A worst-case scenario? Exploring low-energy cosmic-ray neutron signal dynamics in wetlands

In the past 15 years, Cosmic-Ray Neutron Sensing (CRNS) has evolved to a useful tool for monitoring soil moisture at the field scale. Given the large measurement radius of up to 200 metres and measurement depth of 20 to 30 centimetres, it overcomes small-scale heterogeneities and allows to estimate soil moisture at spatio-temporal scales which are required to e.g., inform environmental models or validate soil moisture products from remote sensing data.

CRNS relies on the inverse relationship between soil moisture and observed low-energy cosmic-ray neutrons. Higher soil moisture results in lower neutron intensities but also a higher statistical noise in the data. In combination with the strongly non-linear relationship between soil moisture and observed low-energy cosmic-ray neutrons, this leads to larger uncertainties for soil moisture estimates when the soil moisture is high. Therefore, CRNS is expected to provide most accurate soil moisture estimates at monitoring sites with generally drier soils. Knowledge gaps remain with respect to the use of CRNS and the response of measured neutron intensities at observation sites with very wet soils and even partial water cover.

Against this background, we explore the signal dynamics of observed thermal and epithermal neutron intensities in a wetland in north-eastern Germany. Placing two identical neutron detectors at two different locations in the wetland and with different fractions of water cover in their respective measurement footprint allows for an investigation of the sensitivity of observed neutron signals to variations in partial water cover and soil moisture changes in water-free areas. Site-specific signal dynamics are modelled using neutron transport simulations conducted with the URANOS model code as well as simplified approaches to gain understanding on the influence of water cover and soil moisture on thermal and epithermal neutron signals. Ultimately, the possibility of deriving soil moisture information in water-free areas from observed neutron intensities is explored.

Our analyses shed additional light on the potential of CRNS for soil moisture estimation and its sensitive measurement footprint at extreme and unfavourable monitoring sites.