Assessing an empirical approach to derive SWE from CRNS for pre‑alpine to high‑alpine locations

When high‑energy cosmic rays strike the upper atmosphere, they produce cascades of secondary particles, including fast neutrons that reach the Earth's surface. These neutrons are efficiently moderated by collisions with hydrogen atoms; consequently, the intensity of the neutron flux above ground decreases in proportion to the amount of water present—whether stored in the soil, in liquid form, or frozen as snow.

A stationary cosmic-ray neutron sensing (CRNS) detector records counts of these epithermal neutrons, and a single local water‑content reference is sufficient to convert the count rate into a quantitative estimate of soil moisture. The count‑versus‑moisture relationship has been shown to be remarkably consistent across diverse soils, climates, and geographic regions.

Because the calibration curve is essentially universal, typically only a single in‑situ reference measurement is required; thereafter, and retrospectively, the detector can continuously monitor spatially integrated changes in soil moisture. This simplicity has established CRNS as a valuable tool for agricultural water management, hydrological research, and field‑scale climate monitoring.

In contrast, converting neutron counts to snow water equivalent (SWE) for a sensor positioned above the snowpack has required extensive site‑specific calibration, which has hindered rapid network expansion. This difficulty arises from discrepancies between theoretical models and the limited empirical data available.

Based on a compilation of extensive in‑situ measurements at several montane locations within the Pre‑Alpine Terrestrial Environmental Observatory (TERENO Pre‑Alpine), we derived a set of empirical coefficients for the count–SWE relationship. Most locations in our dataset show good agreement with these empirical coefficients, although some outliers exist. Nevertheless, this empirical approach can reduce the effort required to establish new CRNS stations for SWE monitoring. We also evaluate transferability to alpine–nival sites—characterized by shallow soils, steep topography, and very high SWE—and analyze causes of deviations in the empirical approach’s performance due to site-specific environmental conditions.