Forward and Inverse Modeling of CO 2 and 13 CO 2 in the NCAR CCSM 1 FINAL REPORT

The spatial and temporal distribution of CO2 concentration in the atmosphere contains information regarding sources and sinks at the surface. Quantitative interpretation of this information requires the use of a chemical tracer transport model (CTM) (Enting et al, 1995; Fan et al, 1998; Rayner et al, 1999; Bousquet et al, 2000; Gurney et al, 2002). The problem of inverting the tracer transport model to obtain emissions from concentration data is complicated by the fact that regional sources and sinks may produce nearly identical concentration patterns as seen by the present observational network, which cannot be distinguished on the basis of CO2 concentrations alone. A sink resulting from nitrogen fertilization of the Boreal forest, for example, might produce the same depression in observed CO2 concentration in the marine boundary layer at 50° N latitude as a somewhat weaker sink resulting from uptake by enhanced phytoplankton productivity in the North Atlantic Ocean. In other words, the matrix relating fluxes and concentrations, which must be inverted to estimate these sinks, is nearly singular