Model driven visualisation of climate change scenarios

Climate change will affect us all. In addition to the projected temperature rise, models suggest changes in other climatic averages (e.g., rainfall, sunshine) and also more frequent incidence of special events (e.g., frost, deluge, high winds). These will affect the capacity of the land to support different agricultural or forest products and different modes of production. Land managers need to understand what these changes may be. With this understanding they can better plan their future farm management to be ready and robust in the face of change. This paper introduces regionalised climate change modelling in southwestern Victoria, Australia. Variables from the regional climate models become inputs into comprehensive land suitability analysis for each of a range of agricultural commodities. We then describe some experimental approaches to landscape visualisation which map out the possible changes in land suitability and effective management options for dealing with these. We describe the development and application of a GIS-driven interactive approach to visualisation using a real-time virtual environment augmented by special tools for exploring, in space and time, mapped and statistical data through a video game-like interface. The system, called SIEVE, is based on ESRI ArcMap as the GIS product and Garagegames' Torque Game Engine supporting the virtual environment development. The environment also supports multi-player participation which allows for collaborative review of options by farmers, their neighbours and their advisors. SIEVE is demonstrated in the context of climate change assessment and understanding in what is currently a high productivity farming area supporting dairy, sheep, crop and hardwood timber production. In a recent workshop with regional planners, catchment managers and extension staff, we simulated a farm which had been managed to respond effectively to a specific climate change scenario and one that had not. We then showed how - through interaction in the GIS environment - the less well managed adjacent farm could also improve its productivity. This was portrayed within the collaborative environment with avatars representing three participants (two farmers and an advisor) in the unfolding environmental decision-making. Among the options discussed and visualised were the use of deep rooted perennials and paddock reorganisation to support higher sheep stocking rates; Eucalyptus globulus plantation, a biozone for habitat conservation and low carbon energy production options. Virtual worlds can provide land managers with an interactive tool to explore and adapt to the consequences of climate change. In a formal survey, workshop attendants agreed that SIEVE would be a very useful tool for communicating future scenarios and getting stakeholder participation. They also had suggestions for improvement.