Use of geophysical data for assessing 3D soil variation in a durum wheat field and their association with crop yield

From the perspective of Precision Agriculture, the delineation of management zones in agricultural fields to optimise the use of soil and water resources and increase farmer's profitability requires knowledge of fine-scale variability. The objective of this paper is to delineate management zones in terms of yield performance verifying the suitability of Ground Penetrating Radar (GPR) to assess spatial variation of soil properties. In a durum wheat field in southern Italy, yield data were recorded and a GPR survey carried out after harvesting. On the basis of spatial distribution of the yield, the field was split into four management zones (MZs) and the expected values and standard deviations of GPR signal amplitude data were estimated using polygon kriging for each MZ. In order to interpret the results and assess the resistivity of the soil, the total sand content and bulk electrical conductivity from an electromagnetic induction (EMI) sensor were estimated by polygon kriging. The acquisition of very fine-scale soil information with GPR allowed 3D variability in the field to be imaged. Further, the study has shown the existence of a physically-based relationship between the yield of the wheat and the structural properties of the soil detected by GPR.