Application of IKONOS Imagery for Estimating Population Exposure to Landslide Hazard in Banjarmangu Sub District, Central Java, Indonesia

This paper aims to estimate population exposure to landslide hazard in Banjarmangu sub district (46.5 km
²
), Central Java province, Indonesia. A high resolution imagery (i.e. IKONOS) has been used to assess the distribution and size of individual house over the landslide hazard area. Furthermore, houses are divided into two classes based on their size: <;80 m
²
and >80m
²
to estimate population density, for which it is assumed that house with size <;80m
²
is occupied by 3 people, while house with size >80m
²
is occupied by 5 people. Next, error on the population estimation has been estimated by comparing the results with the population census. The results show that 30,927 people (69%) in the study area are exposed by landslide hazard, where 1% and 40% of these people are located in the very high and high landslide hazard prone area, respectively. Based on the error analysis, our population estimation is overestimation of about 3% indicating a good agreement between remote sensing-derived population estimation and field survey. The results of this study can be useful for the local authorities to reduce landslide disaster risk of the study area.