Interoperability of radar and optical data for forest information assessment

The contribution of the world's forested ecosystems to the global carbon budget is well recognized. Increasingly however, we see the ongoing depletion of forest stocks, biomass burning and land cover change; key contributors to greenhouse gas emissions and ultimately climate change. Through international agreements such as UNFCCC REDD, individual country's are charged with a greater responsibility to account for carbon gains and losses, a move which may spur improved management and more sustainable use of forest and natural resources. With such variability in forest and land cover dynamics, individual country's Measurement, Reporting and Verification (MRV) systems must adapt to detect extensive clearing or more subtle change at the landscape level. The acquisition of both optical and radar remotely sensed data at moderate spatial resolution has, and will continue to support regional assessment of forest and land cover change. In this paper we investigate the interoperability of optical and radar data acquired over Tasmania, Australia, for forest information monitoring. The fundamental requirement of an initial baseline from which subsequent change can be measured is outlined. Classification methodologies and forest cover estimates over the timeframe of image acquisition are presented. The independent and combined use of optical and radar data for forest extent mapping, with the emphasis on processing time-series data, is discussed. The methodologies presented in this paper are robust, consistent over a time-series and issues surrounding interoperability are globally applicable.