Study of Merapi mountain volcano activity by agglomerative cluster analysis

The data available on the global volcanism program of the Smithsonian Institution website, https://volcano.si.edu/, tells about the activeness of volcanoes around the world, such as Mount Etna in Italy, Mount Aso in Japan, Mount Ruapehu in New Zealand, and also Mount Merapi in the Central Java Province, Indonesia. This data has 24 columns indicating the identity of each volcano, and there are more than five random variables. Here, three of those variables are VEI, start date, and end date in the period of the eruptions from 1750 to 2019. Furthermore, two new random variables called the duration of eruption and repose time is formed, which are obtained by subtracting the end date from the start date of the eruption. In this paper, agglomerative clustering will be applied to three random variables, i.e. eruption duration, repose time, and VEI. Based on the results of the analysis, the optimal number of clusters is 4 with cluster 1 having the most members. This cluster has the characteristics that the eruption has great force but the eruption duration is short. Secondly, the ACF and PACF plots as well as the ADF test were used for the stationary test. The results show that the eruption duration data is stationary and does not show periodicity or seasonality.