Requirements analysis for remote sensing of carbon-climate feedbacks

Feedbacks from terrestrial and marine ecosystems will significantly influence future climate, but the magnitude of these feedbacks is poorly constrained by observation and experiment. A major limitation is the mismatch between the local scale of process observations and the coarse resolution at which fluxes are resolved in the atmosphere. New capabilities from GOSAT and OCO-2 will aid significantly bringing the scale of atmospheric inverse estimates of fluxes closer to the process scale, but these powerful instruments are a partial solution. Specifically, resolution and seasonal and spatial sampling biases increase uncertainty for partitioning fluxes between temperate and tropical, terrestrial and marine and land use versus earth system feedbacks. These limitations affect the estimation of feedbacks apparent in mean fluxes, such as effects of increasing CO2 and in flux variability, reflecting climate and land use changes. We show simulations of climate feedback signatures in atmospheric concentration patterns and analyze how these might be best observed. We conclude that separation of land use and earth system driven fluxes is crucial, that resolution of tropical ocean variability is important to estimating the terrestrial sensitivity to climate and that uncertainty from fossil emissions significantly affects estimates of feedback strength.