Machine learning-based Arctic-boreal terrestrial ecosystem CO2 fluxes, 2001-2020

This dataset provides gridded estimates of gross primary productivity (GPP), ecosystem respiration (Reco), and net ecosystem CO2 exchange (NEE) across the circumpolar terrestrial Arctic-boreal region at a 1-km spatial resolution. Monthly CO2 flux data from 2001 to 2020 were generated using terrestrial eddy covariance and chamber CO2 flux observations, combined with geospatial meteorological, remote sensing, topographical and soil data, all within a random forest modeling framework. Aggregated average annual NEE, average annual NEE with direct fire emissions added based on the Global Fire Emissions Database (GFED) product, and temporal trends in annual NEE rasters over 2002-2020 are also included. The data are provided in NetCDF and GeoTIFF formats.