Dataset for EUROCONTROL PRC 2024 Data Challenge on Aircraft Take-off Weight Estimation

This dataset was created for the EUROCONTROL Performance Review Commission (PRC) Data Challenge 2024, which aimed to engage data scientists and aviation enthusiasts in developing open models to estimate aircraft take-off weight from readily available open data sources.<br><br><strong>Overview</strong>---The dataset represents a unique collection of otherwise difficult-to-obtain flight information, combining:- Flight information (origin/destination, take-off/landing times, aircraft type)- ADS-B trajectory data from OpenSky- Actual take-off weight data from airlines<br>This dataset contains 527,162 flights flown throughout Europe in 2022, representing 6.1% of all flights in EUROCONTROL airspace.<br><br><strong>DOI for the data repository</strong>---https://doi.org/10.4121/8cb8484b-dbe7-4750-8b87-a5b1dbc621b4<br><br><strong>Background</strong>---Aircraft take-off weight (ATOW) is a critical parameter for modeling fuel consumption and estimating aviation emissions. This dataset enables the development of open models to predict ATOW from publicly available flight data.<br><br><strong>Citation</strong>---If you use data from this repository, please cite this dataset and also the following paper:<br><strong>Aircraft Take-off Weight Estimation: The EUROCONTROL PRC 2024 Data Challenge</strong>Enrico Spinielli; Junzi Sun; Martin Strohmeier; Xavier Olive; Quinten Goens; Rainer Koelle; Allan Tart; John FitzgeraldJournal of Open Aviation Science, 2025https://doi.org/10.59490/joas.2025.8252<br><br><strong>Dataset composition</strong>---The dataset consists of two main components:<br>1. Actual Take-off Weight (ATOW) Data- Source: EUROCONTROL's Network Manager augmented with airline-provided take-off weight data- Airline information is anonymized with agreements for open data sharing- Total: 527,162 flights with complete weight information<br>2. Trajectory Data- Source: OpenSky Network ADS-B state vectors- Enhanced with meteorological data from Copernicus ERA5 via the fastmeto library- Time window: 5 minutes before actual off-block time to 30 minutes after arrival<br>

本数据集为欧洲空中航行安全组织（EUROCONTROL）性能评审委员会（Performance Review Commission, PRC）2024年数据挑战赛打造，赛事旨在吸引数据科学家与航空爱好者开发开源模型，通过易得的公开数据源估算飞机起飞重量。

**概览**
本数据集收录了此前难以获取的独特飞行信息集合，整合了以下内容：
- 飞行信息（起讫机场、起降时刻、航空器型号）
- 来自OpenSky网络（OpenSky Network）的自动相关监视广播（ADS-B）轨迹数据
- 航空公司提供的实际起飞重量数据
本数据集包含2022年欧洲境内527,162架次航班，占EUROCONTROL管制空域内总航班量的6.1%。

**数据仓储DOI**
https://doi.org/10.4121/8cb8484b-dbe7-4750-8b87-a5b1dbc621b4

**研究背景**
飞机起飞重量（Aircraft Take-off Weight, ATOW）是建模燃油消耗与估算航空排放的关键参数。本数据集支持通过公开飞行数据开发开源模型以预测ATOW。

**引用说明**
若使用本仓储中的数据，请引用本数据集及以下论文：
**《飞机起飞重量估算：EUROCONTROL PRC 2024数据挑战赛》**
Enrico Spinielli; Junzi Sun; Martin Strohmeier; Xavier Olive; Quinten Goens; Rainer Koelle; Allan Tart; John Fitzgerald
《Journal of Open Aviation Science》, 2025
https://doi.org/10.59490/joas.2025.8252

**数据集构成**
本数据集包含两大核心组成部分：
1. 实际起飞重量（Aircraft Take-off Weight, ATOW）数据
- 数据来源：EUROCONTROL网络管理系统结合航空公司提供的起飞重量数据
- 航空公司信息已通过匿名化协议处理以支持开源数据共享
- 总计包含527,162架次具备完整重量信息的航班
2. 轨迹数据
- 数据来源：OpenSky网络ADS-B状态向量数据
- 通过fastmeto库整合了来自哥白尼计划ERA5（Copernicus ERA5）的气象数据
- 时间覆盖区间为实际离停机位时刻前5分钟至抵达后30分钟