Instruction manual

CS650 and CS655 Water Content Reflectometers
Quadratic form:
θ
v
(K
a
) = C
0
+ C
1
*K
a
+ C
2
*K
a
2
with θ
v
the volumetric water content, K
a
the bulk dielectric permittivity of the
soil, and C
n
, the calibration coefficient.
3
rd
degree polynomial form:
θ
v
(K
a
) = C
0
+ C
1
*K
a
+ C
2
*K
a
2
+ C
3
*K
a
3
with θ
v
the volumetric water content, K
a
the bulk dielectric permittivity of the
soil, and C
n
, the calibration coefficient.
Linear form:
θ
v
(K
a
) = C
0
+ C
1
*K
a
0.5
with θ
v
the volumetric water content, K
a
the bulk dielectric permittivity of the
soil, and C
n
, the calibration coefficient.
Two data points from careful measurements can be enough to derive a linear
calibration. A minimum of three data points are needed for a quadratic
calibration. With three evenly spaced water contents covering the expected
range, the middle water content data point will indicate whether a linear or
polynomial calibration equation is needed.
A minimum of four data points are required for derivation of a 3
rd
degree
polynomial. Data points should be spaced as evenly as practical over the
expected range of water content and include the wettest and driest expected
values.
8.4.3 Collecting Laboratory Data for Calibration
Water content reflectometer data needed for CS650 calibration are the CS650
permittivity reading and an independently determined volumetric water
content. From this data, the probe response to changing water content can be
described by a linear or polynomial function as described in Section 8.4.2,
User-Derived Calibration Equation
(p. 23).
Required equipment:
• CS650 connected to datalogger programmed to measure permittivity
• Cylindrical sampling devices to determine sample volume for bulk
density, such as copper tubing with diameter of ≥ 1 inch and length at
about 2 inches.
• Containers and scale to measure soil sample mass
• Oven to dry samples (microwave oven can also be used)
The calibration coefficients are derived from a curve fit of known water
content and probe permittivity output. The number of data sets needed to
24