A time resolution methodology for assessing the quality of lake sediment cores that are dated by 137Cs

A methodology is proposed for grading the utility of lake sediment cores used to reconstruct pollution histories. The observed distribution of 13’Cs with depth in the core is compared to that expected from independent, historic measurements of fallout deposition. The width of the 13’Cs peak in the core profile that corresponds to the fallout maximum of 1963, or the combined maxima of 1959 and 1963, is used to infer the inherent time resolution of the core, i.e. the ability to distinguish events in the deposition history of the watershed. The method is applied to a number of core profiles from various lakes in the U.S. and appears to provide self-consistent results. Lake sediment cores are used to reconstruct the history of the input of trace substances into the ecosystem (Edgington and Robbins 1977; Galloway and Likens 1979; Heit et al. 1980; Norton et al. 1980). Local, regional, and global depositions have all been inferred from changes observed in the distribution of a given substance throughout a core. Currently, there is much interest in using this technique to reconstruct the history of acid precipitation in certain areas of the country (Galloway et al. 1983; Heit et al. 198 1; Ouellet and Jones 1983). A common technique used to date a core involves use of the peak in the 137Cs concentration in the core profile as a marker of the 1963 peak fallout deposition on the earth’s surface from nuclear weapons tests conducted before the atmospheric test ban treaty (Jaakola et al. 1983; Pennington et al. 1976). Examples of other dating techniques include use of the naturally occurring radionuclide 210Pb (half-life, 22 years) (Robbins and Edgington 1975; Koide et al. 1973), Pu (Heit et al. 1984), and pollen (Robbins et al. 1978). Unfortunately, many investigators establish only a rough dating of their cores and then proceed to interpret variations in other measured trace substances, ascribing definitive dates to each core section. Only a qualitative treatment of the error associated with the dating is usually given. We have taken sediment cores in 30 different lakes and reservoirs across the U.S. since 1980 and have found a wide range in apparent definition We believe that some quantitative measure of the resolution of a core and how it can be used to delineate real temporal changes in pollutant flux to a watershed is needed. For example, we have cores that clearly show two separate 137Cs peaks, correlated with the periods of heavy atmospheric nuclear weapons testing in the late 1950s and again in the early 1960s. Such two-peak profiles are indicative of an extremely settled and undisturbed sediment. On the other hand, many cores have a single broad 137Cs peak with a pronounced sign of tailing, which indicates mixing, diffusion, or some integrative process within or external to the sediment. Clearly, the former case is much better suited for analysis of short term trends, whereas the latter may be capable only of distinguishing patterns that occur over a timespan of a decade or more. We therefore propose that an estimate of time resolution be used to assess the utility of a sediment core for reconstructing the deposition history in an area. A suitable parameter, in units of years, for instance, would then be a measure of the minimum time that two separate events in the deposition history of the watershed could be distinguished in a core profile. Such a concept has actually been used to map southern Lake Huron by Robbins (1982); the parameter he used was defined as the ratio of the mixed depth layer to the sedimentation rate and was ultimately derived from applying a model to account for mixing in lake sediments. We introduce here an empirically