| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The U.S. Salinity Lab USDA-ARS 450 W Big Springs Road Riverside, CA 92507
darobinson001{at}yahoo.co.uk
Abbreviations: TDR, time domain reflectometry
The paper presented by Morgan et al. (1999) deals with the question of obtaining an accurate calibration between a dielectric sensor and soil volumetric water content. Since the pioneering work presented by Hoekstra and Delany (1974) and Topp et al. (1980) with time domain reflectometry, the use of dielectric sensors to estimate water content in field studies has become a common method. Broadly instruments can be split into transmission line techniques such as time domain reflectometry (TDR) and probes which measure capacitance. Basic to these techniques is the measurement of soil relative permittivity (dielectric constant) which is then related to volumetric water content.
Calibration can be seen as a single- or two-stage process. One can either calibrate between sensor output and soil volumetric water content, or the sensor output can be converted to permittivity and permittivity then related to the volumetric water content. The advantage of using a two stage approach being that once the instrument output has been converted to relative permittivity, one of the many calibrations relating soil relative permittivity to volumetric water content can be used (Topp et al., 1980; Dirksen and Dasberg, 1993; Robinson et al., 1999). The use of a single step calibration for capacitance probes has generally been used by workers interested in monitoring changes in water content using a specific instrument calibrated to a specific access tube and soil. The major disadvantage is that one cannot separate errors due to the instrument from errors due to the calibration between capacitance and soil water content. Even though capacitance probes are now manufactured to a high level of quality, calibrations between frequency and capacitance (permittivity) tend to be sensor specific because of the electrode geometry and the tolerance of components used in instrument construction (Robinson et al., 1998). These differences between instruments can be overcome by calibrating each instrument in terms of permittivity.
Morgan et al. (1999) suggest that the oscillation frequency is proportional to the soil capacitance, implying that as capacitance increases so the oscillation frequency also increases. This is not the case; frequency decreases with increased capacitance and the relationship is nonlinear as demonstrated by the equation governing the resonant frequency response (F) of a tuned circuit:
 | (1) |
Field calibration is very difficult in the best of conditions, it is often hard to obtain the desired range of water content, and volumetric sampling has its own difficulties. Though this approach has merit, it would have been better if the field calibration was used for comparison with existing permittivity–water content calibrations. Alternatively the field calibration could have been compared with a carefully obtained laboratory calibration in controlled conditions. A calibration for each soil could then have been obtained and the results compared, rather than using their approach of lumping the soils together for calibration. This may or may not be justified depending on soil bulk density, mineralogy and clay content, soil temperature, and organic matter content, none of which appear to be considered. The use of either an established calibration or a good laboratory calibration would, I'm sure, improve their estimates of water content, which is the desired result of the work.
In order to be of benefit to both users and the wider scientific community the scaled count should be replaced by relative permittivity. The authors can then compare their calibration results with previously published work, the estimated error of which is generally smaller than Morgan et al. (1999) obtained with their work. Topp et al. (1980)(Fig. 5) presented a calibration for Rubicon (sandy, mixed, frigid Entic Haplorthods) sandy loam with an error for the 0 to 0.4 water content range of 0.0089 showing that a 9% change in bulk density had no measurable effect on the measurement of permittivity. More recently, Robinson et al. (1999) published calibrations specifically for sandy soils for both TDR and an I.H. capacitance probe (Dean, 1994). The square root of the measured permittivity was related to the soil water content and bulk density of four soils with r2 values of 0.990 for the former and 0.984 for the latter instruments. The findings of most authors who have presented work on calibration in sandy soils (Topp et al., 1980; Zegelin et al., 1989; Drungil and Gish, 1989; Whalley, 1993; Gregory et al., 1995; Robinson et al., 1999) suggests there is little difference between soils. It would be interesting and constructive to determine if the calibration presented in Morgan et al. (1999) differs from these in any considerable way.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  T. J. Kelleners, R. W. O. Soppe, J. E. Ayars, and T. H. Skaggs Calibration of Capacitance Probe Sensors in a Saline Silty Clay Soil Soil Sci. Soc. Am. J., May 1, 2004; 68(3): 770 - 778. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
