Calibration Equation of a Capacitance Soil Water Probe and the Estimation of Soil Water Content.

The estimation of soil water content in a soil using a capacitance type soil probe can be improved by calibrating the probes for that particular soil as opposed to using the factory calibration. The sandy soils (>95% fine sand) in the Central Florida Ridge holds a narrow margin of soil water (about 0.06 m³ m^-3) as available for plant use.  Therefore, the accuracy and precision of soil water estimates are important in irrigation scheduling in this soil.  The sensor output in raw count when plotted against the measured volumetric soil water content obtained in the laboratory, it yielded a linear calibration.  The calibration equation obtained during this calibration using Excel and SAS was a little different when all the decimals were accounted but appeared to be identical when the number of decimals was reduced to an agreeable level.  The calibration equation with Excel was y= 0.0007x -0.3639 and with SAS was y = 0.0006598x – 0.35761.  When the number of decimals in these two equations was reduced, the equation can be written as 0.0007x – 0.36.  However, when the estimates of soil water using this common equation, Excel generated, and the SAS generated were compared, it was revealed that only the SAS generated equation estimated the soil water content to an agreeable level.  The objective of this presentation is to demonstrate that SAS generated equation is better in estimation of water content and that the number of decimals appear in a calibration equation also plays an important part in the estimation.