The Impact of Industrial Activity on the Amount of Atmospheric O2

Concurrent measurements of atmospheric O₂ and CO₂ amount fractions have been used for decades now to estimate fluxes of carbon to and from the oceans and the land biosphere. The equations used in these estimates explicitly include fossil fuel combustion but are built on the assumption that fossil fuels are oxidized solely by atmospheric O₂, ignoring the small fraction of fossil fuel oxidation by oxide ores that occurs during refining of metals, and thereby overestimating losses of atmospheric O₂. Here, we address this deficiency by quantifying the effective O₂ fluxes associated with the processing of iron, aluminium and copper. We also consider the potential impact of sulfur. We find that consideration of the oxygen mobilized during metal processing (equivalent to a net O₂ flux of 12.0_–0.4^+0.2 Tmol a^–1) leads to an increased estimate of ocean carbon uptake in the years 2000-2010 of (0.144_–0.005^+0.002) Pg a^–1 of carbon with a corresponding decrease in estimated land uptake. A rough estimate of oxygen uptake due to sulfur chemistry during fossil fuel combustion (2.4 Tmol a^–1) decreases ocean carbon uptake but by a much smaller amount. These corrections are small compared to existing estimates of the fluxes and their uncertainties ((2.27 ± 0.60) and (1.05 ± 0.84) Pg a^–1 of carbon for ocean and land respectively (Keeling and Manning, 2014)) but should be employed in future analyses.