The definitive version is available at:
http://www.sciencedirect.com/science/journal/00221694
Time Domain Reflectometry (TDR) has been widely accepted as a reliable technology for the measurement of volumetric water content (θ) in soils. Here, we present a new procedure for the graphical interpretation of TDR waveforms to determine the variation of the apparent dielectric constant (K) along a single TDR probe in non-conducting media and its application for profiling the moisture content in soils. The method is based on the influence of K on the reflection coefficient (ρ) and travel time (t) of the TDR signal along a transmission line. A ρ(t) function is initially defined for a length l of 10 cm and plotted together with the TDR waveform for a three-rod probe of length L (L=nl). The interception point of both lines defines t11 as the reference time to build a second ρ(t) line that intercepts the waveform at time t2l. By repeating this process iteratively, a series of ρ(t) lines intercepting the TDR trace at times t3 l,t4 l,… and tnl is obtained, making it possible to calculate K for apparent probe lengths equal to l,2l,3l… and nl. To test the consistency of the method, two sand column experiments were conducted with two different TDR probe geometries. A total of 144 values of K were measured during a wetting and draining cycle of the sand column with vertical 10-, 20-, 30-, and 40-cm long three-wire uncoated TDR probes (diameter d: 2.8 mm; spacing of the outer conductors s: 32 mm) and 10-, 20-, 40-, and 60-cm long coated TDR probes (d: 10 mm; s: 80 mm) using the standard double reflection waveform analysis. A satisfactory relationship (R2=0.99) was found between these K values and those obtained for the same depths by the proposed method applied to the 40- and 60-cm long probes, respectively. In this case, a probe-specific correction factor was used for the different probe geometries. Likewise, an excellent match was found between the θ profiles measured in the sand column applying the new approach to the 60-cm long probe and the θ profiles measured with horizontal probes placed at 10, 20, 30 and 50 cm depth. The results show that the proposed method is sound and suitable for determining the variation of K at fixed intervals along a single probe and therefore for soil water content profiling.
This work was supported by the Comisión Interministerial de Ciencia y Tecnología of Spain (grants AGF98-0261-CO2-02 and AGL2001-2238-CO2-01).
Peer reviewed