Grains Research and Development

Date: 01.08.2004

Looking into your soil from on high and on low

A gammaradiometric image from a 4000ha farm in WA farmed by Glenn Fretwell and family.

By Ian Maling and Dr Matthew Adams

The GRDC SIP09 initiative is using plane-acquired gamma radiometric and ground-based electromagnetic surveys to look into our cropping soils. These are a bit like X-rays that reveal the underlying characteristics of the soil. We hope they will bring us a step closer to cheaply mapping soil characteristics on a farm, paddock and sub-paddock basis.

Unlike alpha and beta particles, gamma rays are detectable to remote sensors due to the low absorption of gamma radiation through hundreds of metres of air. Radioactive forms of the elements potassium, uranium and thorium within soil are the three major elements detected by a gamma ray spectrometer.

A “total count” measurement is made by integrating counts over the whole spectrum measured by the spectrometer. Most of the gamma rays observed from dry rock or soil are from the top 0.30 metres of the soil profile - much of the rooting zone of most crops.

Gamma radiometrics have been applied as an efficient land resource assessment tool for soil type or soil landscape mapping in Australia. If acquired at sufficient line/traverse spacing the potential exists to acquire short-range variation over paddocks and farms that are indicative of soil-forming processes and soil-parent material - a key component for those wishing to manage within-paddock variation.

The main external factors that affect gamma-radiometrics at the time of overflight are soil moisture, the sensor"s height above ground, the speed of the plane, and the size of the detector.

Dr Matthew Adams, of Silverfox Solutions in Western Australia, has been coordinating this effort for the GRDC and SIP09 research teams, and says the contractor, Perth-based UTS Geophysics, has nearly finished its job.

Data already acquired in WA, SA, Victoria, and southern NSW has been provided to teams involved in SIP09 to analyse and compare soil physical and chemical properties being measured on the ground. The remaining farms in northern NSW and southern Queensland were due to be flown over in July.

Magnetics and digital elevation data were acquired at the same time as gamma radiometrics. Magnetic data is useful in hydrological applications. In the west, it is useful for locating dykes, faults and fractures in the bedrock which affect how water moves through the landscape.

At the top on the right is a gamma radiometric image from a 4000 hectare farm in WA farmed by Glenn Fretwell and family. Red areas are high in potassium, green are high in thorium, and blue areas are high in uranium. The red areas are often associated with potassium-containing minerals near the surface, particularly granites with a high component of feldspar minerals.

In the image you can identify a hilltop and alluvial fan from the gamma radiometrics associated from granite outcrops. The blue and green-blue areas are associated with ironstone gravels, which are also associated with poorer performing areas on this farm. A dyke was located via magnetics in the south-eastern section of the farm.

Glenn"s comment on the dyke shown in this image was “that"s the bit of sloshy ground where we always bog the tractor during seeding”. He also said that gamma radiometrics was probably the most useful layer of information he has acquired for his farm.

Crop performance zones resulting from analysis of historical crop biomass/NDVI imagery (image in the centre) have been overlaid on a section of the gamma-radiometric data above and show a reasonable correlation between the red areas and high-performing zones of the paddock, with the poor performing zones being correlated with the bright light blue areas.

Such good correlations between radiometrics and crop performance do not always exist, as bright red colours (lower crop growth) may also indicate shallow soils over a rock outcrop. Restricted rooting depth is then the overriding factor in the crop"s performance. However, this does not diminish the utility of gamma radiometric data. It can help to interpret the cause of yield variability on a sub-paddock scale.

Electromagnetic (EM) surveys are also being undertaken on SIP09 focus paddocks in all states. Three measurements are being made that will “section” the soil profile from the surface to 6 metres below the surface.

The instruments being used for the survey are the Geonics EM31 and EM38 (there are other types of electromagnetic survey instruments) towed by a quad bike.

The EM31 and EM38 measure depth-averaged bulk conductivity. That is, they sample the apparent soil and soil/water conductivity over a certain depth of soil.

EM38h (EM38 in the horizontal mode) measures conductivity primarily in the top 75 centimetres of soil with two-thirds of that in the top 60cm. EM38v (EM38 in the vertical mode) measures the top 1.5m of soil, with one-third in the top 60cm, and two-thirds coming from below that depth.

Roughly 50 percent of the signal comes from the 25cm to 100cm range. The EM31 is similar to the EM38, except it measures to much greater depths in the soil profile.

The application of the different modes of EM38 varies across the country.
In WA, EM surveys have been used primarily for detecting and predicting future salinity areas.

In the east, considerable work by various SIP09 project teams has shown a good correlation between EM measurements and soil properties, such as exchangeable sodium percentage, cation exchange capacity, and soil depth at sub-paddock scales. These have been useful for applying zone management.

Silverfox Solutions is developing a range of PA methods from header yield and biomass analysis to determine the management zones used to control variable rate seeders and spreaders.

In our experience WA clients are saving about $15 to $30/ha by reducing inputs on poor performing zones, and making an extra $40 to $200/ha by increasing the inputs on high performing zones.

This is especially the case when biomass and yield analyses are combined with other datasets like gamma radiometrics and electromagnetics to support the interpretation of low and high performance zones.

A gammaradiometric image from a 4000ha farm in WA farmed by Glenn Fretwell and family.

Overlay of Silverfox zones for VRT on gammaradiometrics

Overlay of Silverfox zones for VRT on gammaradiometrics

Silverfox zoning from biomass analysis (Note: Blue = High; Green = Average; Red = Low:
Pastel colours are variable through time).

Silverfox zoning from biomass analysis

For more information:
Ian Maling, at Silverfox Solutions, 08 9361 0955 or 0418 922 893, or

GRDC Research code: SFX 00001

Region North, South, West