Following an earlier paper, which classified tropical rainfall stations using 34 variables and a single clustering approach, this paper compares the results with those utilizing three other methods. A variety of measures were used to assess ‘performance’ of the methods. Three of the methods (Ward, Average Linkage, and Lance‐Williams) produced very similar results, and even the other one (Centroid) showed a reasonably similar pattern. The similarity in results from the different methods suggests that the station groups obtained are meaningful. However, minor differences are worthy of consideration when climatological interpretation is undertaken at a later stage of the project.
No takes yet. Share an insight, caveat, or question.
H.C. Weinand (1995) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: