Bridging Synthetic Modeling and Field Reality: Assessing Dry-Region Dominance in Cosmic-Ray Neutron Sensing via Geophysical Integration

The accurate quantification of field-scale volumetric water content (VWC) is a critical requirement across multiple disciplines, from optimizing irrigation in precision agriculture to assessing slope stability and managing regional water resources. Cosmic-Ray Neutron Sensing (CRNS) is a pivotal non-invasive technology, providing integrated VWC estimates over large footprints (10–20 hectares) and significant depths (up to 80 cm). However, the interpretation of CRNS data in heterogeneous environments remains challenging. The inherently non-linear relationship between neutron intensity and hydrogen content, combined with a complex spatial weighting function, leads to "dry-region dominance," where the sensor response is disproportionately influenced by the drier portions of the soil. This research investigates these effects through a multidisciplinary workflow that integrates CRNS monitoring with preliminary geophysical spatial characterization. The first stage involved a purely synthetic investigation using the URANOS Monte Carlo neutron transport code to replicate the subsurface heterogeneity of the Borgo Grignanello site (Siena, Italy). To ensure a controlled and quantifiable comparison, the site was represented through a simplified two-region ground model characterized by distinct VWC values, constrained by several high-resolution Electrical Resistivity Tomography (ERT) transects and Electromagnetic Induction (EMI) data. This simplified framework provided a robust "forward model" and numerical proof of the dry-region bias: the derived VWC in the heterogeneous domain demonstrated an agreement with RMSE of 1.01% with the values of the drier region.

To provide empirical evidence for these synthetic findings, the second part of the research compares real CRNS time series with local TDR sensors during selected infiltration events. Given that the local sensors are positioned within the wetter units of the site, a significant incongruence between the two datasets is observed. This discrepancy serves as a direct experimental validation of the dry-region dominance predicted by the forward model, confirming that the CRNS signal is governed by the drier soil components, which effectively overshadow the moisture values of the wetter units in such heterogeneous contexts.

In conclusion, this work demonstrates that a multidisciplinary geophysical strategy is key to a more accurate interpretation of CRNS datasets. By integrating synthetic modeling with prior site characterization, this framework provides the reliable, spatially-aware insights necessary for effective hydrological modeling, natural hazard mitigation, and sustainable land management