Hierarchical Clustering and Dissimilarity Polygon Analyses. Optimizing the Polish Deposition Network.

The potential re-design of the current deposition monitoring network in Poland was assessed by hierarchical clustering analysis. This statistical method determines the inherent or natural groupings of datasets, and/or to provide a summarization of data into groups using different metrics to assess the (di)similarity. The metrics are based on the correlation, to assess the temporal similarity, the Euclidean distance, to assess the magnitude similarity, and the combination of both. This method was used to assess the areas with similar deposition patters across the country based on measurement and model data for acidic compounds and heavy metals. The analysis clearly identified stations potentially redundant or measuring unique deposition patters and regions that represent the potential location of a single station.