y uncertainty.</p>\n<p>Also is there any information on the equipment used for the depth (z) measurement? Typical oil field depth encoders have an accuracy of +/-1.5m per 3050m</p>",
  "messages": [
    {
      "id": "2109296",
      "postDate": "01/21/2023 09:43:38",
      "content": "<p>For a vertical string we would expect a single x-y value.</p>\n<p>Sensor ids 1200-1259 seem to have slight x-y variation, suggesting a slightly deviated well (we're talking about a step out of less than 1m, so nothing significant for this task)</p>\n<p>Question to the organisers - Is this because this was the only well to receive a gyro survey and the other 85 wells are assumed vertical (constant x-y)? If that is the case it gives us some sort of idea of x-y uncertainty.</p>\n<p>Also is there any information on the equipment used for the depth (z) measurement? Typical oil field depth encoders have an accuracy of +/-1.5m per 3050m</p>",
      "rawMarkdown": "For a vertical string we would expect a single x-y value.\n\nSensor ids 1200-1259 seem to have slight x-y variation, suggesting a slightly deviated well (we're talking about a step out of less than 1m, so nothing significant for this task)\n\nQuestion to the organisers - Is this because this was the only well to receive a gyro survey and the other 85 wells are assumed vertical (constant x-y)? If that is the case it gives us some sort of idea of x-y uncertainty.\n\nAlso is there any information on the equipment used for the depth (z) measurement? Typical oil field depth encoders have an accuracy of +/-1.5m per 3050m",
      "votes": null
    },
    {
      "id": "2109477",
      "postDate": "01/21/2023 12:18:54",
      "content": "<p>The horizontal sensor positions are only known to, say, within a meter. This comes from drilling deep to 2.5 km as you point out. But also keep in mind that the inter-string distances in the main array are &gt;100m, so we're still talking about ~1% accuracy. (In the DeepCore region the distances are closer to 40m.)</p>\n<p>The vertical positions are known to a better precision, especially the spacings between sensors on one string are known to cm precision.</p>",
      "rawMarkdown": "The horizontal sensor positions are only known to, say, within a meter. This comes from drilling deep to 2.5 km as you point out. But also keep in mind that the inter-string distances in the main array are >100m, so we're still talking about ~1% accuracy. (In the DeepCore region the distances are closer to 40m.)\n\nThe vertical positions are known to a better precision, especially the spacings between sensors on one string are known to cm precision.",
      "votes": null
    },
    {
      "id": "2109530",
      "postDate": "01/21/2023 12:58:47",
      "content": "<p>Thanks for the reply. Yes totally understand that the relative position between sensors are known to high accuracy but I’m trying to understand the absolute accuracy since this might provide interesting avenues for data augmentation. </p>\n<p>In deep oil and gas wells for example, total depth is the most uncertain measurement, since different measurement techniques have different error (drillers depth, wireline depth, survey/gyro depth)</p>",
      "rawMarkdown": "Thanks for the reply. Yes totally understand that the relative position between sensors are known to high accuracy but I’m trying to understand the absolute accuracy since this might provide interesting avenues for data augmentation. \n\nIn deep oil and gas wells for example, total depth is the most uncertain measurement, since different measurement techniques have different error (drillers depth, wireline depth, survey/gyro depth)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2109477,
      "author_name": "pellerphys",
      "author_url": "",
      "post_date": "01/21/2023 12:18:54",
      "content": "<p>The horizontal sensor positions are only known to, say, within a meter. This comes from drilling deep to 2.5 km as you point out. But also keep in mind that the inter-string distances in the main array are &gt;100m, so we're still talking about ~1% accuracy. (In the DeepCore region the distances are closer to 40m.)</p>\n<p>The vertical positions are known to a better precision, especially the spacings between sensors on one string are known to cm precision.</p>",
      "votes": null,
      "replies": [
        {
          "id": 2109530,
          "author_name": "anjum48",
          "author_url": "",
          "post_date": "01/21/2023 12:58:47",
          "content": "<p>Thanks for the reply. Yes totally understand that the relative position between sensors are known to high accuracy but I’m trying to understand the absolute accuracy since this might provide interesting avenues for data augmentation. </p>\n<p>In deep oil and gas wells for example, total depth is the most uncertain measurement, since different measurement techniques have different error (drillers depth, wireline depth, survey/gyro depth)</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2109296": "For a vertical string we would expect a single x-y value.\n\nSensor ids 1200-1259 seem to have slight x-y variation, suggesting a slightly deviated well (we're talking about a step out of less than 1m, so nothing significant for this task)\n\nQuestion to the organisers - Is this because this was the only well to receive a gyro survey and the other 85 wells are assumed vertical (constant x-y)? If that is the case it gives us some sort of idea of x-y uncertainty.\n\nAlso is there any information on the equipment used for the depth (z) measurement? Typical oil field depth encoders have an accuracy of +/-1.5m per 3050m",
    "2109477": "The horizontal sensor positions are only known to, say, within a meter. This comes from drilling deep to 2.5 km as you point out. But also keep in mind that the inter-string distances in the main array are >100m, so we're still talking about ~1% accuracy. (In the DeepCore region the distances are closer to 40m.)\n\nThe vertical positions are known to a better precision, especially the spacings between sensors on one string are known to cm precision.",
    "2109530": "Thanks for the reply. Yes totally understand that the relative position between sensors are known to high accuracy but I’m trying to understand the absolute accuracy since this might provide interesting avenues for data augmentation. \n\nIn deep oil and gas wells for example, total depth is the most uncertain measurement, since different measurement techniques have different error (drillers depth, wireline depth, survey/gyro depth)"
  },
  "source": "meta"
}