AI-Augmented Decision Support System for Evidence-Based Climate Adaptation in Regional Victoria

Natural Resource Managers (NRMs) rely on sectoral modelling approaches that are system specific such as agriculture, forests or rivers. Though the tools provide insights for individual natural systems, they are limited by lack of a holistic evaluation approach, that understands the nexus (trade-offs) between natural ecosystems. As a result, NRMs are limited by lack of integrated evidence on how and why adaptation strategies fail to deliver the intended outcomes.

This study developed an integrated decision support framework, that explicitly links agricultural, forests and rivers systems, to support regional NRMs in Southwest Victoria. This multi-framework foundation ensures that outputs align with real planning processes used by NRMs. The framework evaluates agricultural productivity, habitat distribution and water availability under changing climatic conditions. Using geospatial tools, AI-augmented climate modelling and integrating a multi-framework approach – the tool provides a robust streamlined analysis. The pilot workflow integrates an ensemble of machine learning models to map the impacts. Downscaled climate projections (ACCESS-CM2 SSP585, 2020–2100) were combined with biophysical and land-use data to model land suitability for canola, habitat probability for Kangaroo Grass, and stream yield for the Moorabool River. A rigorous preprocessing pipeline of normalisation, correlation check (IrI>0.7), Variance Inflation Factor (VIF<10), and ML ensemble-based feature selection, improved predictive accuracy. System-specific outputs were combined using an index-based overlay approach using Shapely packages. The analytical workflow progresses from Vulnerability zones for each ecosystem > Trade-offs/Synergies > and arriving at Adaptation Tipping Points. The decision support system (DSS) translates fragmented systems into a comprehensive model to support evidence-based decision making.

The DSS is conceptually anchored in an integrated decision-making framework, that adopts key aspects of decision-making frameworks from Integrated Catchment Management, scenario testing from Resilience Thinking framework, and Adaptation Tipping points from Dynamic Adaptive Policy Pathways, enabling outputs to align with decision processes used by regional authorities. It identifies vulnerable zones, trade-off zones, and possible adaptation tipping points under changing climatic and development pressures.

By translating complex model outputs into accessible spatial layers and scenario-based decision products, the DSS lowers technical barriers for NRMs and strengthens evidence-based planning. The framework is scalable and transferable, providing a replicable pathway for integrating ecosystem service assessments into climate adaptation policy and land-use planning across diverse regions.