Hierarchical Clustering Pada Data Time Series Hotspot Provinsi Riau

Abstract

Land and forest fires become yearly issues in Indonesia, particularly in Riau province. This work applied the hierarchical clustering method for hotspot time series data using average distance formula to analyze hotspots distribution in Riau province. This work focuses on the hotspots distribution in June, July and August for the period 2001-2012. Time series data decomposition was generated to obtain the hotspot occurrence patterns over the period. Hierarchical clustering on the data results 4 clusters at the height of dendrogram about 300. Most of hotspots in these clusters are located in plantation.