Data from the TESS All-Sky Rotation Survey (TARS)

This Zenodo contains the large data files for the TESS All-Sky Rotation Survey (TARS; Boyle, Bouma, and Mann 2026). TARS is an all-sky catalog of stellar rotation period estimates for 944,056 stars with TESS magnitude T < 16 and distances within 500 pc, derived from TESS full-frame image light curves spanning Sectors 1–96 (2018 July – 2025 September). The catalog more than doubles the number of bright stars with known rotation periods within 100 pc and increases the count within 500 pc by a factor of 3.7. Two random forest classifiers provide per-measurement probabilities for separating instrumental systematics from astrophysical signals and for identifying half-period aliases, enabling users to tune the trade-off between completeness and reliability for their science case. Files tars_table_2.feather — Our default catalog of 944,056 adopted rotation periods. This file contains one row per star, with the adopted rotation period, uncertainty, quality flags, Gaia astrometry and photometry, and sector-level classifier probabilities. The structure and column descriptions are given in Table 2 of the manuscript. This is the catalog most users will want to start with. tars_table_2.csv.zip -- Same as tars_table_2.feather but in .CSV format. tars_table_4.feather — The full catalog of all ~39 million sector-level rotation measurements for 7.5 million target stars. **Warning: this file is approximately 18 GB.** Each row represents one star–sector combination with Lomb-Scargle periodogram parameters, classifier probabilities, and Gaia cross-matched stellar properties. The column descriptions are given in Table 4 (Appendix C) of the manuscript. You will need this file if you want to regenerate the rotation catalog with different classifier thresholds or quality cuts using the master_validation.py script. tars_table_4.csv.zip -- Same as tars_table_4.feather but in .CSV format. **Warning: This file is ~18 GB zipped and ~40 GB unzipped. It will take longer to read into python than the feather file. validation_samples.zip — The four external validation samples used in Section 5 of the paper: Kepler/McQuillan (McQuillan et al. 2014), K2 (Reinhold & Hekker 2020), ZTF (Lu et al. 2022), and open clusters (Long et al. 2023). Each file contains the TESS rotation measurements for stars in the corresponding reference sample, with columns matching tars_table_4.feather plus a Prot column containing the literature rotation period. If you want to run the master_validation.py script described in the paper, download this file, uncompress it, and place the resulting validation_samples/ directory alongside the script. master_validation.py -- A user can use this script to generate their own catalog of rotation periods with different choices than our default catalog. This will load the validation samples and apply the user's given input criteria to the validation samples to give an estimate of the completeness and reliability of the output period catalog built with the user's input parameters. Running this on my 2023 Mac Studio (M2 Ultra chip, 64 GB memory, 24 cores) took ~17 minutes to make the default TARS catalog. master_validation_lowmem.py -- Same as master_validation.py, except optimized to run on computers that have less memory available. Running this on my 2021 MacBook Pro (M1 Max chip, 32 GB memory, 10 cores) took ~45 minutes to make the default TARS catalog. master_validation_requirements.txt - The python package versions that we used to run the master_validation.py script. HOW_TO_USE.md -- Description of how to use the master_validation.py script. Reproducing and Customizing the Catalog We provide a Python script, master_validation.py, that allows users to regenerate the rotation catalog with different systematics classifier thresholds, alias classifier thresholds, quality flag selections, and alias-handling strategies. The script also runs the same validation tests from the paper (Section 5) so users can assess the completeness and reliability of their custom selections. Full usage instructions are available in HOW_TO_USE.md. Light Curves and Vetting Plots The 39 million TESS light curves and vetting plots used in this analysis are available as a MAST High-Level Science Product at https://archive.stsci.edu/hlsp/tars. This HLSP will become active upon paper acceptance. Citation If you use these data, please cite Boyle, Bouma, & Mann (2026).本Zenodo知识库收纳了TESS全天自转巡天（TESS All-Sky Rotation Survey，简称TARS；Boyle、Bouma与Mann，2026）的全部大型数据文件。 TARS是一份全天恒星自转周期估算星表，涵盖944056颗TESS星等T<16、距离小于500秒差距的恒星，数据源自2018年7月至2025年9月期间TESS第1至96天区的全帧图像测光光变曲线。该星表将100秒差距范围内已知自转周期的亮星数量提升了一倍以上，并使500秒差距范围内的该类恒星数量增长至原有3.7倍。本研究采用两套随机森林分类器，为每一项测量提供分类概率，以区分仪器系统误差与天体物理信号，并识别半周期混叠现象，使用者可根据自身科学研究场景，调整完整性与可靠性之间的权衡比例。 **文件** - `tars_table_2.feather`：本研究的默认星表，包含944056颗恒星的采用自转周期。该文件以单恒星为一行，存储了采用自转周期、不确定度、质量标记、盖亚（Gaia）天体测量与测光数据，以及天区级分类器概率。星表结构与字段说明详见论文的表2，为多数使用者的首选起始数据集。 - `tars_table_2.csv.zip`：与`tars_table_2.feather`内容一致，仅格式为压缩CSV格式。 - `tars_table_4.feather`：包含750万颗目标恒星全部约3900万条天区级自转测量数据的完整星表。**警告：该文件大小约18 GB。** 每一行对应一颗恒星在单个天区的测量结果，存储了隆-斯卡尔戈周期图（Lomb-Scargle periodogram）参数、分类器概率以及盖亚交叉匹配得到的恒星属性。字段说明详见论文的表4（附录C）。若使用者需通过`master_validation.py`脚本以不同分类器阈值或质量筛选条件重新生成自转星表，则需要该文件。 - `tars_table_4.csv.zip`：与`tars_table_4.feather`内容一致，仅格式为压缩CSV格式。**警告：该文件压缩后约18 GB，解压后约40 GB，在Python中读取耗时较feather格式文件更长。** - `validation_samples.zip`：包含论文第5节所用的四组外部验证样本，分别为开普勒/麦克奎兰样本（McQuillan等，2014）、K2样本（Reinhold与Hekker，2020）、ZTF样本（Lu等，2022）以及疏散星团样本（Long等，2023）。每个文件存储了对应参考样本中恒星的TESS自转测量数据，字段与`tars_table_4.feather`一致，额外包含一列`Prot`，存储已发表的自转周期。若需运行论文中介绍的`master_validation.py`脚本，请下载该文件并解压，将生成的`validation_samples/`目录与脚本置于同一路径下。 - `master_validation.py`：使用者可通过该脚本生成自定义自转星表，可选用与本研究默认设置不同的参数。该脚本将加载验证样本，并应用使用者指定的筛选条件，从而基于自定义参数生成的自转星表估算其完整性与可靠性。在搭载M2 Ultra芯片、64 GB内存、24核心的2023款Mac Studio上运行该脚本生成默认TARS星表，耗时约17分钟。 - `master_validation_lowmem.py`：与`master_validation.py`功能一致，针对内存较小的设备进行了优化。在搭载M1 Max芯片、32 GB内存、10核心的2021款MacBook Pro上运行该脚本生成默认TARS星表，耗时约45分钟。 - `master_validation_requirements.txt`：记录了运行`master_validation.py`脚本所需的Python包版本信息。 - `HOW_TO_USE.md`：`master_validation.py`脚本的使用说明文档。 **星表复刻与自定义** 本研究提供Python脚本`master_validation.py`，使用者可通过调整系统误差分类器阈值、混叠分类器阈值、质量标记筛选条件以及混叠处理策略，重新生成自转星表。该脚本同时运行论文第5节中的验证测试，以帮助使用者评估自定义参数下生成星表的完整性与可靠性。完整使用说明详见`HOW_TO_USE.md`文件。 **光变曲线与校验绘图** 本分析所用的3900万条TESS光变曲线及校验绘图，可作为米库尔斯基空间望远镜档案馆（MAST）高级科学产品（HLSP）从https://archive.stsci.edu/hlsp/tars获取。该HLSP将在论文录用后正式上线。 **引用说明** 若使用本数据集，请引用Boyle、Bouma与Mann（2026）的研究成果。