Modeling vegetation, carbon, and nutrient dynamics in the savanna woodlands of Australia with the AussieGRASS model

Models of savanna and grassland systems in Australia are used for two distinct aims. The first aim is an applied operational role, where models are run in near real-time and results are used to address the information needs of public policy decisions, for example, drought situation analysis; State of Environment reporting; and provision of climate risk assessments of pasture resources for tactical management of grazing and fire. The second aim is to improve the scientific understanding of processes and to simulate the likely long-term outcomes and risks of current management practices for strategic policy and planning purposes. AussieGRASS (Carter et aI., 2000) provides a simulation of the soil water balance and pasture dry matter fluxes over the Australian continent. Simulations are spatial and are performed on individual grid cells (approximately 5 km x 5 km) across Australia. AussieGRASS uses a version of the GRASP model (Rickert et aI., 2000) and simulates the hydrology and dry matter flow of pasture communities at a daily time step. It includes, as gridded inputs (5 km x 5 km), information layers of daily climate data, soil attributes (e.g., plant available water range), tree cover, pasture community, and densities of grazing animals (i.e., domestic livestock, feral and native herbivores). The information layers of soil type and pasture community allow specification of parameters for available water range of four soil layers and pasture attributes affecting plant growth, senescence, detachment, and decomposition. A key feature of GRASP is that hydrological processes such as pasture transpiration and runoff are driven by dynamic groundcover estimates that are updated daily. Thus, AussieGRASS is a dynamic model, responding to daily variation in climatic variables and moderated by management, fire, and herbivory.