juliensimon/lunar-sample-geochemistry

--- license: cc-by-4.0 pretty_name: "Lunar Sample Geochemistry (Astromat Synthesis)" language: - en description: "Geochemical analyses of 58,289 lunar samples from Apollo, Luna, and Chang'e 5 missions. Major oxides and trace elements from 980 publications." task_categories: - tabular-classification - tabular-regression tags: - space - moon - lunar - geochemistry - apollo - astromat - petrology - planetary-science - open-data - tabular-data - parquet size_categories: - 10K<n<100K configs: - config_name: default data_files: - split: train path: data/lunar_geochemistry.parquet default: true --- # Lunar Sample Geochemistry (Astromat Synthesis) <div align="center"> <img src="banner.jpg" alt="The Moon seen from Apollo 8, showing craters and surface detail" width="400"> <p><em>Credit: NASA/Apollo 8</em></p> </div> *Part of the [Planetary Science Datasets](https://huggingface.co/collections/juliensimon/planetary-science-datasets-69c2d4683bd6a66c34fb4af2) collection on Hugging Face.* A comprehensive compilation of **58,289** geochemical analyses from **14,379** unique lunar samples collected by the Apollo, Luna, and Chang'e 5 missions. Covers major oxide compositions, trace element abundances, and rare earth elements from **980** published studies, compiled by the Astromat/EarthChem project. ## Dataset description Between 1969 and 2020, six Apollo missions (382 kg), three Soviet Luna missions (326 g), and China's Chang'e 5 mission (1.73 kg) returned samples from the Moon's surface. These samples — basalts, breccias, soils, and crustal rocks — have been analyzed in laboratories worldwide for over 50 years, producing a rich geochemical database that underpins our understanding of the Moon's origin, differentiation, and volcanic history. This dataset compiles the results of those analyses into a single tabular format. **Major oxides** (SiO2, TiO2, Al2O3, FeO, MgO, CaO, etc.) in weight percent define the bulk composition and are used to classify lunar rock types (high-Ti basalt, low-Ti basalt, KREEP-rich, ferroan anorthosite). **Trace elements** and **rare earth elements** in ppm provide diagnostic signatures for petrogenetic processes: partial melting, fractional crystallization, and impact mixing. The data was compiled by the Astromat Synthesis project (EarthChem Library) from peer-reviewed publications spanning 1970 to 2025, making it the most comprehensive single source for lunar sample chemistry. ## Sample types | Type | Analyses | |------|----------:| | ROCK>>BRECCIA | 17,455 | | ROCK>>BASALT | 15,949 | | SOIL | 12,703 | | ROCK>>CRUSTAL | 1,713 | | ROCK>>CORE | 1,179 | | ROCK>>UNCLASSIFIED | 132 | ## Missions | Mission | Analyses | |---------|----------:| | Other | 58,289 | ## Quick stats - **58,289** total analyses from **14,379** unique samples - **980** published references - **55** columns (metadata + major oxides + trace elements + REEs) ## Usage ```python from datasets import load_dataset ds = load_dataset("juliensimon/lunar-sample-geochemistry", split="train") df = ds.to_pandas() # TAS-like plot for lunar basalts import matplotlib.pyplot as plt basalts = df[df["sample_type"].str.contains("BASALT", na=False)] valid = basalts.dropna(subset=["sio2_wt_pct", "tio2_wt_pct"]) plt.scatter(valid["sio2_wt_pct"], valid["tio2_wt_pct"], alpha=0.3, s=10) plt.xlabel("SiO2 (wt%)") plt.ylabel("TiO2 (wt%)") plt.title("Lunar Basalt Compositions") plt.show() # Compare missions print(df.groupby("mission")[["sio2_wt_pct", "feo_wt_pct", "mgo_wt_pct"]].mean()) # Filter Apollo 15 samples a15 = df[df["mission"] == "Apollo 15"] print(f"Apollo 15: {len(a15)} analyses") ``` ## Data source Astromat Synthesis Compilation: Lunar samples v.1 (February 2025). EarthChem Library, [doi:10.60520/IEDA/113696](https://doi.org/10.60520/IEDA/113696). CC-BY-4.0. ## Related datasets - [Lunar Craters (Robbins 2019)](https://huggingface.co/datasets/juliensimon/lunar-craters-robbins) — 2M+ Moon craters - [Meteorite Database](https://huggingface.co/datasets/juliensimon/meteorite-database) — Named meteorites with classification - [Meteorite Landings](https://huggingface.co/datasets/juliensimon/meteorite-landings) — 45K+ meteorite fall records - [IAU Planetary Nomenclature](https://huggingface.co/datasets/juliensimon/planetary-nomenclature) — Named lunar features ## Pipeline Source code: [juliensimon/space-datasets](https://github.com/juliensimon/space-datasets) ## Support If you find this dataset useful, please give it a ❤️ on the [dataset page](https://huggingface.co/datasets/juliensimon/lunar-sample-geochemistry) and share feedback in the Community tab! Also consider giving a ⭐ to the [space-datasets](https://github.com/juliensimon/space-datasets) repo. ## Citation ```bibtex @dataset{lunar_geochemistry, author = {Simon, Julien}, title = {Lunar Sample Geochemistry (Astromat Synthesis)}, year = {2026}, publisher = {Hugging Face}, url = {https://huggingface.co/datasets/juliensimon/lunar-sample-geochemistry}, note = {Based on Astromat Synthesis Compilation via EarthChem Library} } ``` ## License [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)--- license: cc-by-4.0 pretty_name: "月球样品地球化学（Astromat合成版）" language: - en description: "本数据集包含阿波罗（Apollo）、月球号（Luna）以及嫦娥五号（Chang'e 5）任务返回的共计58289份月球样品的地球化学分析数据，其主量氧化物（major oxides）与微量元素（trace elements）数据源自980篇已发表文献。" task_categories: - tabular-classification - tabular-regression tags: - 太空（space） - 月球（moon） - 月球样品（lunar） - 地球化学（geochemistry） - 阿波罗（Apollo） - Astromat - 岩石学（petrology） - 行星科学（planetary-science） - 开放数据（open-data） - 表格数据（tabular-data） - Parquet格式（parquet） size_categories: - 10K<n<100K configs: - config_name: default data_files: - split: train path: data/lunar_geochemistry.parquet default: true --- # 月球样品地球化学（Astromat合成数据集） <div align="center"> <img src="banner.jpg" alt="阿波罗8号拍摄的月球表面，展示陨石坑与地表细节" width="400"> <p><em>图源：NASA/阿波罗8号（Apollo 8）</em></p> </div> *本数据集属于Hugging Face平台上的[行星科学数据集（Planetary Science Datasets）]合集。* 本数据集全面整合了阿波罗（Apollo）、月球号（Luna）与嫦娥五号（Chang'e 5）任务采集的14379份独立月球样品的共计58289项地球化学分析结果。数据涵盖主量氧化物（major oxides）组成、微量元素（trace element）丰度以及稀土元素（rare earth elements），源自980项已发表研究，由Astromat/EarthChem项目整合完成。 ## 数据集说明 1969年至2020年间，6次阿波罗（Apollo）任务（带回样品总重382千克）、3次苏联月球号（Luna）任务（带回样品总重326克）以及中国嫦娥五号（Chang'e 5）任务（带回样品总重1.73千克）从月球表面取回了月球样品。这些样品包括玄武岩（basalt）、角砾岩（breccia）、月壤与地壳岩石，已在全球各地实验室开展了超过50年的分析，构建了丰富的地球化学数据库，为我们理解月球起源、分异演化与火山历史提供了核心支撑。 本数据集将上述分析结果整合为统一的表格格式。**主量氧化物（major oxides）**（SiO₂、TiO₂、Al₂O₃、FeO、MgO、CaO等）以重量百分比计，定义了样品的整体组成，可用于月球岩石类型分类（如高钛玄武岩、低钛玄武岩、富KREEP岩石、铁镁质斜长岩）。**微量元素（trace elements）**与**稀土元素（rare earth elements）**以ppm为单位，可为成岩作用（partial melting部分熔融、fractional crystallization分离结晶、impact mixing撞击混合）提供诊断性地球化学标识。 本数据集由Astromat合成项目（EarthChem Library）从1970年至2025年的同行评议文献中整合而来，是目前最全面的月球样品地球化学单一数据源。 ## 样品类型 | 样品类型 | 分析项数 | |---------|---------:| | 岩石>>角砾岩（ROCK>>BRECCIA） | 17455 | | 岩石>>玄武岩（ROCK>>BASALT） | 15949 | | 月壤（SOIL） | 12703 | | 岩石>>地壳岩（ROCK>>CRUSTAL） | 1713 | | 岩石>>岩芯（ROCK>>CORE） | 1179 | | 岩石>>未分类（ROCK>>UNCLASSIFIED） | 132 | ## 任务类型 | 任务类型 | 分析项数 | |---------|---------:| | 其他任务（Other） | 58289 | ## 核心统计数据 - **58289** 项总分析结果，覆盖**14379** 份独立样品 - **980** 篇已发表参考文献 - 共包含**55** 列数据（元数据 + 主量氧化物 + 微量元素 + 稀土元素） ## 使用示例 python from datasets import load_dataset ds = load_dataset("juliensimon/lunar-sample-geochemistry", split="train") df = ds.to_pandas() # 月球玄武岩TAS类绘图 import matplotlib.pyplot as plt basalts = df[df["sample_type"].str.contains("BASALT", na=False)] valid = basalts.dropna(subset=["sio2_wt_pct", "tio2_wt_pct"]) plt.scatter(valid["sio2_wt_pct"], valid["tio2_wt_pct"], alpha=0.3, s=10) plt.xlabel("SiO₂ (wt%)") plt.ylabel("TiO₂ (wt%)") plt.title("月球玄武岩组成") plt.show() # 对比不同任务的元素丰度 print(df.groupby("mission")[["sio2_wt_pct", "feo_wt_pct", "mgo_wt_pct"]].mean()) # 筛选阿波罗15号样品 a15 = df[df["mission"] == "Apollo 15"] print(f"阿波罗15号：共{len(a15)}项分析结果") ## 数据来源 Astromat合成汇编：月球样品v1版（2025年2月），EarthChem图书馆，[DOI:10.60520/IEDA/113696](https://doi.org/10.60520/IEDA/113696)，采用CC-BY-4.0许可协议。 ## 相关数据集 - [月球陨石坑（Robbins 2019）](https://huggingface.co/datasets/juliensimon/lunar-craters-robbins) — 包含超过200万个月球陨石坑 - [陨石数据库](https://huggingface.co/datasets/juliensimon/meteorite-database) — 带分类信息的命名陨石样本 - [陨石降落记录](https://huggingface.co/datasets/juliensimon/meteorite-landings) — 包含45000余条陨石降落记录 - [国际天文学联合会行星命名规范（IAU Planetary Nomenclature）](https://huggingface.co/datasets/juliensimon/planetary-nomenclature) — 已命名的月球地貌特征 ## 数据处理流程 源代码：[juliensimon/space-datasets](https://github.com/juliensimon/space-datasets) ## 支持与反馈 若您认为本数据集对研究有所帮助，请前往[数据集页面](https://huggingface.co/datasets/juliensimon/lunar-sample-geochemistry)点赞，并在社区标签页分享反馈！也欢迎为[space-datasets](https://github.com/juliensimon/space-datasets)代码仓库点亮Star。 ## 引用格式 bibtex @dataset{lunar_geochemistry, author = {Simon, Julien}, title = {月球样品地球化学（Astromat合成数据集）}, year = {2026}, publisher = {Hugging Face}, url = {https://huggingface.co/datasets/juliensimon/lunar-sample-geochemistry}, note = {基于Astromat合成汇编项目与EarthChem图书馆数据构建} } ## 许可证 [CC-BY-4.0知识共享署名4.0国际许可协议](https://creativecommons.org/licenses/by/4.0/)