LEAD VARIABILITY IN THE WESTERN NORTH-ATLANTIC OCEAN AND CENTRAL GREENLAND ICE - IMPLICATIONS FOR THE SEARCH FOR DECADAL TRENDS IN ANTHROPOGENIC EMISSIONS

As Patterson and coworkers have shown, most of the lead in the modern ocean and atmosphere is of anthropogenic origin. Reductions in the uti lization of leaded gasoline over the past two decades should decrease lead deposition from the atmosphere in remote locations. The search fo r trends in Pb deposition within a single decade is bedeviled by large -amplitude short-term variability due to the inherent noisiness of the atmosphere/ocean system. We find that, over the course of a year, lea d concentrations in the surface waters of the western North Atlantic O cean are variable (factor of 2), and in the snow deposited in central Greenland, highly variable (order of magnitude). In the western North Atlantic, Pb-210 normalization minimizes this problem because Pb-210 a nd Pb sources are spatially correlated and continental Pb-210 emission s are constant. It is clear the Pb in surface waters of the western No rth Atlantic has decreased by a factor of 4 during the 1980s. Pb-210 n ormalization does not help in the Arctic because stable Pb and Pb-210 are not spatially correlated. Because of the order-of-magnitude variab ility in Greenland snow Pb linked to annual cycles, any discontinuous time series is likely to be affected by the phenomenon of aliasing. Al iasing makes it difficult to determine if there is a trend in Pb depos ition in central Greenland during the 1980s; present evidence suggests that the reduction in Pb concentration in snow during the 1980s is le ss than a factor of two; certainly quite a bit less than observed in t he western North Atlantic and less than the factor of >7 reduction in leaded gasoline utilization in the United States during the decade. Al though we expect that decadal-scale trends in the 1970s and 1980s are in fact occurring due to the phasing out of leaded gasoline, the repor ted magnitude of decadal-scale trends should be regarded with some res ervation until confirmed by independent samplings.