The calibration of a neutron probe deals with 4 UNRELATED areas of moisture content measurement. In many cases, not understanding the underlying principles of the calibration procedures leads to unscientific guesses and procedures.

The 4 aspects related to calibration are:

1. The Neutron Probe
2. The Moisture Content Definition
3. The Soil
4. The quality of Water used and Access Tubes used

(A) THE PROBE.

The basic operation of the probe is well known. High Energy (fast) neutrons are emitted from an Am241-Be source. These high-energy neutrons can be subjected to 4 basic actions:

1. Nothing. The neutron leaves the source and carries on into space indefinitely
2. Elastic Collision. The neutron collides with another atom and leaves with most of its energy intact. Sodium, Aluminium, Silicon, Phosphorus, etc.
3. Absorption. The neutron is absorbed into the nucleus of the atom and an energy photon released. Elements such as Boron, Cadmium, Chlorine, etc.
4. Inelastic collision. The neutron collides with another atom and loses energy; it slows down. Elements such as Hydrogen.

We need to understand that, in soil, ONLY these activities (unless the soil itself is radioactive!) can affect calibration.

The element that has the most influence on a neutron is Hydrogen (because it is of similar nucleus size).

By using a He-3 detector tube and some electronics, slow neutrons in soil can be detected. A calibration curve gives a relationship between slow neutron counts per unit time and the moisture content in the soil.

(B) MOISTURE CONTENT – DEFINED.

THE DIFFERENCE IN MASS BETWEEN A WET SOIL SAMPLE AND A DRY SOIL SAMPLE, WHERE THE SAMPLE HAS BEEN DRIED TO CONSTANT MASS BETWEEN 105°C AND 110°C

AT STANDARD ATMOSPHERIC PRESSURE, EXPRESSED AS A PERCENTAGE OF EITHER THE DRY MASS –GRAVIMETIC MOISTURE CONTENT, OR, THE UNDISTURBED VOLUME OF THE SOIL SAMPLE – VOLUMETRIC MOISTURE CONTENT.

(C) THE SOIL.

Soil contains mineral compounds of differing particle shapes and size, Air and Water. Note: this water is part of the soil and NOT what is considered to be "free" or measurable by oven drying. This moisture component may vary with soil density.

When water is added to the soil, Air is replaced with Water, AND, some of the soil with

Water (within the given volume of soil).

If the soil contains organic matter (containing hydrogen), another aspect is introduced.

The soil could also contain traces of elements such as boron, cadmium, salts containing chlorine not precipitated – and other elements causing neutron absorption.

CALIBRATION

Because different soils (elements within the soils) could present with varying proportions of the effects described in (1)-(4) above, and also because of HOW MOISTURE CONTENT IS DEFINED, a calibration method is required to determine a relationship between probe readings and required moisture content.

In order to find a relationship between "free water" (the moisture content of the definition) and other effects described above, and the neutron probe counts, a logical and scientific approach is needed that will allow for the systematic quantifying and accounting for all and any of the above effects.

This approach is unique to the WaterMan probe, but can be freely utilised with other probes (with due acknowledgement).

THE CALIBRATION PROCEDURES

STEP 1 – Carried out by the Supplier

FOR ANY NEUTRON PROBE there exists a unique calibration curve between 0% and 100% by volume of water if the calibration is done in a medium that is otherwise inert or completely devoid of any influences described above. This curve is known as the PRIMARY CALIBRATION CURVE (PCC). It has no relationship to any soil, or any other compound, except WATER (also refer later to water quality variation), the primary measurement objective. This curve is supplied by the manufacturer.

[It is essentially a 3^rd degree polynomial that gives an exact fit across the full range of the PCC]

Count Ratio (CR) = [COUNT – AIRCOUNT]/[WATERCOUNT – AIRCOUNT]

Where COUNT is probe counts per second (or probe counts per minute)

AIRCOUNT is count taken in air (zero moisture)

WATERCOUNT is count taken in clean water.

Moisture = A + B x CR + C x CR^2 + D x CR^3 (m/v)

And A, B, C, D are coefficients of calibration.

STEP 2 – Moisture in the soil

A soil is oven-dried to conform with definition of 0% moisture (m/v). A reading is taken with the probe using the PCC to obtain a moisture reading.

If the probe returns a 0% moisture reading, then clearly there is NO further trapped moisture in the soil.

If the probe does return a moisture reading (M0) (as read off from the PCC, R0), this is as a result of hygroscopic, crystal bound or organic material being present in the soil. It may then be necessary to determine the density of the soil, and check to see what the change in moisture reading is against change in density across the expected density variation in the field. In most cases the changes are negligible and much smaller than the expected standard deviation of the field readings. This reading is introduced as a bias or offset and subtracted from the moisture readings obtained from the PCC. By subtracting this bias only the "free" moisture is reported.

* ** IMPORTANT: Obviously, as the moisture content in the soil is increased, and soil is replaced with water, the bias effect is being reduced until at 100% water, there is no longer can be any bias. The WaterMan gauge automatically takes this into account, but users of other gauges need to take note of this and reduce the effect proportionally.

Count Ratio = [COUNT – AIRCOUNT]/[WATERCOUNT – AIRCOUNT]

Where COUNT is probe counts per second (or probe counts per minute) and

Ro is Count Ratio reading as a result of hygroscopic or crystal bound water trapped in soil

K1 is the moisture offset correction to be applied as a bias to remove this from "free" water reading.

STEP 3 – Neutron absorption in the soil (Slope Correction of PCC – Finding the Slope Corrected Curve )

ADD known amounts of water to the soil sample in small stages. If the probe reports a lower moisture content than that which is added, we deduce that the soil is absorbing some of the high-energy neutrons. Again, at 0% moisture and at 100% moisture the effect is not measurable. By changing the slope of the PCC, this soil effect is taken into account (the simplest curve that will fit the correction is a parabolic transform of the PCC- the effect is 0% at 0% moisture and increases to a point, then drops down to 0% again at 100% moisture).

The PCC in effect requires a steeper slope to account for the missing absorbed neutrons.

Soils containing Boron, Cadmium, Iron and other neutron absorbing elements exhibit this effect (about 2% or less of irrigable land). The writer has only encountered this effect in soils used for road building in the Civil Engineering discipline, seldom in soils used for farming. Most often on soils containing Iron (Fe) present intercept and sometimes slope adjustment offsets.

Again, the WaterMan probe does allow for a parabolic correction to take care of this effect, but users of other gauges need to develop their own models.

Count Ratio = [Count – AirCount]/[WaterCount – AirCount]

Correction = Count Ratio x [Watercount-Count]/[WaterCount – AirCount]

Slope Corrected curve = PCC x (1+ Correction x K2) (SCC)

where K2 is a constant dependent on soil moisture difference between PCC moisture and actual moisture obtained.

For most soils K2 = 0

STEP 4 – Water impurities and access tube effects

This is a field user exercise. By taking Air Counts and Water Counts in the field, using the Water and access tubes available, and inserting these values into the equation to determine Ratio, any impurities or effects of these are discounted.

The WaterMan probe deals effectively with this issue in that the 16 calibrations slots will allow for Water Counts, Air Counts, Slope Offset, and Moisture offsets.

Some of the major difficulties in the calibration of problematic soils are Chlorine, as found in common salt (NaCl), and other similar water-soluble compounds introduced into irrigation water.

The neutron absorption effect of Chlorine and its compounds should be recognised as a primarily water borne rather than a soil effect and this step brings this into account. The same principle is applied to effects introduced by different access tube types (thin walled PVC, thick walled PVC, thin walled steel, aluminium etc).

STEP 5 - Verification

Once all these steps have been gone through, using the oven dried as per definition sample, and adding known quantities of water, the calibration can be checked.

Alternatively, in the field, if the actual precipitation per hectare is known, the probe reading should reflect this precipitation (run-off, evaporation etc. accounted for) and any slope and or intercept adjustments (usually very small), made.

Notes:

The calibration procedures outlined above can be done in standard 200 liter drums. The author has successfully used 100 liter bin-type drums to perform all of the above procedures.

It is important to note that the AMOUNT of water retained by stony, sandy, silty, clayey, and mixtures of these, in terms of mechanical and chemical properties (e.g. Sodium Silicate, Sodium Kaolinite, Sodium Illite, Sodium Montmorillonite), is not in any way related to the calibration of a neutron probe in respect of a soil. The rate of soil moisture depletion IS affected by these mechanical and chemical properties in the sense that sand is more permeable than clay and is less moisture retentive than clay etc., but further discussion lies outside the scope of this paper.

July 2001