Prediction of susceptible areas of future earthquake induced by landslides in Padang Pariaman district, west Sumatera, Indonesia

Abstract

The island of Sumatera sits on top of Southeast Asia’s tectonics plate, which overrides the subducting Indian and Australian oceanic tectonic plates that convergence obliquely at about 50–60 mm/year. Based on the focal depth and dip angle, sub-duction zones are divided into interplate (megathrust) and intraplate (benioff) zones. Megathrust zone earthquake event occurs at a depth of less than 50 km and while benioff zone earthquake event occurs at a depth of 50–200 km. The earthquake is generally assumed to occur throughout the region and according to landslide experts, the tectonic plate in this megathrust zone produces major earthquakes every 200 years. The new 200 year cycle has predictably come in recent years. Therefore, based on the assumption of past and present landslides are keys to the future landslides. This means that future landslides will occur in similar geologic, geomorphic and hydrologic conditions. This assumption permits an estimation of future occurrences based on historical data of the earthquake with a 7.6 Richter Scale (RS) which struck Padang Pariaman District, West Sumatera Province, Indonesia on Wednesday, September 30, 2009 that caused 375 death buried by the landslides. Other factor such as geographical condition of West Sumatera Province that is normally triggered by earthquake caused landslide. The topography of this region is mountainous where the hills are easily fall down especially after earthquakes and heavy rain. These factors caused severe landslide. This paper predicts the increase of zonation of susceptible areas of major earthquakes occurrence in the future using remote sensing, GIS and calculation of peak ground acceleration (PGA) using global attenuations. The methods are implemented to generate GIS-based landslide susceptibility maps for Padang Pariaman District which include the use of remote sensing data to map existing landslide locations. Other factors that are important in landslide initiation include slope, slope aspect, elevation, land use/land cover, vegetation mass and mapping the ground acceleration map using global attenuations. The frequency ratio models are also used to map landslide susceptibility.