Calibration of an FDR soil moisture sensor accounting for organic matter and andic properties in Jeju volcanic ash soils

Dielectric sensors are calibrated using soils of various types, but default factory calibration functions may not be valid under all conditions.Soils with high soil organic matter (SOM) content and volcanic ash soils often require soil-specific calibration to achieve acceptable accuracy.This study was aimed to developed and evaluated practical calibration methods for the TEROS 12 FDR (frequency domain reflectometry) soil moisture sensor using 39 soil samples from Jeju Island, covering a wide range of SOM (10 -300 g kg⁻¹) and contrasting andic properties.During controlled drying of repacked soils in containers at field bulk density, raw sensor counts were measured twice daily and paired with gravimetrically determined reference volumetric water content (θ ref ).We compared the factory calibration equation (FCE) with the OM-group (OMCE) and integrated (ICE) calibration equations using bias and soil-level RMSE.The FCE exhibited moisture-dependent bias, overestimating θ ref under dry conditions and underestimating at medium to high moisture, with greater variability in high SOM soils and Andisols.Across all soils, RMSE decreased from 0.067 m³ m -3 (FCE) to 0.048 m 3 m -3 (OMCE; 28.0% reduction) and 0.047 m 3 m -3 (ICE; 29.8% reduction).OMCE performed best at low moisture, whereas ICE most effectively reduced errors in Andisols under high moisture.In Jeju volcanic soils, OMCE provides a practical option when SOM information is available, whereas ICE offers a robust general calibration for soils with mixed andic properties, improving the reliability of TEROS 12-based moisture monitoring.