Dataset of Calibration of Low-cost Soil Moisture Sensors based on IoT approach

Description

This dataset is a fundamental resource for precision agriculture research, providing the necessary foundation for the empirical calibration and performance comparison of low-cost soil moisture sensors. Pre-calibration is crucial for efficient irrigation management, as most of these devices do not come with factory calibration. For this purpose, two commercial sensors were tested: the 3001-TH (TH) model, classified as a Time Domain Reflectometry (TDR) type sensor, and the SKU: CE09640 (SKU) model, a low-cost capacitive sensor. The SKU sensor is widely recognized and utilized in various research studies, whereas the TH model has been highlighted in recent studies for its precision, although detailed information regarding its operation is less common. Data collection was conducted in an uncontrolled environment at the Meteorological Station of the FCA–UNESP, Botucatu, during the period from November 21 to December 21, 2025. The adopted methodology focused on the simultaneous and continuous acquisition of raw sensor readings in relation to the gravimetric moisture reference value (the standard method). To enable this collection, a 'Do It Yourself' (DIY) Data Acquisition Unit (DAU), based on IoT architecture, was employed. This DAU was designed to circumvent the limitations of the gravimetric method and hardware constraints: the Arduino Nano microcontroller was responsible for the precise reading of analog and digital sensors, while the ESP32 handled serial communication and data transmission to the cloud. The dataset described here contains 13 essential variables, recorded every five minutes, resulting in a total volume of 3744 records per day. An initial exploratory analysis confirmed the sensors' sensitivity to variations in soil moisture and local climatic conditions, attesting to the quality of the data for modeling purposes. The dataset is composed of the following variables, acquired using the respective low-cost instruments: • Ambient temperature (oC), measured by the DS18B20 sensor. • Gravimetric soil moisture (%). • Soil temperature (oC), measured by the 3001-TH sensor. • Output voltage (mV), returned by the 3001-TH sensor, used for gravimetric moisture estimation. • SKU sensor reading, represented by the analog value (microcontroller reading unit) of the SKU: CE09640 capacitive sensor. • Soil sample weight (g) measured using a precision balance, necessary for estimating volumetric moisture via the gravimetric method.

Institutions

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

Categories

Funders

• National Council for Scientific and Technological Development

  Ministry of Science, Technology and Innovation

  Brazil

Licence