Our results, reported in Hansen et al. (1) in a recent issue of PNAS, illustrate the value of earth observation data sets in generating internally consistent quantifications of land cover change, specifically, gross forest cover loss at the global scale. In that article, we state that our results are limited to this dynamic and that a more complete characterization of forest change would include forest cover gain and the proximate drivers of change. Wernick et al. (2) take issue with (i) our purported dismissal of other data and findings and (ii) the lack of reforestation/carbon uptake analyses. Concerning existing global data sets on forest extent and change, the issues regarding definitions and the difficulty in producing global scale syntheses is well described by others and referenced in our paper (3, 4). We believe that improved consistency in both definitions and methods is needed to advance global forest monitoring. Our results are a step in that direction and illustrate the value of deriving internally consistent and directly comparable results at the global scale. Our study does not replace or supersede other results, and we do not present our result as “the only one based on sound data.” Our results are unique and limited to the gross forest cover loss dynamic.

Concerning reforestation, it is clear from our abstract, main body text, and discussion that our study does not address this dynamic. Our objective was to demonstrate the capability of quantifying a global forest cover loss dynamic, something not yet achievable for other forest parameters such as growing stock. For example, what are the trends in growing stock for the Democratic Republic of Congo during the study period? Such data do not exist. Critiques focusing on data-rich environments ignore our overall objective of global scale assessment. Our approach may be used to assess gross forest cover gain, although adjustments would be needed to address the temporally gradual nature of forest gain compared with loss. Our methodology as implemented enables a direct comparison between biomes, continents, and select countries for forest extent and gross forest cover loss. Improved earth observation data sets, including those that directly observe forest structure, such as lidar data, will allow us to pursue more complete assessments of global forest change in the future.

We acknowledge the importance of local and regional scale forest inventory data emphasized by Wernick et al. (2), and we share their concern that our results be interpreted in light of the limitations explicitly stated in our article. Our objective is to provide a globally consistent comparison of one component of forest change, gross forest cover loss, and the interpretations and conclusions expressed in our article are consistent with that objective.