Calibration of Arduino Sensors for Soil Moisture Monitoring in Sandy Soils

Abstract

Soil moisture is one of the important parameters in sandy soil, which is known for low water retention when considering plant growth, effective water management, and soil health. The current study investigates the readings of Arduino-based soil moisture sensors and compares them with those of a commercial GS3 sensor from Decagon. In controlled conditions, using sandy soil from Rhu Tapai, Terengganu, was studied. A 32-day drying period was conducted. A third-order polynomial equation was used to correlate with the commercial sensor. The calibration model's accuracy was high, with R-squared values between 0.995 and 0.998, enabling conversion of Arduino sensor readings to volumetric soil moisture content (m³/m³). The results imply that inexpensive Arduino-based sensing can be used alongside, or even as an alternative to, the commercial sensor. Hence, the calibration method on Arduino-based sensors for monitoring soil conditions in sandy areas can be achieved. Further implications of the study show that a step toward integration into real-time monitoring, as in the Internet of Things (IoT), is achievable. The Arduino sensor readings have been demonstrated as an extension for estimating soil air porosity and wet bulk density, which are applicable to agricultural, engineering, and environmental monitoring. This research highlights the economic viability and adaptability of Arduino sensors for soil moisture monitoring, with practical applications in the sustainable management of water resources and in agricultural soil water monitoring.