John Marshall

Professor of Ocean and Climate Science

Evaluating carbon sequestration efficiency in an ocean circulation model by adjoint sensitivity analysis

Evaluating carbon sequestration efficiency in an ocean circulation model by adjoint sensitivity analysis.

(Hill, C and Bugnion, V and Follows, M and Marshall, J), JOURNAL OF GEOPHYSICAL RESEARCH-OCEANS, vol. 109, no. C11, 2004.

Abstract

We demonstrate the application of the adjoint method to develop three-dimensional maps of carbon sequestration efficiency and mean residence time in an ocean general circulation model. In contrast to perturbation sensitivity experiments, the adjoint approach provides a computationally efficient way to characterize both temporal and spatial variations of sequestration efficiency and residence time for a complete global model domain. Sequestration efficiency ( the percentage of carbon injected at a continuous point source that remains in the ocean after an elapsed time), for injections at the base of the main thermocline ( similar to 900 m), is initially lowest in the North Atlantic basin ( except for regions of deep convection in the Labrador Sea) relative to the North Pacific. For injection periods of the order of a century or more, however, the model suggests that Pacific injection sites are generally less efficient for a constant rate injection source. The mean residence time (defined as the average period that impulsively injected carbon from a particular point source remains within the ocean) is also evaluated and mapped. This measure also suggests that Atlantic sequestration is more efficient in the long term. Our calculations draw out the dual role of convective mixing, both exposing shallow sequestration sources to the atmosphere and also, in the subpolal Atlantic and Labrador Sea, feeding carbon from shallow injection sources into the deep circulation away from the atmosphere.

doi = 10.1029/2002JC001598