To measure the effectiveness of efforts to reduce emissions from deforestation and forest degradation (REDD+), one requires data on
both forest cover (change) and biomass stocks. Regional to global datasets with increasing levels of coverage, spatial and temporal detail, and
accuracy stem from innovations in remote sensing and forest monitoring. Still, these datasets do not necessarily align with each other, and it
remains unclear how their uncertainties influence carbon emission estimates.
Our study area covers six REDD+ subnational initiatives in five countries across the tropics. We compared approaches to quantifying impacts on
carbon emissions. We performed an accuracy assessment on the activity data using several tree cover change datasets such as locally
calibrated products based on dense time series data and a global dataset on annual tree cover change. For the error estimation, we used
validation tools for human interpretation based on Landsat Time Series data and high resolution optical imagery. Next, we calculated carbon
emission estimates based on pantropical biomass maps and field inventory data. We differentiated emissions from before and after the start of
the REDD+ initiatives and also considered control areas to estimate REDD+ impact.
We found that forest change products based on locally calibrated algorithms had a higher accuracy than the global product assessed. Biomass
datasets built on both remote sensing and local field inventory data led to better carbon emission estimates. REDD+ impact was limited but
varied considerably between initiatives. Still, the choice of datasets and assessment methods had an influence on the measured local and
regional REDD+ performance. This study contributes to carbon measurement, reporting and verification by providing insight into what extent both
activity data and biomass data influence the uncertainty of carbon emission estimates.