Climate change has increased the frequency of drought and flooding, making water management challenging in open-field orchards. Real-time management of soil moisture supports decision-making for appropriate irrigation levels and prevents water stress,...
Climate change has increased the frequency of drought and flooding, making water management challenging in open-field orchards. Real-time management of soil moisture supports decision-making for appropriate irrigation levels and prevents water stress, leading to stable crop yields. Soil moisture sensors are widely used to optimize irrigation management by quantifying the available water capacity, which is defined as the difference between the field capacity (FC) and the permanent wilting point and exhibits spatiotemporal variability. In this study, a data-driven approach was introduced by analyzing time series data of Frequency Domain Reflectometry (FDR) sensor acquired in open-field orchards to dynamically estimate FC, offering practical advantages over static conventional methods. The soil water status was partitioned into increasing, saturation, and drainage phases based on the sensor-derived volumetric water content, with the classification achieving an F1-score of 0.74. Only drainage events with a duration of more than four days were selected for FC estimation. The estimates were validated against van Genuchten soil water retention curves derived from Rosetta3, consistently confirming that the estimated FC were within the conventionally assumed range of −10 to −33 kPa. Data-driven analysis of volumetric water content can provide a robust and non-destructive method for estimating site-specific FC, enabling irrigation management that dynamically responds to changing field conditions without relying on generalized soil databases.