Developing collaborative planning support tools for optimised farming in Western Australia

Abstract

Land-use (farm) planning is a highly complex and dynamic process. A land-use plan can be optimal at one point in time, but its currency can change quickly due to the dynamic nature of the variables driving the land-use decision-making process. These include external drivers such as weather and produce markets, that also interact with the biophysical interactions and management activities of crop production.The active environment of an annual farm planning process can be envisioned as being cone-like. At the beginning of the sowing year, the number of options open to the manager is huge, although uncertainty is high due to the inability to foresee future weather and market conditions. As the production year reveals itself, the uncertainties around weather and markets become more certain, as does the impact of weather and management activities on future production levels. This restricts the number of alternative management options available to the farm manager. Moreover, every decision made, such as crop type sown in a paddock, will constrains the range of management activities possible in that paddock for the rest of the growing season.This research has developed a prototype Land-use Decision Support System (LUDSS) to aid farm managers in their tactical farm management decision making. The prototype applies an innovative approach that mimics the way in which a farm manager and/or consultant would search for optimal solutions at a whole-farm level. This model captured the range of possible management activities available to the manager and the impact that both external (to the farm) and internal drivers have on crop production and the environment. It also captured the risk and uncertainty found in the decision space.The developed prototype is based on a Multiple Objective Decision-making (MODM) - á Posteriori approach incorporating an Exhaustive Search method. The objective set used for the model is: maximising profit and minimising environmental impact. Pareto optimisation theory was chosen as the method to select the optimal solution and a Monte Carlo simulator is integrated into the prototype to incorporate the dynamic nature of the farm decision making process. The prototype has a user-friendly front and back end to allow farmers to input data, drive the application and extract information easily.