Pitoro et al. (2025) Dataset of Calibration of Low-cost Soil Moisture Sensors based on IoT approach

Identification

• Journal: Mendeley Data
• Year: 2025
• Date: 2025-12-08
• Authors: Pitoro, Valdemiro, Fernandes, Caio, Sánchez-Román , Rodrigo
• DOI: 10.17632/bwdy4z47c4

Research Groups

• Universidade Estadual Paulista Julio de Mesquita Filho, Faculdade de Ciencias Agronomicas Campus de Botucatu
• Universidade Lurio

Short Summary

This paper presents a dataset for the empirical calibration and performance comparison of two low-cost soil moisture sensors (TDR and capacitive) against gravimetric moisture, collected using a custom IoT-based data acquisition unit.

Objective

• To provide a fundamental dataset for the empirical calibration and performance comparison of low-cost soil moisture sensors, crucial for efficient irrigation management in precision agriculture.

Study Configuration

• Spatial Scale: Meteorological Station of the Faculdade de Ciencias Agronomicas (FCA)–UNESP, Botucatu, Brazil.
• Temporal Scale: Data collected continuously from November 21 to December 21, 2025, with readings acquired every 5 minutes.

Methodology and Data

• Models used: Not applicable; the study focuses on data acquisition for sensor calibration. A custom 'Do It Yourself' (DIY) Data Acquisition Unit (DAU) based on IoT architecture (Arduino Nano microcontroller for sensor reading, ESP32 for communication and cloud data transmission) was developed and employed.
• Data sources:
  • Direct measurements from two commercial low-cost soil moisture sensors: 3001-TH (Time Domain Reflectometry type) and SKU: CE09640 (capacitive type).
  • Reference gravimetric soil moisture (percentage) determined by standard laboratory methods using a precision balance for soil sample mass.
  • Environmental measurements: Ambient temperature (degrees Celsius) measured by a DS18B20 sensor, and soil temperature (degrees Celsius) measured by the 3001-TH sensor.
  • Sensor raw outputs: Output voltage (volts) from the 3001-TH sensor, and analog reading (microcontroller reading unit) from the SKU: CE09640 capacitive sensor.
  • Soil sample mass (kilograms) measured using a precision balance.
  • The dataset comprises 13 essential variables, recorded every 5 minutes, totaling 3744 records per day.

Main Results

• A comprehensive dataset for low-cost soil moisture sensor calibration was generated, containing high-resolution measurements over a one-month period.
• Initial exploratory analysis confirmed the sensitivity of both tested sensors (3001-TH and SKU: CE09640) to variations in soil moisture and local climatic conditions.
• The collected data quality was attested as suitable for modeling purposes, providing a robust foundation for future calibration and performance studies of low-cost sensors.

Contributions

• Development and deployment of an innovative IoT-based DIY Data Acquisition Unit designed to overcome limitations in continuous and simultaneous collection of raw sensor readings and gravimetric soil moisture data.
• Creation of a unique, high-resolution dataset specifically for the empirical calibration and performance comparison of widely used low-cost soil moisture sensors, addressing the common lack of factory calibration.
• Provision of a fundamental resource for precision agriculture research, enabling advancements in efficient irrigation management through improved sensor accuracy and reliability.

Funding

• National Council for Scientific and Technological Development (Brazil)
• Ministry of Science, Technology and Innovation (Brazil)

Citation

@article{Pitoro2025Dataset,
  author = {Pitoro, Valdemiro and Fernandes, Caio and Sánchez-Román , Rodrigo},
  title = {Dataset of Calibration of Low-cost Soil Moisture Sensors based on IoT approach},
  journal = {Mendeley Data},
  year = {2025},
  doi = {10.17632/bwdy4z47c4},
  url = {https://doi.org/10.17632/bwdy4z47c4}
}