Disaster vulnerability prediction modeling using GIS in the Agra Khola watershed, central Nepal

Abstract

Disaster vulnerability analysis is an important concern in the mountainous terrains of Nepal Himalaya. The mountain hill-slope of Agra Khola watershed, central Nepal had suffered from a large number of landslides, debris flows, and floods in 1993. Considering watershed as the most representative disaster site, Geographic Information System (GIS) database after 1993 i.e. year 1993-2007 were spatially integrated to model the disaster vulnerability prediction compiling the extensive field surveyed data and derivatives of Digital Elevation Model (DEM). In the prediction model, input variable were adjusted in interactive manner and the GIS process adopted the statistical method (multivariate) to combine the disaster causative variables that have led to past disaster event. The predicted vulnerability is descretized into five different levels using natural junk break method with some adjustment by overlying the past disaster evidences. The disaster vulnerability has showed the successful modelling utilizing the key indicators as landslide, debris flow and flood. The modelling result found that very high vulnerability areas are mainly confined in upper and middle reaches of the Agra Khola watershed and such spatial localization of very hazards is attributed to the presence of favourable geo-environmental conditions. Also a few localized scattered distributions of very high vulnerable zones are seen to be embedded in the medium to high vulnerable zones in the rest of the area. Detrital materials on the steep hill-slope in the very high to high vulnerable zones may remobilize that may cause catastrophic disaster in the downstream catchments if the extreme weather event again recur in the area. Journal of Nepal Geological Society, 2007, Vol. 36 (Sp. Issue) p.22