The possibility of controlling atmospheric carbon dioxide accumulation and attendant climatic effects from fossil-fuel burning by diverting a fraction of the combustion product and injecting it into the deep-ocean, as proposed by Marchetti, is analyzed using an atmosphere/mixed layer/diffusive deep-ocean model for the carbon cycle. The model includes the nonlinear buffering of CO2 at the air/sea interface, and considers the long term trends associated with consuming an assumed fossil-fuel reserve equivalent to 7.09 × 1015 kg carbon as a logistic function of time as in the projections of Siegenthaler and Oeschger, except that atmospheric carbon dioxide levels are computed for five alternate strategies: (a) 100% injected into atmosphere, (b) 50% injected at oceanic depth of 1500 m and 50% into atmosphere, (c) 50% injected at sea floor (4000 m) and 50% into atmosphere, (d) 100% at 1500 m depth and (e) 100% at sea floor. Since no carbon leaves the system, all runs approached the same post-fossil fuel equilibrium after several thousand years, Ca ∼- 1150 ppm, almost four times the pre-fossil fuel value (∼- 300 ppm). But the 'transient' response of these cases showed a marked variation ranging from a peak overshoot value of 2800 ppm in the year 2130 for 100% atmospheric injection to a slight decrease to the pre-fossil fuel 300 ppm lasting till 2300 with a subsequent slow approach to equilibrium for the 100% deep-ocean injection. The implications of these results for an oceanic injection strategy to mitigate the climatic impact of fossil-fuel CO2 is discussed, as are the ingredients of a second generation carbon cycle model for carrying out such forecasts on an engineering design basis. © 1979 D. Reidel Publishing Company.
