Despite recommendations, uncertainty results are rarely incorporated in Life Cycle Assessment (LCA) studies, especially regarding characterization factors (CF). Part 1 of this study conducted AWARE CF regionalization for Brazil, concluding that the Semiarid region had maximum scarcity values. The goal of this study is to evaluate the uncertainties of regionalized AWARE CF in the Semiarid region. Data used to obtain the AWARE BR CF for Brazil were qualitatively and quantitatively assessed. An adapted Pedigree Matrix was adopted to assess qualitative uncertainties. Classical statistical analysis was used for quantitative uncertainty assessment, and 10,000 Monte Carlo simulations were computed for uncertainty propagation. Qualitative results indicated that the natural flow’s parameter was very uncertain due to poor spatial correlation and low reliability, as it is based on empirical models. Quantitative results showed that water availability data, which had large temporal variability, typical of the Brazilian Semiarid region, was the main responsible for uncertainties in input data. Area uncertainty had a good performance in both qualitative and quantitative assessments. Regarding output data, moderate CF were found to be more uncertain, while more extreme CF exhibited lower variation, corroborating with previous analyses. Moreover, the adoption of shorter datasets led to a reduction in average and standard deviation values for CF. Findings from this study showed two important reasons why the quantitative and qualitative assessments should be conducted simultaneously. The first one was to avoid bias, as availability data and natural flow performed differently in each evaluation. The second one was to confirm results, as the area proved to be very little uncertain in both assessments. An adaptation of Pedigree Matrix and a penalty factor for missing data could be used as a base for quantitative uncertainty parameters for LCIA. Generating SD and k-factor was very positive in terms of results for AWARE method and comparison with other methods. Both indicators had similar results and led to a common conclusion: uncertainties are mainly low and very low for AWARE BR CF in the Semiarid region.

中文翻译：

---

巴西水资源短缺：第 2 部分——区域特征因子的不确定性评估

尽管提出了建议，但不确定性结果很少被纳入生命周期评估 (LCA) 研究，尤其是在特征因子 (CF) 方面。本研究的第 1 部分对巴西进行了 AWARE CF 区域化，得出的结论是半干旱地区具有最大的稀缺价值。本研究的目的是评估半干旱地区区域化 AWARE CF 的不确定性。用于获得巴西 AWARE BR CF 的数据进行了定性和定量评估。采用了改编的谱系矩阵来评估定性的不确定性。经典统计分析用于定量不确定性评估，并计算了 10,000 次蒙特卡罗模拟以用于不确定性传播。定性结果表明，自然流参数是基于经验模型的，由于空间相关性差和可靠性低，因此具有很大的不确定性。定量结果表明，巴西半干旱地区典型的具有较大时间变异性的水资源可用性数据是造成输入数据不确定性的主要原因。面积不确定性在定性和定量评估中都有很好的表现。关于输出数据，发现中度 CF 更具不确定性，而更极端的 CF 表现出较低的变化，这与之前的分析相符。此外，采用更短的数据集导致 CF 的平均和标准偏差值减少。这项研究的结果显示了为什么应该同时进行定量和定性评估的两个重要原因。第一个是避免偏见，因为可用性数据和自然流量在每次评估中的表现都不同。第二个是确认结果，因为在两次评估中都证明该区域几乎没有不确定性。谱系矩阵的改编和缺失数据的惩罚因子可用作 LCIA 定量不确定性参数的基础。就 AWARE 方法的结果以及与其他方法的比较而言，生成 SD 和 k 因子是非常积极的。这两个指标具有相似的结果并得出了一个共同的结论：半干旱地区的 AWARE BR CF 的不确定性主要是低和非常低的。因为可用性数据和自然流量在每次评估中表现不同。第二个是确认结果，因为在两次评估中都证明该区域几乎没有不确定性。谱系矩阵的改编和缺失数据的惩罚因子可用作 LCIA 定量不确定性参数的基础。就 AWARE 方法的结果以及与其他方法的比较而言，生成 SD 和 k 因子是非常积极的。这两个指标具有相似的结果并得出了一个共同的结论：半干旱地区的 AWARE BR CF 的不确定性主要是低和非常低的。因为可用性数据和自然流量在每次评估中表现不同。第二个是确认结果，因为在两次评估中都证明该区域几乎没有不确定性。谱系矩阵的改编和缺失数据的惩罚因子可用作 LCIA 定量不确定性参数的基础。就 AWARE 方法的结果以及与其他方法的比较而言，生成 SD 和 k 因子是非常积极的。这两个指标具有相似的结果并得出了一个共同的结论：半干旱地区的 AWARE BR CF 的不确定性主要是低和非常低的。谱系矩阵的改编和缺失数据的惩罚因子可用作 LCIA 定量不确定性参数的基础。就 AWARE 方法的结果以及与其他方法的比较而言，生成 SD 和 k 因子是非常积极的。这两个指标具有相似的结果并得出了一个共同的结论：半干旱地区的 AWARE BR CF 的不确定性主要是低和非常低的。谱系矩阵的改编和缺失数据的惩罚因子可用作 LCIA 定量不确定性参数的基础。就 AWARE 方法的结果以及与其他方法的比较而言，生成 SD 和 k 因子是非常积极的。这两个指标具有相似的结果并得出了一个共同的结论：半干旱地区的 AWARE BR CF 的不确定性主要是低和非常低的。