Determination of site specific calibration functions for the estimation of soil moisture from measurements of cosmic-ray neutron intensity

Abstract:

The recently-developed cosmic-ray neutron intensity method measures area-average soil moisture at an intermediate scale of hectometers. Calibration has proven difficult at that scale because of spatial variability of soil water and the presence of other pools of water, such as that in vegetation, also spatially and temporally variable.

Soil moisture is determined using a standard calibration function that relates the neutron intensity to soil water, and that has been parameterized by fitting a curve to neutron intensities modelled at different soil moistures. Neutron transport was simulated using the MCNPX model in which a simple setup of bare ground and sandy homogeneous soil only composed of SiO₂ was used. The standard procedure is that only one parameter of the calibration function should be fitted, which is determined from at least one independent soil moisture calibration.

In this study, site-specific calibration functions are determined to obtain some insights on the effect of other pools of hydrogen than soil moisture. Insights will elucidate whether the calibration scheme for field sites with other major pools of hydrogen should be adapted. The calibration functions are obtained similarly to the standard calibration function, but site specific model-setups are used. We obtained calibration at field sites within HOBE - the Danish Hydrologic Observatory. The field sites represent three major land covers within the catchment; farmland, forest and heathland, and the model-setups are based on site-specific data for soil chemistry, soil organic carbon, litter layer and above- and below ground biomass.

The three models provided three different calibration functions and, additionally, they were all different from the standard calibration function. The steepness of the curve and the dynamic range of neutron intensity modeled were found to be particularly dependent on the above-ground biomass and the thickness of the litter layer. Three-to-four independent soil moisture calibrations were conducted at each field site to validate the modeled site-specific calibration functions. A satisfactory fit of modeled and measured neutron intensity was found. The calibration scheme is recommended to be modified for field sites of considerable amounts of above ground biomass and thick organic rich litter layer.