finds that measured CO, concentration gradients in the atmosphere over the oceans off Tiana Beach, Long Island, New York during December 1978 are 'about two orders of magnitude greater than the values that one would expect based on net global mean flux into oceans'. The global mean flux was computed to be 2.2 x lo-ii g cme2 s-l of C02. (Note: on both occasions when this figure was mentioned in the paper, the negative sign on the exponent was omitted.)
This figure was deduced from a release of 5 x 10" g yr-' (Baes et al., 1976), a value that is expressed in terms of the mass of carbon, not the mass of CO,. The computed mean global flux must therefore be revised upward by a factor of 44/12 k 3.7 if it is to refer to mass of CO,. In the discussion below we shall adhere to the more common procedure of working with carbon masses so as to preserve comparability between the various carbon reservoirs, except when we specifically quote amounts of CO, for the purposes of comparison.
Assuming that possible contributions from deforestation and soil organic losses can be ignored, the above value is a reasonable estimate of the global mean flux. It must however be appreciated that net fluxes representative of specific latitudinal zones and specific times of the year can differ considerably from the global average. In a recent attempt to model the global atmospheric distribution of CO,, including exchanges to and from the oceans (Hyson and Pearman, 1980), we estimated an annual release over tropical oceans (12" N to 12" S) of 0.22 Gt yr-' and an uptake at higher latitudes of 3.3 Gt yr-', equivalent to mean annual surface fluxes of -0.9 x 10-l' and 3.7 x 10-l' g cme2 s-l respectively. Using the same model to give a more detailed description of the spatial and temporal variations in surface flux gives an estimate of 10 x IO-" g cme2 s-l for the mean December flux at about 40" N. (This corresponds to 37 x lo-" g cme2 s-l of CO,.) It is of interest to note that measurements using large-scale sampling (both vertically and horizontally) gave winter and summer fluxes of plus and minus 3.5 x 10-l' g cmm2 s-l, respectively, for 40" S (Garratt and Pearman, 1973). It should also be mentioned that our model estimates of zonal fluxes are smaller than those determined by . We believe therefore that SethuRaman's use of a global Boundary Lajer Meteorology 22 (1982) 257-259.