Farmers trial climate and decision software

The diagram shows that rainfall at Loxton to 18 May 1997 is at about the decile three level which falls within the driest 30 per cent of years recorded at Loxton. The average rainfall for Loxton at 1 June (decile 5) is about 45 mm. (A decile records accumulated rainfall to date relative to previous years.)

A small 1996 road test to assess computer aids which link climate data to wheat planting and management decisions has grown to a full-scale evaluation by 160 farmers in two states in 1997.

Climate Risk Officer with Primary Industries SA, Jacqui Balston, said that two decision-support computer programs developed by Agriculture WA, TACT and PYCAL, had been tested in 1996 by 22 farmers and advisers from northern Eyre Peninsula and the eastern wheatbelt of WA.

She said TACT was a seasonal wheat-sowing decision aid providing yield and gross margin predictions for a range of probable outcomes based on rainfall to date and historic climate data. Farmer data could be entered into the program to generate farm-specific information and yield simulations.

PYCAL calculated stored soil water and potential yield for a number of different crops using current season rainfall and cumulative rainfall data. It is expected to be offered as a TOPCROP® Australia product in 1998.

160 farmers participating

Ms Balston said that this year Primary Industries SA, Agriculture WA and the Kondinin Group had invited 160 farmers to participate in the extended trial called the Climate Risk and Yield Information Service. At this stage the service was restricted to a representative group of growers from districts with an average annual rainfall of 350 mm or less.

The service analyses farm-specific data and compares it with historical rainfall patterns for the region.

In February farmers provided the service with details such as daily rainfall since November, descriptions of two important soil types on their farm, amount of rain required to start seeding, expected variable input costs, and on-farm wheat price forecasts. From 1 April to 30 June the farmers update the rainfall and input cost information weekly.

In return each farmer receives an individualised weekly update including information on:

  • accumulated rainfall to date relative to previous years (deciles)
  • predicted stored water
  • probable simulated yields
  • an economic analysis and
  • the probability of rainfall based on historic data.
  • "The information on climate risk and yield prospects can support decisions to adjust inputs and crop area in response to seasonal conditions," Ms Balston said.

    Past experience indicates that rainfall variability can account for a potential yield difference up to 1.5 t/ha within a year on neighbouring farms.

    "A number of other decisions such as crop variety, herbicide and fertiliser use and even machinery purchase may be influenced by the information supplied."

    She said that in addition to the evaluation by farmers on the northern Eyre Peninsula and in eastern WA, the TACT and PYCAL programs are also being tested by 17 farmers in the SA Mallee in 1997.

    TACT and PYCAL can be run on any desktop or laptop IBM computer with 386 chip or above. According to Ms Balston, you would need up to 6 mB of hard drive space to run both programs. For further information, contact Ms Balston.

    Farmers in the eastern wheatbelt can obtain the programs from Doug Abrecht, Agriculture WA, 3 Baron-Hay Court, South Perth WA 6151. Eyre Peninsula farmers can obtain them from Jacqui Balston, Primary Industries SA, CRC for Soil and Land Management, Private Bag 2, Glen Osmond SA 5064. At this stage both programs are free.

    The following tables indicate the type of information provided by TACT and PYCAL.

    Subprogram 3.7.2 Contact: Ms Jacqui Balston 08 8303 6731

    Yield predictions Long run railfall records: Minnipa
    &nbspWheat: MinnipaSL Wheat: Cungena LS
    75% (3 in 4 chance) of over0.3 t/ha0.8 t/ha
    50% (1 in 2 chance) of over1.1 t/ha1.4 t/ha
    25% (1 in 4 chance) of over2.2 t/ha2.5 t/ha

    Seen here, baseline yield probabilities for any year based on 100 years of climate data. The long-term average (1 in 2 chance) simulated yield for Minnipa in South Australia is 1.1 t/ha on the heavy land (Minnipa sandy loam) and 1.4 t/ha on the light land (Cungena calcareous soil) for a crop sown according to the standard sowing rule for the district. In wetter years these probabilities would improve and, in dryer years, may decline.