Calibração e desempenho de sensores capacitivos de baixo custo em solos com diferentes atributos químicos

Abstract

Accurate monitoring of soil water content is essential for irrigation management efficiency and the sustainability of water resources. This study aimed to evaluate the performance and perform the calibration of low-cost capacitive sensors based on the Frequency Domain Reflectometry (FDR) technique across seven soil classes with distinct physicochemical characteristics. The experiment was conducted in a laboratory setting, using the gravimetric method as the standard reference. Samples were subjected to controlled saturation and drying (desorption) cycles, correlating the electronic output signals of the sensors (models AP1 and AP2) with the actual volumetric moisture. The results demonstrated that the generic factory calibration is inadequate for the studied edaphic conditions. The AP2 sensor showed significant operational limitations, exhibiting non-linear behavior and insensitivity in high moisture (saturation > 80%) and low moisture ranges, restricting its practical utility. The AP1 sensor presented a better linear fit and higher sensitivity; however, its precision was influenced by the soil matrix composition. It was observed that high potassium content and high base saturation increased electrical conductivity, generating noise and overestimates in the readings, while the organic matter content affected the intercept parameters of the regression equations. It is concluded that the technical feasibility of these low-cost sensors, especially the AP1 model, imperatively depends on specific calibrations that consider not only texture but also salinity and soil bulk density, ensuring the necessary reliability for irrigation management.