Use of pan-tropical biomass maps and deforestation datasets to derive carbon loss estimates for the Amazon biome

IPCC Tier 1 above-ground biomass (AGB) default values per ecological zone have high uncertainties. Remote sensing based pan-tropical biomass maps can be used to derive more realistic Tier 1 values and furthermore allow a pixel-level analysis. Such approach enables more robust AGB estimates at ecological scale as the geospatial pattern of AGB in tropical forests is taken into account. Our study investigates the impact of different activity (deforestation) datasets and carbon emission factors on carbon loss over the last decade for the Brazilian Amazon. Estimates of the carbon loss vary strongly: up to 83% and 66% relative differences depending upon the emission and activity datasets used respectively. While the Brazilian carbon map delivers higher carbon estimates than the remote sensing based AGB datasets, the Brazilian activity dataset shows lower deforestation rates than the Tree-cover product. However, the sample-based TREES approach delivers deforestation estimates which are quite close to the official Brazilian data. The combination of emission and activity data over the period 2007-2012 leads to carbon loss estimates that range from 59 to 172 megatons per year. When applying a spatially explicit approach low uncertainties at pixel-level are required. Thus, a combination between highly accurate activity data and spatially explicit AGB data, such as provided by the newly developed data fusion model, is recommended.

LANGNER Andreas Johannes;  SHIMABUKURO Yosio Edemir;  ACHARD Frederic;  SIMONETTI Dario;  MITCHARD E. T. A.; 

2015-12-18

EUROPEAN SPACE AGENCY

JRC96335

0379-4067,   

https://publications.jrc.ec.europa.eu/repository/handle/JRC96335,