Model outputs were updated on Dec 20, 2017. This project used a machine learning data-driven model to predict the distribution of soil carbon under mangrove forests globally. Specifically this dataset contains: 1) a compilation of georeferenced and harmonized soil profile data under mangroves compiled from literature, reports and unpublished contributions 2) global mosaics of soil carbon stocks to 1m and 2m depths produced at 100 m resolution 3) tiled predictions of soil carbon stocks produced at 30 m resolution 4) shape file containing the tiling system 5) shape file containing country boundaries used for calculating national level statistics For detailed methodologies, please contact JS directly until the paper is published. 30m data can be quickly visualized at: https://storage.googleapis.com/gfiske1/global_mangrove/index_w_slider.html (2017-12-20)
