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.