Capacitance moisture probes: To calibrate or not?

Lance Pendergast and Jenelle Hare, Department of Primary Industries and Fisheries and Cotton Catchment Communities CRC
Recent efforts to improve on-farm water-use efficiencies have seen the increased use of soil moisture monitoring tools – particularly capacitance probes, such as C-probe, Enviroscan and Diviner.

Probes are typically used to:

• schedule irrigations – determining when an irrigation is required
• determine soil water deficit (to determine how much water to apply) or plant water use.

The question often arises: ´Do probes require to be specifically field calibrated´? While the answer depends on the use above, the answer for most commercial applications is usually ‘no!’ However they do need to be normalised.

Normalisation

Probes are normalised by matching the raw readings from each sensor at both 0% and 100% water levels (i.e. held in the air, and submerged in water). Without normalising, these devices would only provide a range of irrelevant raw data that varies slightly with each sensor. By matching the raw reading from each sensor to both 0% and 100% water levels, a comparison of readings taken by different sensors can be made on a common scale. Equipment suppliers should provide you with procedures for normalising their product.

Probes for irrigation scheduling

Following the normalisation process, readings can be displayed (typically graphically), permitting irrigators to monitor their soil water levels based on trend changes. This accepted practice, using a default calibration equation within a product´s software, negates the requirement for a complex site specific calibration.

In many cases the graphs employ millimetres as the vertical axis units. Herein lies a problem – the readings from probes, presented as millimetres, are actually only an estimation of volumetric soil moisture levels and therefore should not be taken literally.

In most irrigation scheduling the main aim is to identify when to apply the next irrigation. The optimum time to irrigate can be determined by observing soil water levels as they move from field capacity to the refill point. Field capacity and the refill point can be determined through relating known soil moisture conditions to probe values, and analysis of soil moisture trends.

Uncalibrated capacitance probes will show changes in soil water content over time and will easily satisfy the requirements of an irrigator seeking to schedule irrigations based on variation in plant water use. Probes do not require a calibrated unit of measurement in order to achieve this task, as the change in rate of water extraction is used instead. Scheduling irrigation based on the slope of the line is acceptable; however it must be recognised that a change in the slope of the line can be due to soil moisture availability, change in evaporative demand in response to cooler or cloudy weather, or other factors.

Probes for determining soil water

Where a more accurate unit of measurement is required, there are higher order calibration methods available to correlate the relationship between the raw data counts, actual soil water content levels and the water-holding capacity of the soil. This calibration is rarely done in a commercial agricultural situation as it is possible to schedule irrigations using soil moisture trends rather than absolute numbers.

Calibration becomes necessary when a device is to be used to determine the magnitude of plant water use, soil moisture deficit or to calculate a full soil water balance. Usually this is required in research where a precise control over the amount of irrigation water applied and/or information to calculate a soil water balance is required.

While soil moisture values will differ between a calibrated and uncalibrated probe, the pattern of the soil water trends will be similar. Figure 1 shows soil moisture data collected from a calibrated site which is then compared to the same data recalculated using the default calibration equation.

Figure 1: A single data set using a default and site specific calibration. Source: (2001) Calibration of the Sentek Pty Ltd Soil Moisture Sensors

In this figure, it is clear that the data represents the same trends, regardless of the default or site specific calibration. However if we look at the first irrigation event at point A:

• the calibrated soil moisture (red line) goes from 144mm to 173mm – a change of 29mm
• the default calibration soil moisture (black line) goes from 105mm to 146mm – a change of 41mm.

If the default calibration was used and assumed to be correct, the change in soil moisture would be overestimated by 40%.

The calibration of soil water monitoring devices and the comparison of the resulting data requires much more effort than simple irrigation scheduling. The process is both exacting and site specific as it involves the accurate assessment of a range of parameters (including gravimetric water content and bulk density) which can vary significantly between soil types. This often necessitates taking into account changes in the soil profile. WATERpak (pgs 101 – 106) and Sentek Pty Ltd provide a detailed account of the calibration procedure if required. The Scheduling II workshops (as part of the Irrigated Cotton and Grains Workshop series) also explain the various parameters required and procedures involved for those interested.

It is important to remember, particularly when not completely refilling the soil profile, that probes give a point source of data, and even calibrated data may not give an average picture of soil moisture status. Figure 2 shows the field setup of three capacitance probes placed close to each other: one in the plant line, and one on either side, only 250mm away. Water was placed in the furrow, 500mm from the plant line, as illustrated by the blue arrow. In this case, water was applied by a LEPA sock, but the same effect would be experienced from a deficit surface irrigation event.

Figure 2: Probe setup for data in Figure 3.

Figure 3 illustrates the data obtained from the three probes. The blue line indicates probe position 1 from Figure 3, with the pink line indicating probe position 2 and the yellow line indicating probe position 3. As demonstrated, the soil water response varies between probes indicating spatial variability will have an impact on the accuracy of probe data, regardless of the accuracy of calibration.

Figure 3: Even calibrated probes may not give a good picture of average soil moisture due to spatial variability.

During the irrigation indicated at point A, you can clearly see that increase in soil moisture is greatest in the pink and yellow lines, compared with the blue line. Obviously with such variations in soil moisture across even a 0.5 metre cross section, the variation over a field of many hectares could be significant.

Summary

Capacitance probes are extensively used for scheduling of irrigation events based on changes in soil water levels over time. Most irrigators with scheduling devices determine irrigation events based on graphed data depicting actual plant water use over time.

The debate over calibration should not detract from this ability to support the process of day-to-day irrigation management. It should, however, highlight the dangers in taking uncalibrated data literally.

Calibration may be undertaken to obtain more accurate soil moisture data, although the likelihood for spatial variability should be considered when contemplating the onerous task of calibration.