Abstract
The application of time domain reflectometry (TDR) principles to determine soil water content has been widely accepted as an alternative method since Topp et al. (1980) proposed an empirical relationship between the dielectric constant of soil and soil moisture content. However, a wide range of studies have made it clear that the proposed function has only restricted validity. Although the response of the TDR signal shows a steady correlation with soil moisture, it is also affected by others factors related to the physical and chemical nature of soil and the solute characteristics of soil solution. It is therefore necessary to determine a particular calibration for each situation. Three methods are used to calibrate a TDR sensor. First, two well‐known methods are compared: non‐continuous wetting on one hand, and continuous wetting by dripping of a soil sample on the other. The third consists of the progressive introduction of a sensor in a soil sample with a known moisture content. The three methods all lead to similar consistent relationships. The methods are validated using field data taking into account spatial variation of measurement at a given date and water inputs and uptakes under an irrigated corn crop during a growing season. Copyright © 2003 John Wiley & Sons, Ltd.