Abstract
The use of accurate and near real time volumetric soil moisture (θv, m3 m−3) measurements informs not only precise irrigation scheduling but also decisions regarding fertiliser applications, feed supply and stock management in non-irrigated farming systems. To assist θv monitoring, capacitance-based AquaCheck (AquaCheck, South Africa) sensors are being increasingly used in precision agriculture applications. However, the performance of these sensors have not been investigated extensively under field conditions with dynamic soil moisture regimes.For evaluation, 20 probes (4 sensors/unit) were installed in predominantly silt loam soils on a pastoral hill country farm in the southern east coast of the North Island of New Zealand. Raw readings, and θv derived from a manufacturer-recommended calibration, were compared with 400 reference θv values over a one-year timespan (01/11/2016 – 01/12/2017). In addition, three custom calibration methods were developed and assessed, each tailored to suit the availability of soil information.Over all depths (100 to 400 mm), the manufacturer calibration resulted in θv readings with a mean RMSE of 0.106 m3 m−3, a mean MBE of -0.099 m3 m−3 (indicating underestimation), and a coefficient of determination (R2) of 0.58 when correlated to reference θv values. A single custom formula, relevant to the local soils resulted in an improved RMSE of 0.039 m3 m−3 and R2 = 0.58, while probe-specific calibrations achieved an RMSE of 0.029 m3 m−3 and R2 of 0.77. The application of sensor-specific calibration resulted in an RMSE of 0.019 m3 m−3 with R2 = 0.9. Temporal sensor responses to the magnitude of rainfall events closely agreed. Sensor performance and accuracy errors were observed to vary as a function of soil wetness, bulk density (ρb gcm-3) clay and total organic carbon (TOC) content. Increasing clay content and ρb resulted in increased raw output for a given θv whereas TOC demonstrated the opposite trend. These effects were significant (P value < 0.001) but eliminated by the sensor-specific custom calibration.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call