Precision of Evapotranspiration Estimates Using Neutron Probe

A simple hydrologic balance model based on neutron‐probe measurements of soil moisture content was used to estimate evapotranspiration for alfalfa. The model results indicated large variations in the amount of estimated daily evapotranspiration for individual neutron‐probe access tubes, although field and crop conditions were the same. To identify the sources of variation in the depth of stored soil moisture that would affect an evapotranspiration estimate an analysis was performed. The variation in the amount of daily evapotranspiration was associated with a calibration component and an instrument component, both of which contributed to variance in probe readings at a given time and location. An experiment was conducted in the field to quantify the expected variation in local soil moisture measurements. From this experiment, procedures for reducing the variation due to local measurement errors were derived. The instrument component of the variance in the depth of stored water over the soil profile was linearly reduced by decreasing the soil layer thickness, e.g., decreasing the vertical distance between successive neutron probe readings. The instrument component of the standard error in the depth of stored water was reduced by increasing the number of neutron‐probe measurements per soil layer. A method was developed to quantify the minimum number of days between soil moisture measurements to achieve a specified error criterion.