A statistical analysis of 295 cloud samples collected at the Puy de Dôme station in France (PUY), covering the period 2001–2018, was conducted using principal component analysis (PCA), agglomerative hierarchical clustering (AHC), and partial least squares (PLS) regression. Our model classified the cloud water samples on the basis of their chemical concentrations and of the dynamical history of their air masses estimated with back-trajectory calculations. The statistical analysis split our dataset into two sets, i.e., the first set characterized by westerly air masses and marine characteristics, with high concentrations of sea salts and the second set having air masses originating from the northeastern sector and the “continental” zone, with high concentrations of potentially anthropogenic ions. It appears from our dataset that the influence of cloud microphysics remains minor at PUY as compared with the impact of the air mass history, i.e., physicochemical processes, such as multiphase reactivity.

中文翻译：

---

根据化学测量和空气质量历史矩阵对在Puy deDôme站（法国）采样的云进行分类

使用主成分分析（PCA），聚集层次聚类（AHC）和偏最小二乘（PLS）对法国Puy deDôme站（PUY）收集的295个云样本进行了统计分析，涵盖了2001-2018年期间）回归。我们的模型基于云水样本的化学浓度和通过反向轨迹计算估算的空气动力学历史来对云水样本进行分类。统计分析将我们的数据集分为两组，即第一组的特征是西风气团和海洋特征，海盐浓度高，第二组的特征是来自东北地区和“大陆”区的气团，高浓度的潜在人为离子。