Isotope mixing scenarios for: To what extent are the source mixing models accurate: evaluation of the model accuracy and guidelines for the site-specific model selection

We selected 10 types of distinct isotope signatures that can be found in the samples of natural water. Every 3–10 types of hypothetical isotope signatures were conceptually grouped together. There would be 968 possible combinations based on combinatorics theory. However, we needed distinct mixing polygons to facilitate our determination of model capacity in dealing with uncertainties. Therefore, we kept only 240 such groups in the final analysis. Each group was designated with a predefined mixing ratio. After that, we ran all the examined models through these mixing scenarios to obtain the model estimation of the mixing ratios.

我们选取了天然水样本中可检测到的10种不同同位素特征（isotope signatures）。我们将每3至10种假想的同位素特征归为概念性分组。依据组合数学理论，此类分组共计可产生968种可能的组合。但为便于评估模型应对不确定性的能力，我们需要获取互不重叠的混合多边形（mixing polygons），因此在最终分析中仅保留了其中240组。为每组预设了固定的混合比例（mixing ratio）。随后，我们将所有待测试的模型应用于上述混合场景，以得到模型对混合比例的估算结果。