Past production choices and short-term seasonal weather forecasts direct food and fibre production, inhibiting preparedness for operating under a changed climate, which requires larger planning horizons. Simulations have been the primary tool for investigation of alternative production strategies. This approach is resource hungry and lacks goal objectivity due to agent biased input decisions. As an alternative, optimisation can assist as it is resource parsimonious and objective focussed. However, mathematical representation of natural environments is inherently compromised due to its reduction of complex land-crop-climate interactions to searchable values, adding uncertainty to model outputs. Qualification and quantification of uncertainty is frequently omitted in model design and reporting. This compromises model credibility and usefulness as a Decision Support System (DSS).
This need for a robust temporal dynamic agricultural tool is confronted through a spatiotemporal optimiser of rotation sequences (STORS). Using its ability to forecast Climate Smart Landscapes (CSL) where production choices are sympathetic to climate allowing us to use Ecological Intensification of Agriculture (EIA) to optimise food and fibre commodities.
A generic framework connects abiotic parameters with biotic attributes, allowing the evolutionary algorithm, STORS, to generate Pareto-optimal land use alternatives from which farms and regional planners can select for decision making. A case study of the Murrumbidgee Irrigation Area (MIA) is used, where five agricultural production systems, across six soil types, under five water scenarios is investigated.
Outcomes will facilitate answering two key research questions 1) Can STORS identify feasible rotation sequences? 2) Is there a temporal shift in crop species.