Short, Multineedle Frequency Domain Reflectometry Sensor Suitable for Measuring Soil Water Content

Abstract

Time domain reflectometry (TDR) is a well‐established electromagnetic technique used to measure soil water content. Time domain reflectometry sensors have been combined with heat pulse sensors to produce thermo‐TDR sensors. Thermo‐TDR sensors are restricted to having relatively short needles to accurately measure soil thermal properties. Short needle lengths, however, can limit the accuracy of the TDR measurement of soil water content. Frequency domain reflectometry (FDR) sensors are an alternative to TDR sensors that can provide an inexpensive measurement of soil water content. The objective of this study was to determine whether short FDR sensors can accurately measure soil water content. We designed and constructed a short FDR sensor. For four soil types across a range of water contents, temperatures, and salt contents, we measured soil dielectric spectra with the short FDR sensor. A vector network analyzer was used to obtain soil dielectric spectra in the 1‐MHz to 3‐GHz frequency range. The ideal frequency of a short FDR sensor is the frequency at which the permittivity is not altered by changing temperature or salt content. The 47‐ to 200‐MHz range was an ideal frequency range for measuring soil water content, and 70 MHz was the frequency least influenced by temperature and salt content. The short FDR sensor provided quick, continuous, stable, and cheap measurements of soil water content. Because of the promising performance of the short thermo‐FDR sensor in laboratory studies, sensors should be evaluated in future field studies.

Development and Application of the Heat Pulse Method for Soil Physical Measurements

“…The short physical probe length of a thermo‐TDR sensor also limits the accuracy of soil water determination. A possible alternative is to use a thermo‐FDR (Sheng et al, ), because capacitance or impedance sensors (FDR) allow more flexibility in electrode configuration with reduced measurement complexity compared to TDR, in addition to lower costs of the electronics system and minimal postprocessing (Sheng et al, ; Xu et al, ; Zent et al, , ). Sheng et al () coupled an electromagnetic sensor to determine water content with a penta‐needle thermo‐FDR, which demonstrated a significant improvement in soil water content determination (with RMSE = 0.012 cm 3 /cm 3 compared to RMSE = 0.042 cm 3 /cm 3 of the HP methods).…”

Measurement of Soil Water Content with Dielectric Dispersion Frequency

“…A third-order polynomial equation fit the data well: indicating that the dielectric dispersion frequency may be used to estimate soil water contents for different soils at different temperatures. The root mean square error (RMSE) of the q estimated by the f d measurement is 0.021, which is better than 0.027, the RMSE of q estimated at 70 MHz (Xu et al, 2012) and better than the RMSE of f d = 0.0288, which is estimated by the Topp equation.…”

“…From the complex permittivity spectra of the soils, we can determine the ideal measurement frequency at which the apparent permittivity of the four soils was relatively constant for a range of texture and temperatures. Xu et al (2012) reported that 70 MHz was the best measurement frequency at which the apparent permittivity of four select soils did not change for a range of temperatures and salinities. It should be noted that fixed-frequency methods have limitations for soil water content measurements fig.…”

“…The open standard produced an open trace on the Smith chart. We started measurements of the soil 2 h after the VNA was powered on to ensure the stability of the instruments (Xu et al, 2012). The scattering parameter (S 11 ) was obtained from the VNA and converted to permittivity using the procedure of Logsdon and Laird (2002), who based their procedure on that of Campbell (1990) and Kraft (1987).…”

“…Many recent scientific research efforts have addressed the development of durable q measurement systems that are accurate, offer fast response times, are nondestructive, and are relatively low cost (Noborio, 2001;Robinson et al, 2003;Xu et al, 2012).…”