Sensor network builds soil moisture knowledge bank

Key points

  • Thirty-three public-access soil moisture probes exist across the WA wheatbelt
  • Data from soil moisture probes is being used with Yield Prophet® modelling to improve soil type characterisations
  • The data can better inform on-farm decisions in dry years and help growers better understand water-holding capacity in wet years
Photo (from left) of Frank D'Emden, Brad Tonkin and Kristin Lefroy

(From left) Frank D’Emden  with Brad Tonkin and Kristin Lefroy from the Moora-Miling Pasture Improvement Group at the Bindi Bindi soil moisture probe site. Data from soil moisture probes is being used with Yield Prophet® modelling to improve soil type characterisations.

Theories on the water-holding capacity of soils across Western Australia’s wheatbelt are being tested after the wet winter in 2016, utilising a network of moisture probes

A wet winter across parts of Western Australia’s wheatbelt in 2016 has helped researchers recalibrate soils’ ‘bucket size’. The recalibrations are part of a broader project designed to give growers access to real-time soil moisture and weather data and periodic Yield Prophet® reports.

Moisture dynamics

Understanding soil moisture dynamics is an important consideration for growers when making in-season and, in some cases, pre-season management decisions.

Yield Prophet® uses the Agricultural Production Systems sIMulator (APSIM) computer model together with paddock-specific soil, crop and climate data to generate information about the likely outcomes of farming decisions.

APSIM, developed by CSIRO and internationally recognised as a highly advanced simulator of agricultural systems, contains a suite of modules that enable the simulation of systems that cover a range of plant, animal, soil, climate and management interactions.

A network of soil moisture probes and Yield Prophet® sites was established in the Esperance, Albany, Kwinana East and Kwinana West port zones in 2014 and monitored in 2015. Most of these sites were established through the GRDC’s Regional Cropping Solutions Network initiative, with some grower groups installing additional sites.

Probing the soil

EnviroPro capacitance probes were buried at a depth of 25 centimetres, with sensors spaced at 10cm intervals, providing estimates of soil water content to 65cm or 105cm, depending on soil depth.

Automatic rain gauges and temperature/humidity sensors are also at each site, enabling site-specific rainfall and temperature data to be fed directly into the Yield Prophet® model.

Soil cores for physical and chemical analysis were taken during probe installation, with samples generally taken at 10cm increments to a depth of 50cm, and subsoil samples taken from 50 to 70cm and 70 to 100cm or shallower depending on the soil depth.

All samples were analysed for ammonium, nitrate, electrical conductivity (as a salinity measurement, EC1:5), pH (calcium chloride), chlorine and particle size (sand, silt and clay).

The probes were calibrated to 104 individual sensor settings and particle size analysis and soil chemistry data entered into CSIRO’s Soil Water Express calculator. This helped to provide an estimate of plant available water (PAW), by way of subtracting the crop lower limit estimate from the sum of the calibrated soil moisture sensor data. The total PAW estimates require further calibration to account for the influence of subsoil constraints such as salinity, sodicity, acidity or other subsoil constraints.

Precision Agronomics Australia’s Frank D’Emden, who led the Esperance and Albany port zone components, says the network can help growers make in-crop decisions, especially regarding whether to apply top-up nitrogen. However, in 2016 the network has helped in another way with the wetter season allowing researchers to recalibrate some theoretical ideas about soil water-holding capacity, with new drained upper limits being discovered at some sites.

“It is quite difficult to learn these limits without artificially wetting up to calculate a soil’s bucket size,” Mr D’Emden says. “So a wetter year helped us to test the theoretical sizes.”

This was in contrast to the dry 2015 winter when the network’s importance in helping growers make in-crop decisions was clear. Visits to the website hosting the information reached almost 900 in June 2015, compared with a high of just over 300 in 2014. (And despite a wetter 2016, the site had nearly 2000 hits.)

Mr D’Emden says increased traffic to the site in 2015 was possibly a result of growers wanting to see how much moisture they had in the bucket. “It’s showing that the extra information is at least being accessed, with anecdotal evidence that some growers decided not to put down extra nitrogen that year, despite a lot of crops getting a good start.

“In previous years (and without the subsoil moisture data) there would have been the temptation to put extra nitrogen down. But some growers have looked at the data and the Yield Prophet® model results and decided not to.”

Map showing the spread of moisture probes in WA that growers can access

Figure 1 The spread of moisture probes in WA that growers can access.

Mr D’Emden says it was a good decision. “We had very low rainfall across some parts of the wheatbelt for the rest of that winter, so the extra nitrogen would not have been worth it. In other areas where winter rainfall was tracking below average, the probes were showing high levels of stored soil water from late 2014, so in these areas we had the confidence to supply crops with more nitrogen, which was a decision that paid off despite a relatively dry August in those areas.”

Results from the projects showed that half of the Yield Prophet® estimates were within 300 kilograms per hectare of the final crop yield at GS30 (June/July) and 70 per cent at GS90, with the median difference between the Yield Prophet® GS90 estimates and final yield being 0.35 tonnes/ha.

The existing network of sites is also helping a range of different projects – from the Department of Agriculture and Food, WA, CSIRO and the Birchip Cropping Group, and funded via the GRDC’s Regional Cropping Solutions Network – to improve understanding of how soil moisture probes and Yield Prophet® can be used to complement each other; increase the range of soil type selections in Yield Prophet®; and cross-validate soil moisture probe data with Yield Prophet® soil moisture modelling.

Moisture probe and weather data are logged and transmitted via the NextG Global System for Mobile Communications (GSM) network using Outpost Central telemetry.

More information:

Frank D’Emden,


Take the lid off profit potential


PA conference reflects generational change

GRDC Project Code PRE00004

Region West