The ever-growing archive of satellite imagery provides an unparalleled dataset for characterising landscape dynamics. The real challenge is to identify, develop and operationalise ways of converting satellite imagery from raw data into decision/policy support information at a spatial and temporal scale that is relevant to decision makers. To achieve this we need to move beyond the current models of heavily parameterized time series analysis and single-scene classifications to a tailored, phenomena-specific approach that exploits the strengths of the myriad of sensors currently available. For a national scale project this means developing a system that: 1. is based on a robust, nationally consistent land cover classification; 2. has a flexible minimum mapping unit that reflects the level of land use intensity and the scale of resource management decisions; and 3. is underpinned by a suite of time series analysis tools that can be tailored to extract the information contained within a time series that represents a specific land use practice. The National Earth Observation (NEO) Group at Geoscience Australia has developed object oriented image analysis and time series analysis techniques to facilitate a new dynamic approach to mapping land cover and land cover change. The flexibility of this approach is essential because the methodology is being applied to a diverse range of bioclimatic regions that span Australia, to address the diverse ecological and natural resource management issues that coincide with these bioclimatic regions and the land uses they support.