Carbon emissions risk map from deforestation in the tropical Amazon

Assis, Pedro Valle This work aims to estimate the carbon emissions from tropical deforestation in the Brazilian Amazon associated to the risk assessment of future land use change. The emissions are estimated by incorporating temporal deforestation dynamics, accounting for the biophysical and socioeconomic heterogeneity in the region, as well secondary forest growth dynamic in abandoned areas. The land cover change model that supported the risk assessment of deforestation, was run based on linear regressions. This method takes into account spatial heterogeneity of deforestation as the spatial variables adopted to fit the final regression model comprise: environmental aspects, economic attractiveness, accessibility and land tenure structure. After fitting a suitable regression models for each land cover category, the potential of each cell to be deforested (25x25km and 5x5 km of resolution) in the near future was used to calculate the risk assessment of land cover change. The carbon emissions model combines high-resolution new forest clear-cut mapping and four alternative sources of spatial information on biomass distribution for different vegetation types. The risk assessment map of CO2 emissions, was obtained by crossing the simulation results of the historical land cover changes to a map of aboveground biomass contained in the remaining forest. This final map represents the risk of CO2 emissions at 25x25km and 5x5 km until 2020, under a scenario of carbon emission reduction target.