Adamo et al. (2026) Soil Moisture, Irrigation Actuator and Weather Dataset from a Multi-Sector Precision-Irrigation

Identification

• Journal: Mendeley Data
• Year: 2026
• Date: 2026-09-14
• Authors: Tommaso Adamo, Lucio Colizzi, Giovanni Dimauro, Emanuela Guerriero, Nunzia Lomonte
• DOI: 10.17632/c837v6p8ph.2

Research Groups

• University of Bari Aldo Moro (Bari, Apulia, Italy)
• University of Salento (Lecce, Apulia, Italy)

Short Summary

This work provides a comprehensive IoT dataset from a precision-irrigation farm in Italy, integrating soil sensor readings, irrigation actuator telemetry, and weather data across five different crop sectors.

Objective

• To provide a high-resolution, validated dataset that links soil moisture dynamics with actual irrigation actuator states and weather forcing to support the development of precision irrigation and machine learning models.

Study Configuration

• Spatial Scale: Local farm level (Arnesano, Lecce, Italy), encompassing five cultivated sectors: two open-field tomato, one potted tomato, one zucchini, and one blueberry.
• Temporal Scale: One full growing season (February–September 2025), with processed data aggregated on a 10-minute grid.

Methodology and Data

• Models used: Python-based preprocessing pipeline (utilizing pandas, numpy, and requests) for data cleaning and feature engineering; ERA5 reanalysis for weather data.
• Data sources:
  • IoT sensors: Soil moisture, pH, and electrical-conductivity probes.
  • Actuators: Irrigation and fertigation valve state records.
  • Weather: ERA5 reanalysis.

Main Results

• Raw Data Collection: Captured 486,894 sensor measurements and 3,055,944 valve state records.
• Processed Data Generation: Reconstructed and agronomically validated 789 irrigation and fertigation events.
• Analysis-Ready Outputs: Produced five per-sector time series (33,259 rows each) on a 10-minute grid, including engineered features and 24-hour-ahead soil-moisture targets.
• Software: Developed a reproducible Python pipeline to transform raw IoT exports into processed analysis tables.

Contributions

• Actuator Integration: Unlike most public datasets, this includes actuator telemetry, making applied water a measured variable rather than an assumption.
• Benchmarking Capability: By publishing the raw layer (including sensor dropouts and spikes) alongside the processed layer, the dataset allows researchers to benchmark alternative data-cleaning strategies.
• Cross-Sector Generalization: Provides data from four different crop/substrate configurations under the same weather forcing, facilitating studies on model generalizability.

Funding

• Not specified in the provided text.

Citation

@article{Adamo2026Soil,
  author = {Adamo, Tommaso and Colizzi, Lucio and Dimauro, Giovanni and Guerriero, Emanuela and Lomonte, Nunzia},
  title = {Soil Moisture, Irrigation Actuator and Weather Dataset from a Multi-Sector Precision-Irrigation},
  journal = {Mendeley Data},
  year = {2026},
  doi = {10.17632/c837v6p8ph.2},
  url = {https://doi.org/10.17632/c837v6p8ph.2}
}