Energetic closure of the spatially resolved global food system

Abstract: The global food system provides the energy that supports human metabolism, with complex spatial interdependencies between food production, transformation, and consumption. Empirical food system data for these global processes are often fragmented and inconsistent, with only certain components captured in spatially resolved formats. Here we propose a flexible approach to allocating country-level food system data subnationally. We estimate the spatial patterns of food energy production and supply, which we compare to estimates of human metabolism based on average body size. We downscale these rates onto a one-degree resolution grid with 95 corresponding food types to derive an internally consistent, energy-conserving, and spatially resolved dataset. We show that national food supply varies linearly with metabolism per capita, with about half the variation in food supply explained by metabolic rates. Our data processing pipeline is openly available and can readily incorporate new inputs in order to advance trans-disciplinary food system modeling efforts.