Pre-Assessment of the value of remote sensing derived data for reducing uncertainty about carbon emissions estimation in Brazil

The increase in CO2 and other Greenhouse Gases in Earth’s atmosphere is the main driver of climate change, issue which urges for the development and application of global mitigation strategies. Deforestation and land use change in the tropics is known to be responsible for a good part of those emissions, which motivated the creation of mechanisms of compensation for conservation of forests in developing countries, such as Reducing Emissions from Deforestation and Forest Degradation (REDD). Some of the greatest challenges of such mechanisms are the large uncertainties about carbon emissions from deforestation and the costs of carbon monitoring systems to measure and verify the impact of the REDD+. This research will investigate data and methods for uncertainty assessment and develop strategies of data acquisition that confronts information content against the costs of monitoring by 1) fitting and the evaluating error models for input data of carbon emissions estimates; 2) comparing error propagation techniques and identifying the most relevant sources of errors 3) evaluating the value of input data for reducing uncertainties and development of a framework for designing optimal data acquisition strategies. The thesis will be complemented with a stepwise case study showcasing a complete assessment of uncertainties and application of the developed data acquisition optimization scheme, using remote sensing and field data acquired for the Municipality of Paragominas, in Brazil.