Principles:

The solute movement experiment simultaneously, uses the principles of Time Domain Reflectometry (TDR) and the Tension Disc Permeameter , to measure the infiltration rate of solute at specific depths within a soil.

Methods:

A small hole was dug within the pasture sub-site and, four TDR wave guides were inserted at depths; 1cm, 5cms, 9cms, 13cms and 17cms, respectively.

Both the tension disk and TDR experiments were set up to work simultaneously, with the tension disk resting directly above the waveguides.

Time began with the first bubble of the permeameter. Measurements for this were taken at 15 – 30 second intervals, for the first 5 minutes and then at 1 minute intervals thereafter, for a total of 31 minutes. Fresh water was used to fill the permeameter initially, with the solute added after 6 minutes. At the 10 minute and 30 second mark, fresh water was reapplied and from that time onwards, until the experiment ceased.

TDR measurements were taken cyclically from the shallow to deep, with time stated at each measurement. This process continued for a total of 54 minutes, with an estimated time delay of around 25 seconds between each waveguide reading.

Inserting a waveguide.

Wave guides inserted at depth within pasture.

Experiment in progress (example shown is without solute).

Disk permeameter set up above, buried waveguides in figure 3.

Experiment in progress, reading permeameter (example shown is without solute).

Analysis:

Data were recorded from the ponded disc permeameter and the TDR apparatus. As stated in the description of the experimental method above, the initial water level in the permeameter reservoir was recorded at the beginning of the experiment. After infiltration began, water level and cumulative time were recorded at intervals, as described above. Water level readings allowed cumulative infiltration to be calculated. TDR readings of soil volumetric water content were made for each probe in succession from the beginning of infiltration and the cumulative time for each reading noted.

The TDR wave forms were saved for readings from just after the saline water was added to the permeameter reservoir up to the final readings.

The volumetric water content data from the TDR were analysed to produce a plot of predicted volumetric water content by depth at specified cumulative time values. The volumetric water content and cumulative time data for each probe were first separately analysed to derive a prediction formula for volumetric water content over time. This formula was obtained by splining, which was done over three sections for each probe plot, after removing outliers. The spline for the first section of the curve was constrained to ensure that the initial measured volumetric water content values for each depth were correctly given by the prediction formula. The middle section was then splined and, finally, part of the middle section and the last reading were splined. Each spline included an overlap of points from the spline on each side to improve continuity of predictions. Predictions of volumetric water content were then calculated from the prediction formulas for each probe (i.e. depth) at the time points shown on Figure 1. The final set of predictions for each probe was then compiled by choosing predictions from each of the three splined prediction formulas over the appropriate range of times. The plot obtained is shown in Figure 1.

Figure 1 Solute movement

Results and Discussion:

The splined volumetric water content measurements produce infiltration curves shown in Figure 1. These smoothed curves show the general pattern of infiltration over time.

Volumetric water content of the upper part of the profile increases up to about 500 s. The lower part of the profile also increases but lags behind the upper part. Based on a measured mean bulk density for the pasture topsoil of 1.08 g/cu. cm, the porosity of this soil is approximately 0.59 and so, after 500 s, the upper part of the profile is approaching saturation. After about 1000 s, the profile is fairly uniformly wet throughout. It then begins to drain, with the upper part showing more draining more than the lower part and between 5000 s and 10000 s stabilises at around 40% volumetric water content. At this point, water is being held against gravity. The profile will be expected to drain slightly more over the following 1 to 2 days before reaching field capacity. The increased infiltration at a depth of 9 cm is due to a localised layering effect in the profile, with a more porous layer or macropore at that depth.

The strength of the transmitted and reflected TDR pulse can be used to calculate a quantity, Vr, which measures the attenuation of the signal:

where V0 is initial signal voltage, Vf is voltage of the reflected signal when first detected and Vi is minimum signal voltage. These three values can be determined from the TDR wave form. The greater the signal attenuation, Vr, the greater the salt concentration in the soil water.

Figure 2 below shows signal attenuation, vr, over time at each probe depth. The 1 cm depth probe failed between 420 s and 900 s, so the peak for a depth of 1 cm is too far to the right. The plots for depths of 5 cm, 9 cm and 13 cm show peaks after approximately 730 s, 830 s and 950 s respectively. These peaks indicate the passage of the pulse of saline solution through those depths. These peaks can be compared with the raw TDR volumetric water content readings shown in Figure 3. There are corresponding peaks in the plot in Figure 3 at 5 cm and 9 cm, although for the latter the TDR reading peak occurs slightly earlier. At 13 cm, the TDR peak is slight and at 17 cm no peak is discernable.

Figure 2   Signal attenuation over time

Figure 3   Raw TDR volumetric water content readings

The two plots show the effect on TDR volumetric water content readings of a saline soil solution. The increased electrical conductivity of the soil solution caused by the added saline water causes the dielectric constant of the soil solution to increase. Hence, the TDR reading, calculated from the equation of volumetric water content as a function of the dielectric constant, given in the ‘Time domain reflectometry’ section, increases in response to the saline soil solution. This reading gives the combined effect of the increased volumetric water content itself and increased salinity of the soil solution.

From the peaks discussed above on Figure 2, the speed at which the saline pulse moves downwards through the profile can be estimated and is approximately 2.2 cm/min.

Information on solute movement through soil is useful in agriculture in relation to liquid fertiliser application. Knowing the rate at which fertiliser infiltrates can be helpful in timing applications depending on the age and hence root length of plants. Such information is also useful in determining the effect on fertiliser movement through the profile of rain after application. In the context of chemical spills onto soil or contamination of soil by liquids, such information is also clearly very useful.

This experiment is advantageous as it determines the rate of wetting front movement within a soil when solutes are present and at possibly, influential concentrations. This process is desirable to irrigators whom are using salty water or have salt effected land and therefore wish to determine the amount of water being lost to groundwater as recharge or more significantly, remaining within the soil profile and affecting plant growth.

Problems with the Method:

There are a number of limitations for each experiment process however, specifically for this experiment the labour intensiveness of the process limits its ability to be performed readily, as a minimum of six people are needed.

Any cracks around the probes will allow preferential flow and distort the TDR readings. As the soil in the survey site is a Vertosol, this problem is real. The probe set at 9cm depth showed an area of greater porosity than the rest of the profile was this was possibly the result of the presence of a macropore.

Good soil-probe contact is necessary for accurate readings. As the soil is very clayey, good contact is likely to have been achieved.

Furthermore, the time frame for the experiment did not fully capture the receding wetting front and therefore the accuracy of the data is limited. Exacerbating this is the limitations of the TDR and the wave guides; there is significant time delay between depth as only one wave guide could be read at a time and these were placed at limited depths and numbers. When using this experiment for improving irrigation practices it would be advised to use a larger number of wave guides to a deeper depth, so that infiltration is captured more accurately.