The Impact of Satellite Trails on Hubble Space Telescope Observations: satellite classifications
收藏资源简介:
This repository contains the Hubble Space Telescope (HST) observations with satellite classifications, released in the paper "<em>The Impact of Satellite Trails on Hubble Space Telescope Observations</em>" (DOI: 10.1038/s41550-023-01903-3<em>).</em> This table contains 114 607 individual HST images taken in the last 19 years and publicly released in the eHST archive by 3 October 2021, with satellite trail classifications made with machine learning and citizen science. We processed the individual images by adding the two ACS/WFC (WFC3/UVIS, respectively) apertures side-by-side, without correcting for geometric distortions and without the gap between the two detectors (hence why the satellite trails can appear discontinuous in the images). Note that these are not original images from the eHST archive, and therefore are not meant for other scientific analysis. The dataset contains 3072 HST images with satellites (2.7% of the dataset) and 3228 satellite trails in total. The classifications were visually inspected and vetted by the authors (images are flagged with the 'satellite' flag). The table contains the following columns: observation IDs: both individual exposures (simple_id) and composite images, multiple individual exposures processed and stacked (composite_id); instrument: ACS/WFC or WFC3/UVIS; start time and end time of exposure; exposure duration; right ascension (ra) and declination (dec); 'satellite' flag [empty otherwise]; no_sat: number of satellites in the image; image URL - URL for the individual HST observations used for satellite classification; additional metadata columns, as available in the eHST archive. The satellites were classified by volunteers on the Hubble Asteroid Hunter citizen science project and with a machine learning classifier. Please cite the paper (Kruk et al., https://www.nature.com/articles/s41550-023-01903-3) when using the data in this repository.
本仓库包含带有卫星轨迹分类标注的哈勃空间望远镜(Hubble Space Telescope, HST)观测数据,相关成果发表于论文《哈勃空间望远镜观测中的卫星轨迹影响》(DOI: 10.1038/s41550-023-01903-3)。本数据集包含2021年10月3日前公开发布于eHST档案库的、近19年间拍摄的114607张独立哈勃空间望远镜图像,其卫星轨迹分类通过机器学习与公众科学项目完成。我们对单张图像进行了处理,将两个ACS/WFC(分别对应WFC3/UVIS)的成像区域并排拼接,未校正几何畸变,也未消除两个探测器间的间隙——这便是图像中的卫星轨迹可能呈现不连续状态的原因。请注意,本数据集并非eHST档案库中的原始图像,因此不应用于其他科学分析。该数据集包含3072张含卫星的哈勃空间望远镜图像(占数据集总量的2.7%),总计3228条卫星轨迹。所有分类结果均经作者人工检视与审核,图像已标记'satellite'标签。本表格包含以下字段:观测ID:涵盖单次曝光图像(simple_id)与经多帧叠加处理的合成图像(composite_id);仪器类型:ACS/WFC或WFC3/UVIS;曝光开始与结束时间;曝光时长;赤经(right ascension, ra)与赤纬(declination, dec);'satellite'标签(无卫星时该项为空);no_sat:图像中的卫星数量;图像URL:用于卫星轨迹分类的哈勃空间望远镜单张观测图像的链接;以及eHST档案库中包含的其他附加元数据列。本数据集的卫星轨迹分类由哈勃小行星猎人(Hubble Asteroid Hunter)公众科学项目的志愿者与机器学习分类器共同完成。使用本仓库中的数据时,请引用该论文(Kruk等,https://www.nature.com/articles/s41550-023-01903-3)。



