S. Pepin et al., TEMPERATURE-DEPENDENT MEASUREMENT ERRORS IN TIME-DOMAIN REFLECTOMETRYDETERMINATIONS OF SOIL-WATER, Soil Science Society of America journal, 59(1), 1995, pp. 38-43
With the recent development of improved time domain reflectometry (TDR
) probe design, measurement systems, and calibration procedures, it is
now possible to detect and quantify the effect of temperature on the
soil apparent dielectric constant (K-a). We investigated measurement e
rrors in K-a associated with soil temperature variations and compared
measured changes in K-a with those predicted by a dielectric mixing mo
del. After confirming the accuracy and resolution of our measurement s
ystem with a series of measurements on distilled water, we measured ch
anges in K-a with temperature for a range of soil types, including san
d, loam, and peat, at soil water contents (theta(nu)) ranging from 0.0
9 to 0.81 m(3) m(-3). The measured variation with temperature in the d
ielectric constant of distilled water (0.322 degrees C-1) was very clo
se to that reported in the literature (0.356 degrees C-1). In soils, c
hanges in K-a with temperature were highest at high water contents. Fo
r soils near saturation, the overall changes observed in K-a with temp
erature were lower than those predicted by the dielectric mixing model
by 17% for sand, 24% for loam, and 39% for peat. These results sugges
t that the temperature dependence Of the dielectric constant of water
in a soil matrix is lower than that of bulk water. Absolute water cont
ent errors increased linearly with the size of the water fraction, ran
ging from 8.75 x 10(-5) m(3) m(-3)degrees C-1 at 0.05 m(3) m(-3) soil
water content to 1.40 x 10(-3) m(3) m(-3)degrees C-1 at 0.80 m(3) m(-3
) soil water content. To obtain the highest measurement accuracy, part
icularly at higher theta(nu), we suggest that a temperature correction
of 0.001758(nu)degrees C-1 be employed.