Soil Water Sensor Calibration and Soil Water Processes in a Farmer's Field.

The physical sphere of influence of field moisture sensors deviates from the statistical range of influence mainly affected by soil properties and landscape variability. This aspect is relevant for strategies of soil moisture sensor field calibration. The objective of this study was to test different calibration approaches and to describe space-time behavior of soil water content in a farmer’s field in Western Kentucky. For this purpose, 45 access tubes were installed at 10-m distances across a catena in various landscape positions of a field soil with varying textural and morphologic properties. Soil water content was indirectly determined using a capacitance probe at 10-cm-depth intervals between the land surface and 80 cm depth in each of the 45 profiles over a period of two years at time intervals between 8 and 21 days. At five times with different levels of soil water content, soil samples were taken for gravimetric soil water content measurement. The results were used for testing and comparing different calibration approaches of the capacitance probe. For the conversion of gravimetric into volumetric water content, soil dry bulk density was derived from soil auger samples. For this purpose, raw measurements had to be processed using kriging in combination with jack-knifing. The capacitance probe was calibrated a) specifically for each site and depth, b) separately for each of the 45 profiles, c) for each of the eight soil depths across all 45 locations, and d) stratified based on soil clay content. Site- and depth-specific calibration (356 functions) yielded the best results followed by the calibration for each individual depth across all 45 locations (eight functions). Calibration with respect to clay-based stratification of data yielded large errors and uncertainties. Temporal stability, cospectral and state-space analysis were employed to describe space-time water content processes in this field soil.