Near-ground microwave radiometry for on-the-go surface soil moisture sensing in micro-irrigated orchards in California

High-resolution geospatial soil moisture measurements are needed to inform hydrological modeling and to guide water management in agriculture, especially in highly heterogeneous systems such as micro-irrigated orchards. In this research, we used a Portable L-band Radiometer (PoLRa) to map very high-resolution (<2 m) soil surface moisture in micro-irrigated orchards in Southern California. Almond (Prunus dulcis Mill.), olive (Olea europaea L.), and orange (Citrus × sinensis Osbeck) orchards grown on Monserate sandy-loam soil were surveyed from the Summer through the Fall of 2022. The sensor was mounted on an all-terrain vehicle and paired with a centimeter-level positioning system. PoLRa measurements were compared with ground-truth volumetric water content determined from soil cores collected at the study sites. The sensor data were calibrated to estimate surface soil moisture with an analysis of covariance linear regression approach. The lowest estimation errors were observed in the almond orchard, which had flat soil and no canopy interference. There, the root mean square error of the tested linear models ranged between 3.9% and 4.1%. Over the entire dataset, the root mean square error was 5.9%. This new sensor technology may be a means for improving understanding of water dynamics in complex and heterogeneous agricultural systems. Nevertheless, further research is needed to refine calibration models and address environmental variability and its effects on the sensor's measurements.