129I in the SE-Dome ice core, Greenland: A new candidate golden spike for the Anthropocene.

The Anthropocene is a proposed geological epoch that will mark the time when humans have irreversibly affected the Earth. One of the primary requirements to formally establish this is a Global Boundary Stratotype Section and Point or "golden spike" - a record of a planetary signal marking the new epoch's beginning. The leading candidates for the Anthropocene's golden spike are the fallout peaks of 14C (T1/2 = 5730 y) and 239Pu (T1/2 = 24,110 y) from nuclear weapons testing in the 1960s. However, these radionuclides' half-lives may not be long enough for their signals to be observable in the far future and are, thus, not durable. In this regard, here we show the 129I time series record (1957-2007) of the SE-Dome ice core, Greenland. We find that 129I in SE-Dome records almost the entire history of the nuclear age in excellent detail at a time resolution of about four months. More specifically, 129I in SE-Dome reflects signals from nuclear weapons testing in 1958, 1961, and 1962, the Chernobyl Accident in 1986, and various signals from nuclear fuel reprocessing within the same year or a year after. The quantitative relationships between 129I in SE-Dome and these human nuclear activities were established using a numerical model. Similar signals are observed in other records from various environments worldwide, such as sediments, tree rings, and corals. This global ubiquity and synchronicity are comparable to those of the 14C and 239Pu bomb signals, but the much longer half-life of 129I (T1/2 = 15.7 My) makes it a more durable golden spike. For these reasons, the 129I record of the SE-Dome ice core can be considered an excellent candidate for the Anthropocene golden spike.

High-resolution 129I bomb peak profile in an ice core from SE-Dome site, Greenland.

129I in natural archives, such as ice cores, can be used as a proxy for human nuclear activities, age marker, and environmental tracer. Currently, there is only one published record of 129I in ice core (i.e., from Fiescherhorn Glacier, Swiss Alps) and its limited time resolution (1-2 years) prevents the full use of 129I for the mentioned applications. Here we show 129I concentrations in an ice core from SE-Dome, Greenland, covering years 1956-1976 at a time resolution of ∼6 months, the most detailed record to date. Results revealed 129I bomb peaks in years 1959, 1962, and 1963, associated to tests performed by the former Soviet Union, one year prior, in its Novaya Zemlya test site. All 129I bomb peaks were observed in winter (1958.9, 1962.1, and 1963.0), while tritium bomb peaks, another prominent radionuclide associated with nuclear bomb testing, were observed in spring or summer (1959.3, and 1963.6; Iizuka et al., 2017). These results indicate that 129I bomb peaks can be used as annual and seasonal age markers for these years. Furthermore, we found that 129I recorded nuclear fuel reprocessing signals and that these can be potentially used to correct timing of estimated 129I releases during years 1964-1976. Comparisons with other published records of 129I in natural archives showed that 129I can be used as common age marker and tracer for different types of records. Most notably, the 1963 129I bomb peak can be used as common age marker for ice and coral cores, providing the means to reconcile age models and associated trends from the polar and tropical regions, respectively.

Depth profile and mobility of (129)I and (137)Cs in soil originating from the Fukushima Dai-ichi Nuclear Power Plant accident.

The (129)I derived from the FDNPP accident were clearly identified near the surface and showed a trend of rapid decrease with depth. The FDNPP (129)I and (137)Cs was 51.6 ± 1.7 mBq cm(-2) and 88.2 ± 27.1 kBq cm(-2) (average of four cores inventory) respectively. On average, 91% of the FDNPP (129)I existed within the top 5 g cm(-2) and 98% within the top 10 g cm(-2) and average of 100% of the FDNPP (137)Cs existed within the top 5 g cm(-2). From the observation of the temporal variation of depth profiles from the same upland field (Kawauchi village, 20 km away from the FDNPP to the southwest direction), downward migration rates of 0.81 ± 0.32 g cm(-2) yr(-1) for the FDNPP (129)I and 0.19 ± 0.17 g cm(-2) yr(-1) for the FDNPP (137)Cs were estimated. A simple diffusion model was introduced to evaluate the downward mobility of the FDNPP-derived (129)I and (137)Cs. The apparent diffusion coefficients D of 0.0086 ± 0.0034 and 0.0011 ± 0.0010 g(2) cm(-)(4) d(-)(1) were obtained for (129)I and (137)Cs, respectively. These values might be representative for Haplic Gray lowland soils in near the steady state under humid temperate climate.