Advances in Calcium Isotope Geochemistry: Method Development and Application in a Multi-Proxy Investigation of Alaskan Arctic Watersheds.

다국어 초록 (Multilingual Abstract)

The research presented in this dissertation improves double-spike isotope dilution techniques for measuring calcium isotope abundance variations (delta 44/40Ca and delta44/42Ca) by Multi Collector Thermal Ionization Mass Spectrometry (MC-TIMS). In combination with other elemental and isotopic proxies, high precision Ca isotope measurements are applied to the study of mountain and tundra watersheds underlain by continuous permafrost on the North Slope of Alaska to understand how seasonal freeze/thaw cycles of the active layer influence stream geochemistry.
I developed a Monte Carlo error model to optimize analysis of delta 44/40Ca values using a 43Ca-42Ca double-spike. The model maximizes precision and throughput, while minimizing Faraday collector damage and instrumental drift. I implemented the new method to analyze four common Ca isotope standards and found good agreement with model predictions. The global, long-term external reproducibility for the method is +/-0.041‰ (2sigmaSD), which represents a two- to ten-fold improvement over previous Ca isotope methods.
Next, I modified the Monte Carlo error model to optimize simultaneous analysis of delta44/40Ca and delta44/42Ca values using a 48Ca-43Ca double-spike. I used the method to analyze five common Ca isotope standards and found that measured precisions were significantly worse than model predictions, contrary to optimization of the 43Ca-42Ca double-spike technique. Patterns in measured data revealed that mixing of multiple Ca reservoirs on the filament hampered attainment of high-precision results expected from model predictions. Moreover, I observed that mixing had a multi larger effect with the 48Ca-43Ca double-spike compared to the 43Ca- 42Ca double-spike. I employed a filament reservoir-mixing model to examine how different double-spike pairs (i.e., 48Ca- 43Ca, 48Ca-42Ca, 46Ca- 43Ca, and 43Ca-42Ca) magnify propagation of these errors. Measured and modeled data confirmed existence of an "average mass rule," namely that the highest precision data are obtained when the average mass of the double-spike pair is similar to the average mass of the target measurement ratio. By identifying mixing as the dominant source of error for many Ca double-spike pairs, future research can focus on minimizing the effect. At current levels of filament mixing, the best precisions can only be achieved with a 43Ca-42Ca double-spike for delta44/40Ca and a 46Ca-43Ca double-spike for delta44/42Ca.
Finally, I used multiple proxies (major ions, delta34S SO4, delta13CDIC, 87Sr/ 86Sr, and delta44/40Ca) to quantify seasonal controls on the geochemistry of streams draining continuous permafrost on the North Slope of Alaska. The major ion, delta34SSO4 and delta 13CDIC geochemistry indicates the overall dominance of carbonate weathering by carbonic and sulfuric acids, with additional seasonal influences from silicate weathering by carbonic acid, atmospheric deposition, plant decay, and sulfate salt dissolution. While all rivers experience shifts in major ion ratios and delta13CDIC values that coincide with seasonal permafrost thaw and downward expansion of the active layer, streams that predominantly drain bare bedrock in the Brook Range experience a dramatic transition from carbonate weathering in the spring and summer to sulfate salt dissolution in the fall. Riverine delta34S SO4 values confirm the sulfate salts are secondary precipitates rather than primary, bedrock forming marine evaporites. Results suggest that cryoconcentration in soils during the late fall and winter forces precipitation of secondary salts at depth, and exposure during subsequent thaw season yields the observed stream water signals. Streams draining tundra north of the Brooks Range lack definitive evidence for sulfate salt dissolution, presumably because thick peat soils limit the weathering of underlying glacial sediment where the salts are expected to form and dissolve. Appearance of a sulfate salt dissolution signal in tundra streams may correlate with future permafrost degradation. Carbonate weathering dominates riverine 87Sr/86Sr ratios, but the compositional heterogeneity of bedrock and secondary salts limit interpretation of the data. Riverine delta44/40Ca values are higher than those for bedrock, likely due to plant uptake of lighter Ca isotopes. In the tundra watersheds, freshet delta44/40Ca values are 0.10‰ to 0.20‰ lower than summer and fall values. These trends could reflect contributions from plant decay or radiogenic 40Ca contributions from the weathering of old silicate minerals with high K/Ca ratios.