Sperchios River Water Pollution Copernicus Satellite Dataset

Sperchios River Water Pollution Dataset (Sentinel-2) This dataset contains per-tile water quality products derived from Sentinel-2 imagery for the Sperchios River. Tiles are numbered from tile_00 through tile_09 and include CSV metrics, per-date JSON collections, and PNG image layers. All tiles correspond to spatial subdivisions of the Sperchios River. (Github Repository, Kaggle) Overview The Sperchios Gulf Water Pollution Dataset provides spatially and temporally resolved water-quality indicators derived from Sentinel-2 imagery. The dataset is designed to support river-to-coastal water pollution monitoring, spatio-temporal analysis, and machine-learning-based forecasting of environmental parameters. Area of Interest The dataset covers the Sperchios River River, Greece, including the river delta and adjacent coastal waters. The area is subdivided into ten spatial tiles (tile_00 to tile_09) to enable localized analysis of water-quality dynamics. Bounding box: [22.51901286489857, 38.86448942189875, 22.562631861867928, 38.8758260508686] Temporal Coverage Raw collection folders span 2016-01-01 to 2026-03-10 final_merged_data.csv spans 2016-04-19 to 2026-03-03 1D_mean_metrics_interpolated_time_based.csv spans 05-03-2016 to 03-03-2026 Data Products The dataset includes the following water-quality products derived from Sentinel-2: Chlorophyll-a concentration Turbidity Colored Dissolved Organic Matter (CDOM) Cyanobacteria density Dissolved Organic Carbon (DOC) Water color Each product is provided as: Time-series CSV metrics Image visualizations (PNG) Raw Sentinel-2 feature collections (JSON) Preprocessing Raw Sentinel-2 observations were preprocessed to: Remove invalid or missing acquisitions Aggregate per-tile statistics Interpolate missing dates to achieve a daily temporal resolution The files final_merged_data.csv.preprocessed contain the final processed time-series recommended for machine learning applications. Structure tile_00..tile_09/ csv/ : CSV summaries and preprocessed time-based metrics contains processed CSV files: 1D_mean_metrics_interpolated_time_based.csv.preprocessed 1D_mean_metrics_interpolated_time_based.csv max_metrics.csv min_metrics.csv mean_metrics.csv images/ : PNG layers for water-quality products per tile raw_s2_collections/ : per-date JSON collections (raw from Copernicus Dataspace Ecosystem) tiles_info.json : per-tile metadata (tile and acquisition info) Global Metadata metadata.json provides dataset-wide metadata Recommended Usage For machine learning and time-series modeling, we recommend using: final_merged_data.csv.preprocessed Thισ file provide consistent temporal sampling and reduced noise, suitable for forecasting and anomaly detection tasks. Citation [Wating for official publication of the paper] License This dataset is licensed under the Creative Commons Attribution 4.0 International (CC-BY 4.0) license. Full license text: https://creativecommons.org/licenses/by/4.0/ Acknowledgement This work was funded under the TERRA project, which has received funding from the European Union’s Horizon Europe research and innovation program under grant agreement No. 101189962. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Health and Digital Executive Agency (HADEA). Neither the European Union nor the granting authority can be held responsible for them.斯佩尔奇奥斯河水污染数据集（Sentinel-2） 本数据集针对斯佩尔奇奥斯河，提供基于哨兵二号（Sentinel-2）影像生成的逐瓦片水质产品。数据集共包含tile_00至tile_09共10个瓦片，每个瓦片均包含CSV格式指标、按日期划分的JSON集合以及PNG格式影像图层。所有瓦片均对应斯佩尔奇奥斯河的空间分区。（GitHub仓库、Kaggle平台发布） ## 概述 斯佩尔奇奥斯湾水污染数据集提供基于哨兵二号影像生成的时空分辨率水质指标。本数据集旨在支撑河-海岸带水污染监测、时空分析以及基于机器学习的环境参数预测任务。 ## 关注区域 本数据集覆盖希腊的斯佩尔奇奥斯河，包括河口三角洲及邻近近岸海域。该区域被划分为10个空间瓦片（tile_00至tile_09），以支持水质动态的局域化分析。 边界框（bounding box）： [22.51901286489857, 38.86448942189875, 22.562631861867928, 38.8758260508686] ## 时间覆盖范围 原始采集文件夹的时间跨度为2016年1月1日至2026年3月10日 final_merged_data.csv的时间跨度为2016年4月19日至2026年3月3日 1D_mean_metrics_interpolated_time_based.csv的时间跨度为2016年3月5日至2026年3月3日 ## 数据产品 本数据集包含以下基于哨兵二号影像生成的水质产品： - 叶绿素a浓度 - 浊度 - 有色溶解有机物（CDOM） - 蓝藻密度 - 溶解有机碳（DOC） - 水体颜色 每种产品均以下列形式提供： - 时序CSV格式指标 - 影像可视化结果（PNG格式） - 原始哨兵二号特征集合（JSON格式） ## 预处理流程 原始哨兵二号观测数据经以下预处理步骤： 1. 剔除无效或缺失的采集数据 2. 聚合各瓦片的统计指标 3. 对缺失日期进行插值，以实现日级时间分辨率 final_merged_data.csv.preprocessed文件包含经最终处理的时序数据，推荐用于机器学习相关应用。 ## 数据结构 tile_00至tile_09/ ├── csv/：CSV汇总文件与预处理时序指标，包含以下已处理CSV文件： - 1D_mean_metrics_interpolated_time_based.csv.preprocessed - 1D_mean_metrics_interpolated_time_based.csv - max_metrics.csv - min_metrics.csv - mean_metrics.csv ├── images/：各瓦片的水质产品PNG图层 ├── raw_s2_collections/：按日期划分的JSON原始集合（源自哥白尼数据空间生态系统（Copernicus Dataspace Ecosystem）） └── tiles_info.json：各瓦片元数据（包含瓦片与采集信息） ## 全局元数据 metadata.json文件提供全数据集范围的元数据。 ## 推荐使用方式 针对机器学习与时序建模任务，我们推荐使用final_merged_data.csv.preprocessed。 该文件具备一致的时序采样特性且噪声更低，适用于预测与异常检测任务。 ## 引用方式 [待论文正式发表后补充] ## 授权许可 本数据集采用知识共享署名4.0国际通用许可（Creative Commons Attribution 4.0 International，CC-BY 4.0）进行授权。 完整许可文本：https://creativecommons.org/licenses/by/4.0/ ## 致谢 本研究受TERRA项目资助，该项目获得欧盟地平线欧洲研究与创新计划的资助，项目编号为101189962。本文件所表达的观点仅代表作者本人，未必代表欧盟或欧洲健康与数字执行局（HADEA）的立场。欧盟与资助机构不对本文件内容承担任何责任。