Uncertainties in the measurement of soil water content caused by abrupt soil layer changes, when using a neutron probe

Multigroup diffusion code calculations have been made to model the changes in the thermal flux distribution as a fast neutron source passes through an interface between two soil systems. By integrating this flux over the volume of a detector attached to the source, a relative count rate can be determined for different source positions with respect to the interface. Water densities determined from these count rates and calibration curves can be seriously in error near the interface. The magnitude of these errors depends upon the soil parameters, the detector geometry, and the proximity of the soil strata to the neutron source and detector. Unless the soil structure is well known, it would be unwise to rely on water measurements closer than about 25 cm to a soil discontinuity. It two soils merge over 30-40 cm, it is likely that the measured water density will not deviate far from the real value.