The savanna ecosystem is the second largest biome of tropical South America. Although deforestation of the Amazon rainforest has received much public attention, most of the Brazilian deforestation has taken place, and continues to take place, in the savanna environments (locally called Cerrado). Of the original 2,000,000 km2 of Cerrado that existed in Brazil before about 1940, about 50% has been converted and fragmented by deforestation and expansion of the agriculture frontier. The Cerrado is considered a hot spot for biodiversity and is the headwater region of major rivers of eastern South America. It is also currently the focus of agricultural expansion in response to domestic and international markets for soy and biofuels. The potential for expansion of intensive agriculture onto land currently occupied by degraded cattle pastures (i.e. land that is already altered but underutilized) is often presented as a means of avoiding further deforestation in the Cerrado and minimizing environmental degradation while increasing agricultural productivity. However, the capacity and suitability of land for this transition has not been evaluated or quantified on a regional basis. This proposal aims to quantify these degraded lands and their potential to mitigate future environmental impacts. The specific objectives are: 1) Quantify the area of low productivity pastures using multiple resolution satellite sensors. 2) Assess the impacts of a range of economic scenarios for sugarcane and soybean expansion on future land use changes in the Cerrado. 3) Appraise inter-regional shifts in agricultural production in response to economic drivers within Brazil and the impact they have on deforestation pressure on the Cerrado 4) Estimate historical and potential future changes in carbon stocks and emissions of N2O and CH4 from conversion to agricultural production. 5) Assess the hydrological impacts of historical and future scenarios of deforestation and conversion to intensive agriculture. The first objective will be addressed by using a combination of Landsat-TM, MODIS, and ALOS products. TM-derived reflectance values will be related to field measurements of biomass and will be used to scale up to the MODIS resolution to calculate pasture biomass values for the entire Cerrado region. The second and third objectives will be addressed by expanding and refining an existing Cerrado deforestation model by integrating it with a national and sub-national macroeconomic model. The amount of land allocated to sugarcane crop will be simulated based on national and global market demands for biofuels and conversion from degraded pastures and other land use categories to sugarcane and soybean. The results will be linked to a regionalized model of the Brazilian economy to simulate inter-regional economic effects on the Cerrado. The fourth objective will be addressed by calculating carbon consequences and the net effect on N2O and CH4 emissions from historical deforestation and each future scenario, using a bookkeeping model based on summaries of changes in C stocks and gas fluxes among land uses derived from LBA research and other data on emissions from agricultural fields. The fifth objective will be a follow-on activity of our expiring NASA-LCLUC project (Davidson, Coe, et al), where we are estimating effects of historical land use change on the hydrology of the Cerrado, using coupled global climate, vegetation, and surface hydrology models. The proposed project responds to both components of land-use change detection and identification of economic and social drivers in an area of agricultural expansion that is currently understudied. Because forest conservation efforts in the Amazon under future implementation of Reducing Emissions from Deforestation and Degradation (REDD) could lead to increased pressure for agricultural and pasture land in the neighboring Cerrado region, our work will also contribute to this international program.