VENTILATION RATES OF THE GREENLAND AND NORWEGIAN SEAS DERIVED FROM DISTRIBUTIONS OF THE CHLOROFLUOROMETHANE-F11 AND CHLOROFLUOROMETHANE-F12

Distributions of the chlorofluoromethanes (CFM) F11 (CFCl3) and F12 (CF2Cl2) in the Greenland Sea were measured in late winter 1989 as part of the Greenland Sea Project (GSP), where Norwegian Sea Deep Water (NSDW) was also sampled in the Lofoten Basin. In the deep Boreas Basin inflow of NSDW to Greenland Sea Deep Water (GSDW) could be observed. The CFM distributions within the GSDW were homogeneous and not altered compared to 1982. Thus, deep convection in the Greenland Gyre was reduced since 1982 compared to previous years where various tracer measurements (1972, 1979, 1981 and 1982) prove convective turnover of GSDW during that time. In the deep Norwegian Sea, however, the CFM concentrations have increased since 1982. The surface waters in the Greenland Gyre are saturated to only 79 +/- 5% compared to ambient atmospheric CFM concentrations due to previous ice cover and/or rapid cooling of the surface waters. For the first time, CFM measurements have been made in a convection event reaching down to intermediate depths. A comparison of the CFM inventories of this location with those of the mean CFM depth profiles in the Greenland Gyre reveals no difference, showing homogenization of the water column, but no further equilibration between atmosphere and ocean during convection. This may be caused by the short duration of a convection event compared to the equilibration time for gas exchange between air and sea. Although convection did not reach the bottom, the CFM measurements made during convection are applied to construct a CFM boundary condition for the convective renewal of GSDW. With two different five-box models and the boundary conditions developed, ventilation rates of GSDW and NSDW also take into account the reduced turnover rate of GSDW since 1982. For the first time, model ventilation times are presented coincident with the F11 and the F12 distribution. In model 1 the reduced convective renewal of GSDW is compensated by an appropriate reduction of the GSDW volume. In model 2, mass balance of the GSDW box is achieved by an upwelling term of the same magnitude than the renewal term. After 1982 both terms are reduced in the same manner. The turnover times of GSDW due to convection prior to 1982 of 19-20 (model 1) and 13-15 years (model 2) are much shorter than reported previously. In the years 1983-9 the model-derived ventilation time increases to 190-200 years (model 1) and 42-67 years (model 2).