How Repeatable Is Gamma-Ray Spectrometry for Agricultural Soil Mapping? Preliminary results and future perspective.

Detecting spatial variability and delineating management zones are key steps in supporting good agricultural practices and promoting precision agriculture. In this context, proximal soil sensing techniques, such as gamma-ray spectrometry, have shown promising potential over the last decades. However, their application is not yet fully standardized, and their performance has not always been consistent across the studies conducted to date.

In this study, a novel gamma spectrometer prototype equipped with a 1-L NaI(Tl) scintillator was used to characterize soil properties spatial variability at two experimental sites in the Emilia-Romagna region (Italy). The first site is a 13-ha bi-varietal vineyard with permanent grass cover, whereas the second site is a 60-ha walnut orchard, also grass-covered. During each survey counts associated with the emissions of the radionuclides ⁴⁰K, ²³²Th, and ²³⁸U, as well as total gamma counts, were acquired using a window-analysis approach. The collected data were subsequently processed within a GIS environment and statistical analyses were performed using R software.

At the first site, different surveys were carried out, both in mobile on-the-go mode and in mobile stop-and-go mode. The on-the-go surveys exhibited satisfactory repeatability for total gamma counts, whereas the spatial patterns of individual radionuclide emissions showed limited repeatability. The delineation of homogeneous zones based on total gamma counts was consistent with soil texture analyses; however, an inverse relationship was observed compared to that commonly reported in literature, with higher gamma counts corresponding to areas with higher sand content. The stop-and-go survey yielded results consistent with both the zoning derived from the on-the-go surveys and the textural analysis but only for total gamma counts. The low variability of measurements collected at the same location appears to indicate greater precision and, consequently, higher repeatability for this acquisition mode; however, its main drawback is the loss of spatial resolution compared to the on-the-go approach.

At the second site, an on-the-go survey conducted along a transect showed strong agreement with the 1:50,000 soil map of the Emilia-Romagna region and with laboratory-based texture analyses, both in terms of spatial discontinuities and relationship between soil texture and gamma counts.

Future developments of this research will include the extension to additional study sites, the development of a soil-moisture-based correction model for spatialized gamma counts and the refinement of count-extraction methods for individual radionuclides.