The residence times of trace elements determined in the surface Arctic Ocean during the 2015 US Arctic GEOTRACES expedition

Data collected during the US Arctic GEOTRACES expedition in 2015 are used to estimate the mean residence time of dissolved trace elements (Fe, Mn, Ni, Cd, Zn, Cu, Pb, V) in surface water with respect to atmospheric deposition. The calculations utilize mixed layer trace element (TE) inventories, aerosol solubility determinations, and estimates of the atmospheric trace element flux into the upper ocean. The trace element flux is estimated by the product of the Be-7 flux (determined by the ocean Be-7 inventory) and the TE/Be-7 ratio of aerosols. This method has been established elsewhere and is tested here by comparing Be-7-derived TE flux to the measured TE accumulation in recently deposited snow. Given the variability in snow and aerosol TE concentration observed over the expedition, and the limited timescale of the observations, agreement between the two methods is reasonable. While there are assumptions in these calculations, the distribution of residence times with respect to atmospheric input across the expedition track informs us of additional sources or sinks for each element. The residence time of dissolved Fe was similar to 20-40 y for most stations. However, several stations that display a longer, oceanographically inconsistent apparent Fe residence time of similar to 300-500 years are likely influenced by additional input from the Transpolar Drift (TPD), which has been shown to convey shelf water properties to the central Arctic. This was seen for Cu, Ni and Zn as well. The flux of Fe delivered by the TPD was similar to 10 nmol/m(2)/d for these stations, an order of magnitude greater than the soluble atmospheric input. On the other hand, V and Pb show a decrease in the apparent residence times within TPD water, suggesting removal of these elements from the source region of the TPD. For Mn, there is no obvious trend in residence time among the stations; however the apparent residence time (400-1400 y) is significantly greater than the similar to 20 y calculated for atmospheric input elsewhere, signifying appreciable input from other sources. It has been suggested that about 90% of Mn input to the Arctic Ocean originates from Arctic rivers, shelf sediments, and coastal erosion. Results here suggest a flux from these sources of similar to 30 nmol/m(2)/d which is significantly greater than the atmospheric input of Mn in the Arctic.