Geophysically-informed modelling and interpretation of cosmic-ray neutron sensing response to soil moisture spatiotemporal dynamics

Soil moisture and its dynamics have direct hydrological and ecological impacts, from the local field scale to the regional and global scales. Working on a gently-sloping vineyard with grassed interrows, this study combines soil moisture monitoring from cosmic-ray neutron sensing (CRNS) and soil sensors, geophysical characterizations, and CRNS numerical simulations. Multi-scale electrical resistivity tomography and frequency-domain electromagnetic mapping highlight a fingering transition from a clayey substrate to a fractured-karstic marl, both overlain by a more worked and homogeneous soil. Together with ancillary weather data, this spatial information and the associated infiltration dynamics explain how CRNS and nearby soil sensors respond differently to precipitation, depending on season, soil conditions, and rainfall characteristics. This highlights the critical interplay between spatiotemporal soil moisture dynamics and CRNS signal nonlinearity and nonuniqueness. CRNS numerical modelling further investigates this interplay, specifically in terms of lateral dominance of drier regions and amplified sensitivity to moisture changes at the soil surface, here induced by light rainfall events. A theoretical generalization of the dry-dominance is provided, based on Jensen's inequality and a third-order Taylor expansion of the CRNS response function. Together, field observations and numerical modelling inform a conceptual soil hydrology model that describes the observed soil moisture spatiotemporal dynamics. The results reinforce the value of more spatially informed, hydrology-based CRNS applications, a key step toward capitalizing on recent CRNS methodological advances and achieving a better integration with soil moisture networks and multi-methodological hydrological observatories.