Clustering Haze Trajectory of Peatland Fires in Riau Province using K-Means Algorithm

Abstract

Haze is one of the impacts caused by forest and land fires. In order to determine how far the movement of haze from forest and land fires including in peatland, this research aims to determine the haze trajectory patterns from peatland fires in Riau Province in 2015 using HYSPLIT model. Furthermore, the haze trajectory patterns were grouped using the K-Means algorithm. The results show that the trajectory consists of 4887 positions of haze that moves to the northeast and northwest. The clustering results using K-Means show that the largest cluster has 3702 positions of haze and these haze movements are located at the average of height of 23.554 m AGL. Those haze positions were located in Riau, North Sumatra, Malaysia, and Malacca Strait.