An Evaluation of Three Clustering Procedures for Use in Synoptic Climatological Classification

Laurence S. Kalkstein, Guanri Tan, Jon A. Skindlov

Journal of Climate and Applied Meteorology · 1987 · 515 citations · 14 references

Concepts

TL;DR

Choosing an appropriate clustering method can significantly affect the composition of objective synoptic groupings. The study compares Ward's, average linkage, and centroid clustering to identify the most effective method for synoptic classification. The methods were applied to a temporal synoptic index of days in Mobile, Alabama, to form meteorologically homogeneous units. Average linkage clustering produced the most realistic synoptic groupings, correctly separating extreme weather days, whereas Ward's mixed extremes with less extreme days and centroid produced overly large and single‑day groups.

Abstract

The selection of the proper clustering procedure to use in the development of an objective synoptic methodology may have far-reaching implications on the composition of the final “homogeneous” groupings. The goal of this study is to evaluate three common clustering techniques (Ward's, average linkage, and centroid) to determine which yields the most meaningful synoptic classification. The three clustering procedures were applied to a temporal synoptic index which classified days in Mobile, Alabama into meteorologically homogeneous units. The final meteorological groupings differed widely among the three pressures. Ward's tended to produce groups with relatively similar numbers of days. Thus, many extreme weather days were grouped with less extreme days, and the final meteorological units did not duplicate reality with great precision. The centroid procedure produced one very large group and many single-day groups, yielding unsatisfactory results. The average linkage procedure, which minimizes within-cluster variance, produced the most realistic synoptic groupings and properly combined extreme weather days into distinct meteorological units.

References

14