A geostatistical framework for predicting variations in strontium concentrations and isotope ratios in Alaskan rivers
A geostatistical framework for predicting variations in strontium concentrations and isotope ratios in Alaskan rivers
复制标题
DOI：
10.1016/j.chemgeo.2014.08.030
复制
发表时间：
2014-12-11
期刊：
影响因子：
3.9
通讯作者：
Bowen, Gabriel J.
中科院分区：
文献类型：
--
作者：
Bataille, Clement P.;Brennan, Sean R.;Bowen, Gabriel J.
巴塔耶和鲍恩（2012年）开发了一些模型，用于预测区域物源研究中岩石（基岩模型）和河流（流域水模型）中锶 - 87与锶 - 86（Sr - 87/Sr - 86）比值的变化。在此，我们重新审视这些模型的公式和校准，并将其应用于预测阿拉斯加河流中的锶浓度（[Sr]）和Sr - 87/Sr - 86。首先，我们添加了几个新的组件和/或改进措施以解决模型的局限性，包括：1）一个独立的硅质碎屑沉积子模型，2）明确考虑局部尺度上Sr - 87/Sr - 86的变异性，以及3）对预测不确定性进行完全耦合评估。通过对阿拉斯加各地885个Sr - 87/Sr - 86岩石分析数据的测试，新的基岩模型显著提高了在火成岩和沉积岩环境中Sr - 87/Sr - 86的预测准确性。其次，我们开发了一个完全独立的锶化学风化模型，该模型使用来自北半球高纬度地区河流的339个[Sr]分析数据库进行校准，并预测岩石中锶释放速率随岩性、永久冻土覆盖和坡度的空间变化。我们将基岩模型和锶化学风化模型相结合，以预测阿拉斯加河流中的[Sr]和Sr - 87/Sr - 86。通过对61个水样数据集的测试，所得的流域水模型解释了阿拉斯加河流中82%的Sr - 87/Sr - 86变化。我们将流域水模型估算的阿拉斯加径流的平均[Sr]和Sr - 87/Sr - 86与育空河的观测数据进行比较。阿拉斯加地表径流估算的平均[Sr]和Sr - 87/Sr - 86分别为 - 104.3微克/升和0.7098，与育空河的 - 139.3微克/升和0.7137分别有显著差异。这一结果对仅根据大河估算的[Sr]和Sr - 87/Sr - 86值能代表整个地球表面锶风化通量的假设提出了质疑。这项工作的数据产物为估算阿拉斯加区域物源和化学风化研究中岩石和河流的Sr - 87/Sr - 86值提供了另一种基础。（C）2014爱思唯尔B.V.版权所有。
Bataille and Bowen (2012) developed models to predict variations in the ratio of 87-strontium to 86-strontium (Sr-87/Sr-86) in rocks (bedrock model) and rivers (catchment water model) for regional provenance studies. Here, we revisit those models' formulation and calibration and apply them to predict Sr concentrations ([Sr]) and Sr-87/Sr-86 of Alaskan rivers. In a first step, we add several new components and/or improvements to resolve limitations of the model, including: 1) an independent siliciclastic sediment sub-model, 2) an explicit consideration of Sr-87/Sr-86 variability at the local scale, and 3) a fully-coupled assessment of prediction uncertainty. Tested against a compilation of 885 Sr-87/Sr-86 rock analyses across Alaska, the newbedrock model significantly improves Sr-87/Sr-86 prediction accuracy in both igneous and sedimentary settings. In a second step, we develop a fully independent Sr chemical weathering model calibrated using a database of 339 [Sr] analyses from rivers of Northern Hemisphere high-latitude and predicting spatial variations in the rate of Sr release from rocks as a function of lithology, permafrost cover and slope. We combine the bedrock and Sr chemical weathering models to predict [Sr] and Sr-87/Sr-86 in Alaskan rivers. Tested on a dataset of 61 water samples, the resulting catchment water model explains 82% of Sr-87/Sr-86 variations in Alaskan rivers. We compare the average [Sr] and Sr-87/Sr-86 of Alaskan runoff estimated with the catchment water model to observed data of the Yukon River. The estimated average [Sr] and Sr-87/Sr-86 of Alaskan surface runoff -104.3 mu g/L and 0.7098 respectively -differ significantly from those of the Yukon River -139.3 mu g/L and 0.7137 respectively. This result calls into question the assumption that [Sr] and Sr-87/Sr-86 values estimated only from large rivers are representative of the Sr weathering flux from the entire Earth surface. The data products from this work provide an alternative basis for estimating Sr-87/Sr-86 values in rocks and rivers for regional provenance and chemical weathering studies across Alaska. (C) 2014 Elsevier B.V. All rights reserved.