• Develop a robust method for estimating greenhouse gas emissions according to National Greenhouse Gas Inventory methods following land clearing using remotely sensed and other data, suitable for application at the continental scale
• Undertake sensitivity analyses to examine the impact of various land clearing practices and changes in model parameters such as the proportion of soil carbon lost following clearing on greenhouse gas emissions
• Develop an approach to wall-to-wall carbon accounting following land use change applicable at a continental scale by linking land use history approaches with point/patch scale carbon models.

In 1999, land clearing contributed 13.5 % of the emissions identified in Australia’s National Greenhouse Gas Inventory. There are substantial uncertainties around this estimate. Remote sensing of land cover is providing detailed information on land clearing. Spatially explicit, computationally robust methods are needed to link the remotely sensed results with soil carbon and biomass information to improve the reliability of these emission estimates.

The project focuses initially on developing a land cover history for the Fitzroy Basin at the one hectare (ha) level over the period 1970-1999 by combining information from aerial photography, and Landsat MSS and TM satellites. Methods are established for dealing with missing data, for quantifying sequences of clearing and regrowth, and establishing the land cover following loss of woody vegetation. Spatially explicit biomass and soil carbon surfaces at scales commensurate with the land cover history data are to be used to establish greenhouse gas emissions due to land clearing in the Fitzroy Basin since 1970. Regrowth history, its possible causes and impacts on emission rates are to also be examined.

The challenge for continental scale carbon modelling is to link process based point/patch models with data sets describing the high levels of heterogeneity in carbon stocks, and detailed information on the drivers of change, such as land clearing and fire. At the same time, uncertainties in the data and the results obtained through integrating these approaches need to be quantified. To examine how this might be done, a spatially explicit version of the Roth C soil carbon turnover model will be linked with the land use history for the Fitzroy Basin to generate a complete accounting of changes in soil carbon associated with land clearing 1970-1999.

Future developments in national carbon accounting will require robust, computationally efficient approaches which integrate spatially explicit data on temporal changes in land cover, fire history and land management practices with carbon turnover models and spatially explicit data describing carbon stocks.

• Robust computationally efficient method for linking remotely sensed land cover change information with spatially explicit vegetation and soil carbon data to calculate greenhouse gas emissions due to land clearing
• Understanding of the impact of different land clearing practices on the trajectory of greenhouse gas emission losses
• Quantification of the impact of changing soil carbon loss rate parameters following land clearing for grazing of native pastures
• Methods for linking high resolution remotely sensed and other spatially explicit data sets with point/patch scale models to undertake continental scale carbon modelling

• An improved ability to project the effects of land-use and land-cover change on greenhouse gas fluxes in the future.
• A contribution to ensure that Australia’s carbon accounting system continues to achieve international best practice.