Non-conservative nature of Boron in low salinity Arctic ice and ice melt zones

The Arctic Ocean is one of the most rapidly changing environments on the planet as sea ice extent and thickness have declined extensively over the last 40 years. It is predicted that by 2050, Arctic summers will become mostly ice-free, and the Arctic Ocean will be dominated by seasonally annual, rather than multiyear, sea ice. Arctic sea ice serves as a mediator of biogeochemical processes globally, though the impacts of increased ice melt and water column freshening on Arctic biogeochemistry are uncertain. Specifically, declining sea ice raises significant concerns regarding the future carbon uptake potential of the Arctic and the buffering capacity, or alkalinity, of seawater. Boron (B) is a major element in seawater, and in the form of the borate ion, it serves as the third largest contributor to alkalinity. Boron concentrations in the open ocean are typically conservative and accounted for through relationships with other water components, such as with salinity (S) in the boron to salinity ratio (B/S). Well established B/S ratios have been defined for the open ocean; however, salinity variability can create discrepancies in the open ocean boron corrections for alkalinity. In 2021, work in the marginal ice zone of the Bering and Chukchi Seas revealed non-conservative boron behavior and significant alkalinity system inaccuracies based on the deviation in computed B/S ratios in ice cores and brine. In this study, we investigate the B/S ratio in ice melt zone waters, snow, brine, annual, and multiyear sea ice from the eastern Arctic basin. A total of 169 samples were collected during the onset on melt (May-June 2023) on the ARTofMELT expedition across a range of salinities (2 - 63). High salinity samples (S>29) included 1 lead, 7 brine, 16 under-ice, and 28 open ocean water samples. Low salinity samples (S<29) included 1 brine, 10 snow, and 106 ice core samples. Excluding snow, results indicate deviations from the accepted open ocean B/S ratio (0.1336 mg/kg). For both the entire high salinity sample set and the open ocean subset within it, the B/S average value (0.1304 ± 0.001 mg/kg) was lower. For low salinity samples, the average B/S value (0.1328 ± 0.003) was higher than the high salinity sample value but still lower than the accepted field value. The range of B/S ratios was much larger in low salinity samples (0.1260-0.1425 mg/kg) than high salinity samples (0.1275-0.1350 mg/kg); however, both ranges were significantly smaller than the 2021 B/S ratio range (0.0900-0.1850 mg/kg). The smaller deviation from the accepted B/S ratio in this study resulted in carbon system analysis inaccuracies less than 2 µmol/kg across the entire salinity range. We present the computed B/S ratios and the differences in these datasets using the δ¹⁸O isotopic ratios to understand the heterogeneity of western, annual ice in the marginal ice zone and eastern, multiyear ice in pack ice regions. The marked distinction in the datasets allows potential insight into boron concentrations and the conversion of total alkalinity to carbonate alkalinity across current and future systemic climate-change shifts in the Arctic.