6.4.3: Soil Water Measurement Method

An alternative, or supplement, to the checkbook accounting method is to measure soil water directly for irrigation scheduling. In concept, it is quite simple. Rather than predicting or calculating SWD, the SWD is inferred from measures of f_r, f_d, θ_m, θ_v, or soil water tension. Once SWD or AW is determined, then Equations 6.10 or 6.11 and 6.16 or 6.17 are used to calculate the LD and ED for the location where the measurements were taken. The soil water content must be measured throughout the entire plant root zone. Samples or measurements at 1-foot intervals are usually adequate. If a 3-foot root zone is to be sampled, then sensors could be placed at 6, 18, and 30 inches, respectively, and each sensor would represent a 1-foot interval.

Techniques such as feel and appearance, gravimetric sampling, neutron scattering, and TDR measure water content directly (Chapter 2). Water contents can be used in the LD and ED calculations, just as was done by checkbook accounting in Example 6.5. When soil water potential (soil water tension) is measured, such as with tensiometers, granular matrix sensors, or electrical resistance blocks, a soil water release curve is needed to convert tension to volumetric water content. This is essentially a local calibration. The soil water release curve is not easily determined. Land-grant universities and government agencies, such as the Natural Resources Conservation Service, can provide release curves that represent the soils in question. An example of data used for converting tension to SWD for general soil texture classifications in Nebraska is shown in Table 6.5. More detailed data are provided by Irmak et al. (2016) and Melvin and Martin (2018). The use of soil water sensing to schedule irrigations is illustrated in Example 6.7.

Table 6.5. Example SWD versus tension for selected soil textures in Nebraska.

Tension (cb)	Fine Sand Fraction Depleted (in/in)	Loamy Sand Fraction Depleted (in/in)	Sandy Loam Fraction Depleted (in/in)	Fine Sandy Loam Fraction Depleted (in/in)
0	0.000	0.000	0.000	0.000
10	0.008	0.000	0.000	0.000
20	0.025	0.025	0.025	0.017
30	0.042	0.033	0.042	0.042
40	0.050	0.042	0.050	0.058
50	0.054	0.050	0.058	0.067
60	0.058	0.058	0.067	0.083
70	0.067	0.067	0.071	0.092
80	=	=	0.075	0.100
AWC (in/in)	0.083	0.092	0.117	0.150

Example 6.7

Assume a lettuce field with a root zone depth of 12 in. The soil is a silt loam with an AWC of 0.17 in/in. Assume f_dc = 0.50. The feel and appearance method was used to measure soil water. The fr was 0.60 in Location 1 and 0.80 in Location 2. ET (forecasted) is 0.2 in/d. The d_e = 0.5 in and r_a = 0.3 in. Determine the LD and ED dates for the two locations that were sampled.

Given: ET = 0.2 in/d AWC = 0.17 in/in

R_d = 12 in d_e = 0.5 in

r_a = 0.3 in

Find: LD and ED

Solution

Solution: TAW = (R_d)(AWC) (Equation 6.2)

TAW = (12 in)(0.17 in/in) = 2.0 in

AD = f_dc(TAW) (Equation 6.4)

AD = (0.50)(2.0 in) = 1.0 in

Calculations for Location 1:

f_d = 1 – f_r (Equation 2.10a)

f_d = 1 – 0.60 = 0.40

SWD = f_d(TAW) (Equation 6.13)

SWD = (0.4)(2.0 in) = 0.8 in

\(LD = \dfrac{AD-SWD}{ET_f}\) (Equation 6.10)

\(LD = \dfrac{1.0 \text{ in}-0.8 \text{ in}}{0.2 \text{ in/d}}=1 \text{ d}\)

\(ED = \dfrac{r_a+d_e-SWD}{ET_f}\)

\(ED =\dfrac{0.3 \text{ in}+0.5\text{ in}-0.8\text{ in}}{0.2\text{ in/d}}=0 \text{ d}\)

An alternative to converting soil water tension to water content is to monitor the soil water sensors frequently and irrigate when the soil water tension has reached a “threshold level” (Irmak et al, 2016). In fact, manufacturers of soil water sensing equipment often provide the users with guidelines for these threshold levels for various crops and soil textures. They are usually based on sensing near the vertical center of the root zone. Example thresholds are given in Table 6.6. One problem with this approach is that it is difficult to predict ahead to determine the LD. This is overcome by more frequent monitoring. Graphical extrapolation, as shown in Figure 6.13, can be used to lessen the frequency of monitoring. The graphical method provides a good visual record of soil water variations during the season. A limitation of the threshold level method is that the irrigator does not know how much water the soil can hold during each irrigation.

Table 6.6. Example threshold soil water tensions for irrigation scheduling based on f_dc = 0.35.

Texture	Threshold Tension (cb)
Fine sand	20
Loamy sand	25
Sandy loam	35
Fine sandy loam	45
Silt loam	80
Clay loam	80

Figure 6.13. Graphical method of predicted date of irrigation.

A list of questions to consider when selecting a soil water monitoring system, including sensors, communications, and data storage, has been provided by ITRC (2019).

An important, and often frustrating, consideration is the number of locations that must be sampled to reliably estimate the average soil water condition within the area of interest. You must not only consider the spatial variability of the soil itself, but also the spatial variability of water application from the irrigation system. A minimum of four locations should be sampled in a large, irrigated area that has “relatively” uniform soils and slopes. It is often good to sample stress-prone areas (low AWC and/or shallower root zone), areas where infiltration is low (steeper slopes, etc.), areas where water applications are low due to the inherent nature of the system (e.g., the downstream end of furrow irrigated fields), or where ET is the highest (e.g., nonshaded and wind exposed areas within a landscape). Do this only if the stress-prone area represents a “significant” portion of the irrigated area. Using checkbook accounting in conjunction with the soil water sensing method reduces the number of locations that must be sampled. Another method, which can greatly reduce the uncertainty of θ_v data from sensors, is to monitor and manage trends in SWD instead of θ_v or AW (Singh et al., 2020). In this method, SWD and AD should be calculated using the observational FC (FC_obs) for the specific sensor and location, which is determined from the trend in the sensor data when the field is approaching FC conditions (e.g., at the beginning of the season after the profile becomes saturated and ET is low). In this way, much of the uncertainty from the sensor and spatial variability cancels out during the calculation:

SWD = (θ_FC,obs – θ_v)R_d