Capturing Field‐Scale Soil Moisture Dynamics in Ireland Using Cosmic‐Ray Neutron Sensing

Cosmic‐Ray Neutron Sensing (CRNS) enables non‐invasive monitoring of field‐scale soil moisture, bridging the spatial gap between point‐scale sensors such as Time Domain Reflectometry (TDR) and coarse‐resolution satellite products. Its reliability, however, depends on rigorous atmospheric correction, site‐specific calibration, and proper scale harmonisation. This study evaluates CRNS performance at four temperate maritime grassland sites in Ireland and conducts a controlled comparison between the classical N 0 calibration framework and the physics‐based Unified Transport Solution (UTS). Additionally, weighted TDR observations were used as a spatially and vertically harmonised reference. After full correction, CRNS‐derived volumetric moisture content (VMC) showed strong agreement with weighted TDR across sites ( R 2 = 0.78–0.92), capturing seasonal wetting–drying cycles and event‐scale infiltration dynamics. Bound hydrogen pools contributed 15%–25% of the total neutron signal, demonstrating their critical role in site‐specific calibration. The comparison between N 0 and UTS revealed that increased physical complexity does not universally improve performance; For example, UTS provided measurable gains at sites where dynamic pore‐water redistribution dominated, whereas at sites with high contributions from relatively static hydrogen pools, improvements were limited. Event‐based analysis confirmed that CRNS responses primarily reflected soil layers between 20 and 30 cm depth and integrated transient wetting signals beyond the reach of single‐depth sensors. While CRNS exhibited sensitivity to near‐surface and interception‐related hydrogen during peak wet periods, this behaviour reflects physical signal integration rather than measurement artefacts. Overall, CRNS provides a robust and transferable framework for field‐scale soil moisture monitoring in humid temperate grasslands when atmospheric corrections, hydrogen‐pool characterisation and footprint‐aware validation are consistently applied.