Generating spatially robust carbon budgets from flux tower observations

Estimating global terrestrial productivity is typically achieved by rescaling individual flux tower measurements, traditionally assumed to represent homogeneous areas, using gridded remote sensing and climate data. Using 154 locations from the FLUXNET2015 database, we demonstrate that variations in spatial homogeneity and nonuniform sampling patterns introduce variability in carbon budget estimates that propagate to the biome scale. We propose a practical solution to quantify the variability of vegetation characteristics and uniformity of sampling patterns and, moreover, account for contributions of sampling variations over heterogeneous surfaces to carbon budgets from flux towers. Our proposed space‐time‐equitable budgets reduce uncertainty related to heterogeneities, allow for more accurate attribution of physiological variations in productivity trends, and provide more representative grid cell averages for linking fluxes with gridded data products.